id	sid	tid	token	lemma	pos
ejpam-6087	1	1	european	european	PROPN
ejpam-6087	1	2	journal	journal	PROPN
ejpam-6087	1	3	of	of	ADP
ejpam-6087	1	4	pure	pure	ADJ
ejpam-6087	1	5	and	and	CCONJ
ejpam-6087	1	6	applied	applied	ADJ
ejpam-6087	1	7	mathematics	mathematic	NOUN
ejpam-6087	1	8	2025	2025	NUM
ejpam-6087	1	9	,	,	PUNCT
ejpam-6087	1	10	vol	vol	NOUN
ejpam-6087	1	11	.	.	PROPN
ejpam-6087	1	12	18	18	NUM
ejpam-6087	1	13	,	,	PUNCT
ejpam-6087	1	14	issue	issue	NOUN
ejpam-6087	1	15	2	2	NUM
ejpam-6087	1	16	,	,	PUNCT
ejpam-6087	1	17	article	article	NOUN
ejpam-6087	1	18	number	number	NOUN
ejpam-6087	1	19	6087	6087	NUM
ejpam-6087	1	20	issn	issn	VERB
ejpam-6087	1	21	1307	1307	NUM
ejpam-6087	1	22	-	-	SYM
ejpam-6087	1	23	5543	5543	NUM
ejpam-6087	1	24	–	–	PUNCT
ejpam-6087	1	25	ejpam.com	ejpam.com	X
ejpam-6087	1	26	published	publish	VERB
ejpam-6087	1	27	by	by	ADP
ejpam-6087	1	28	new	new	PROPN
ejpam-6087	1	29	york	york	PROPN
ejpam-6087	1	30	business	business	PROPN
ejpam-6087	1	31	global	global	ADJ
ejpam-6087	1	32	towards	towards	ADP
ejpam-6087	1	33	accurate	accurate	ADJ
ejpam-6087	1	34	fake	fake	ADJ
ejpam-6087	1	35	news	news	NOUN
ejpam-6087	1	36	detection	detection	NOUN
ejpam-6087	1	37	:	:	PUNCT
ejpam-6087	1	38	evaluating	evaluate	VERB
ejpam-6087	1	39	machine	machine	NOUN
ejpam-6087	1	40	learning	learn	VERB
ejpam-6087	1	41	approaches	approach	NOUN
ejpam-6087	1	42	and	and	CCONJ
ejpam-6087	1	43	feature	feature	NOUN
ejpam-6087	1	44	selection	selection	NOUN
ejpam-6087	1	45	strategies	strategy	NOUN
ejpam-6087	1	46	mutaz	mutaz	NOUN
ejpam-6087	1	47	a.	a.	PROPN
ejpam-6087	1	48	b.	b.	PROPN
ejpam-6087	1	49	al	al	PROPN
ejpam-6087	1	50	-	-	PUNCT
ejpam-6087	1	51	tarawneh1,∗	tarawneh1,∗	PROPN
ejpam-6087	1	52	,	,	PUNCT
ejpam-6087	1	53	ashraf	ashraf	PROPN
ejpam-6087	1	54	al	al	PROPN
ejpam-6087	1	55	-	-	PUNCT
ejpam-6087	1	56	khresheh2	khresheh2	PROPN
ejpam-6087	1	57	,	,	PUNCT
ejpam-6087	1	58	,	,	PUNCT
ejpam-6087	1	59	omar	omar	PROPN
ejpam-6087	1	60	al	al	PROPN
ejpam-6087	1	61	-	-	PUNCT
ejpam-6087	1	62	irr1	irr1	PROPN
ejpam-6087	1	63	,	,	PUNCT
ejpam-6087	1	64	,	,	PUNCT
ejpam-6087	1	65	ajla	ajla	PROPN
ejpam-6087	1	66	kulaglic1	kulaglic1	PROPN
ejpam-6087	1	67	,	,	PUNCT
ejpam-6087	1	68	,	,	PUNCT
ejpam-6087	1	69	kassem	kassem	PROPN
ejpam-6087	1	70	danach3	danach3	PROPN
ejpam-6087	1	71	,	,	PUNCT
ejpam-6087	1	72	,	,	PUNCT
ejpam-6087	1	73	hassan	hassan	PROPN
ejpam-6087	1	74	kanj1	kanj1	PROPN
ejpam-6087	1	75	,	,	PUNCT
ejpam-6087	1	76	,	,	PUNCT
ejpam-6087	1	77	ghayth	ghayth	NOUN
ejpam-6087	1	78	almahadin4	almahadin4	NOUN
ejpam-6087	1	79	1	1	NUM
ejpam-6087	1	80	college	college	NOUN
ejpam-6087	1	81	of	of	ADP
ejpam-6087	1	82	engineering	engineering	NOUN
ejpam-6087	1	83	and	and	CCONJ
ejpam-6087	1	84	technology	technology	NOUN
ejpam-6087	1	85	,	,	PUNCT
ejpam-6087	1	86	american	american	PROPN
ejpam-6087	1	87	university	university	PROPN
ejpam-6087	1	88	of	of	ADP
ejpam-6087	1	89	the	the	DET
ejpam-6087	1	90	middle	middle	PROPN
ejpam-6087	1	91	east	east	PROPN
ejpam-6087	1	92	,	,	PUNCT
ejpam-6087	1	93	egaila	egaila	PROPN
ejpam-6087	1	94	54200	54200	NUM
ejpam-6087	1	95	,	,	PUNCT
ejpam-6087	1	96	kuwait	kuwait	PROPN
ejpam-6087	1	97	2	2	NUM
ejpam-6087	1	98	computer	computer	NOUN
ejpam-6087	1	99	science	science	NOUN
ejpam-6087	1	100	department	department	PROPN
ejpam-6087	1	101	,	,	PUNCT
ejpam-6087	1	102	tafila	tafila	NOUN
ejpam-6087	1	103	technical	technical	ADJ
ejpam-6087	1	104	university	university	NOUN
ejpam-6087	1	105	,	,	PUNCT
ejpam-6087	1	106	tafila	tafila	NOUN
ejpam-6087	1	107	,	,	PUNCT
ejpam-6087	1	108	jordan	jordan	PROPN
ejpam-6087	1	109	3	3	NUM
ejpam-6087	1	110	basic	basic	ADJ
ejpam-6087	1	111	and	and	CCONJ
ejpam-6087	1	112	applied	apply	VERB
ejpam-6087	1	113	sciences	sciences	PROPN
ejpam-6087	1	114	research	research	NOUN
ejpam-6087	1	115	center	center	NOUN
ejpam-6087	1	116	,	,	PUNCT
ejpam-6087	1	117	al	al	PROPN
ejpam-6087	1	118	maaref	maaref	PROPN
ejpam-6087	1	119	university	university	PROPN
ejpam-6087	1	120	,	,	PUNCT
ejpam-6087	1	121	beirut	beirut	PROPN
ejpam-6087	1	122	,	,	PUNCT
ejpam-6087	1	123	lebanon	lebanon	PROPN
ejpam-6087	1	124	4	4	NUM
ejpam-6087	1	125	data	data	PROPN
ejpam-6087	1	126	science	science	PROPN
ejpam-6087	1	127	department	department	PROPN
ejpam-6087	1	128	,	,	PUNCT
ejpam-6087	1	129	mutah	mutah	PROPN
ejpam-6087	1	130	university	university	PROPN
ejpam-6087	1	131	,	,	PUNCT
ejpam-6087	1	132	karak	karak	PROPN
ejpam-6087	1	133	,	,	PUNCT
ejpam-6087	1	134	61710	61710	NUM
ejpam-6087	1	135	,	,	PUNCT
ejpam-6087	1	136	jordan	jordan	PROPN
ejpam-6087	1	137	abstract	abstract	PROPN
ejpam-6087	1	138	.	.	PUNCT
ejpam-6087	2	1	the	the	DET
ejpam-6087	2	2	rapid	rapid	ADJ
ejpam-6087	2	3	spread	spread	NOUN
ejpam-6087	2	4	of	of	ADP
ejpam-6087	2	5	fake	fake	ADJ
ejpam-6087	2	6	news	news	NOUN
ejpam-6087	2	7	in	in	ADP
ejpam-6087	2	8	the	the	DET
ejpam-6087	2	9	digital	digital	ADJ
ejpam-6087	2	10	age	age	NOUN
ejpam-6087	2	11	poses	pose	VERB
ejpam-6087	2	12	significant	significant	ADJ
ejpam-6087	2	13	challenges	challenge	NOUN
ejpam-6087	2	14	,	,	PUNCT
ejpam-6087	2	15	necessitating	necessitate	VERB
ejpam-6087	2	16	effective	effective	ADJ
ejpam-6087	2	17	detection	detection	NOUN
ejpam-6087	2	18	methods	method	NOUN
ejpam-6087	2	19	.	.	PUNCT
ejpam-6087	3	1	this	this	DET
ejpam-6087	3	2	study	study	NOUN
ejpam-6087	3	3	presents	present	VERB
ejpam-6087	3	4	a	a	DET
ejpam-6087	3	5	comprehensive	comprehensive	ADJ
ejpam-6087	3	6	evaluation	evaluation	NOUN
ejpam-6087	3	7	of	of	ADP
ejpam-6087	3	8	various	various	ADJ
ejpam-6087	3	9	ensemble	ensemble	ADJ
ejpam-6087	3	10	and	and	CCONJ
ejpam-6087	3	11	machine	machine	NOUN
ejpam-6087	3	12	learning	learn	VERB
ejpam-6087	3	13	classifiers	classifier	NOUN
ejpam-6087	3	14	,	,	PUNCT
ejpam-6087	3	15	combined	combine	VERB
ejpam-6087	3	16	with	with	ADP
ejpam-6087	3	17	different	different	ADJ
ejpam-6087	3	18	feature	feature	NOUN
ejpam-6087	3	19	selection	selection	NOUN
ejpam-6087	3	20	techniques	technique	NOUN
ejpam-6087	3	21	,	,	PUNCT
ejpam-6087	3	22	to	to	PART
ejpam-6087	3	23	improve	improve	VERB
ejpam-6087	3	24	the	the	DET
ejpam-6087	3	25	accuracy	accuracy	NOUN
ejpam-6087	3	26	and	and	CCONJ
ejpam-6087	3	27	reliability	reliability	NOUN
ejpam-6087	3	28	of	of	ADP
ejpam-6087	3	29	the	the	DET
ejpam-6087	3	30	detection	detection	NOUN
ejpam-6087	3	31	of	of	ADP
ejpam-6087	3	32	fake	fake	ADJ
ejpam-6087	3	33	news	news	NOUN
ejpam-6087	3	34	.	.	PUNCT
ejpam-6087	4	1	using	use	VERB
ejpam-6087	4	2	the	the	DET
ejpam-6087	4	3	truthseeker	truthseeker	NOUN
ejpam-6087	4	4	dataset	dataset	NOUN
ejpam-6087	4	5	,	,	PUNCT
ejpam-6087	4	6	this	this	DET
ejpam-6087	4	7	research	research	NOUN
ejpam-6087	4	8	examines	examine	VERB
ejpam-6087	4	9	feature	feature	NOUN
ejpam-6087	4	10	selection	selection	NOUN
ejpam-6087	4	11	methods	method	NOUN
ejpam-6087	4	12	such	such	ADJ
ejpam-6087	4	13	as	as	ADP
ejpam-6087	4	14	recursive	recursive	ADJ
ejpam-6087	4	15	feature	feature	NOUN
ejpam-6087	4	16	elimination	elimination	NOUN
ejpam-6087	4	17	(	(	PUNCT
ejpam-6087	4	18	rfe	rfe	NOUN
ejpam-6087	4	19	)	)	PUNCT
ejpam-6087	4	20	,	,	PUNCT
ejpam-6087	4	21	selectkbest	selectkb	ADJ
ejpam-6087	4	22	,	,	PUNCT
ejpam-6087	4	23	principal	principal	ADJ
ejpam-6087	4	24	component	component	NOUN
ejpam-6087	4	25	analysis	analysis	NOUN
ejpam-6087	4	26	(	(	PUNCT
ejpam-6087	4	27	pca	pca	NOUN
ejpam-6087	4	28	)	)	PUNCT
ejpam-6087	4	29	,	,	PUNCT
ejpam-6087	4	30	and	and	CCONJ
ejpam-6087	4	31	genetic	genetic	ADJ
ejpam-6087	4	32	algorithms	algorithm	NOUN
ejpam-6087	4	33	(	(	PUNCT
ejpam-6087	4	34	ga	ga	NOUN
ejpam-6087	4	35	)	)	PUNCT
ejpam-6087	4	36	,	,	PUNCT
ejpam-6087	4	37	analyzing	analyze	VERB
ejpam-6087	4	38	their	their	PRON
ejpam-6087	4	39	impact	impact	NOUN
ejpam-6087	4	40	on	on	ADP
ejpam-6087	4	41	model	model	NOUN
ejpam-6087	4	42	performance	performance	NOUN
ejpam-6087	4	43	.	.	PUNCT
ejpam-6087	5	1	key	key	ADJ
ejpam-6087	5	2	metrics	metric	NOUN
ejpam-6087	5	3	such	such	ADJ
ejpam-6087	5	4	as	as	ADP
ejpam-6087	5	5	accuracy	accuracy	NOUN
ejpam-6087	5	6	,	,	PUNCT
ejpam-6087	5	7	precision	precision	NOUN
ejpam-6087	5	8	,	,	PUNCT
ejpam-6087	5	9	recall	recall	NOUN
ejpam-6087	5	10	,	,	PUNCT
ejpam-6087	5	11	f1	f1	NOUN
ejpam-6087	5	12	score	score	NOUN
ejpam-6087	5	13	,	,	PUNCT
ejpam-6087	5	14	and	and	CCONJ
ejpam-6087	5	15	auc	auc	NOUN
ejpam-6087	5	16	-	-	PUNCT
ejpam-6087	5	17	roc	roc	NOUN
ejpam-6087	5	18	were	be	AUX
ejpam-6087	5	19	used	use	VERB
ejpam-6087	5	20	to	to	PART
ejpam-6087	5	21	assess	assess	VERB
ejpam-6087	5	22	the	the	DET
ejpam-6087	5	23	effectiveness	effectiveness	NOUN
ejpam-6087	5	24	of	of	ADP
ejpam-6087	5	25	each	each	DET
ejpam-6087	5	26	classifier	classifier	NOUN
ejpam-6087	5	27	.	.	PUNCT
ejpam-6087	6	1	the	the	DET
ejpam-6087	6	2	results	result	NOUN
ejpam-6087	6	3	reveal	reveal	VERB
ejpam-6087	6	4	that	that	SCONJ
ejpam-6087	6	5	ensemble	ensemble	ADJ
ejpam-6087	6	6	methods	method	NOUN
ejpam-6087	6	7	,	,	PUNCT
ejpam-6087	6	8	particularly	particularly	ADV
ejpam-6087	6	9	random	random	ADJ
ejpam-6087	6	10	forest	forest	NOUN
ejpam-6087	6	11	(	(	PUNCT
ejpam-6087	6	12	rf	rf	NOUN
ejpam-6087	6	13	)	)	PUNCT
ejpam-6087	6	14	and	and	CCONJ
ejpam-6087	6	15	gradient	gradient	ADJ
ejpam-6087	6	16	boosting	boosting	NOUN
ejpam-6087	6	17	,	,	PUNCT
ejpam-6087	6	18	demonstrate	demonstrate	VERB
ejpam-6087	6	19	superior	superior	ADJ
ejpam-6087	6	20	performance	performance	NOUN
ejpam-6087	6	21	,	,	PUNCT
ejpam-6087	6	22	achieving	achieve	VERB
ejpam-6087	6	23	high	high	ADJ
ejpam-6087	6	24	accuracy	accuracy	NOUN
ejpam-6087	6	25	and	and	CCONJ
ejpam-6087	6	26	auc	auc	NOUN
ejpam-6087	6	27	-	-	PUNCT
ejpam-6087	6	28	roc	roc	NOUN
ejpam-6087	6	29	scores	score	NOUN
ejpam-6087	6	30	.	.	PUNCT
ejpam-6087	7	1	moreover	moreover	ADV
ejpam-6087	7	2	,	,	PUNCT
ejpam-6087	7	3	feature	feature	NOUN
ejpam-6087	7	4	selection	selection	NOUN
ejpam-6087	7	5	techniques	technique	NOUN
ejpam-6087	7	6	like	like	ADP
ejpam-6087	7	7	rfe	rfe	NOUN
ejpam-6087	7	8	and	and	CCONJ
ejpam-6087	7	9	selectkbest	selectkb	ADJ
ejpam-6087	7	10	significantly	significantly	ADV
ejpam-6087	7	11	improve	improve	VERB
ejpam-6087	7	12	model	model	NOUN
ejpam-6087	7	13	outcomes	outcome	NOUN
ejpam-6087	7	14	by	by	ADP
ejpam-6087	7	15	optimizing	optimize	VERB
ejpam-6087	7	16	the	the	DET
ejpam-6087	7	17	feature	feature	NOUN
ejpam-6087	7	18	set	set	NOUN
ejpam-6087	7	19	,	,	PUNCT
ejpam-6087	7	20	while	while	SCONJ
ejpam-6087	7	21	pca	pca	PROPN
ejpam-6087	7	22	is	be	AUX
ejpam-6087	7	23	less	less	ADV
ejpam-6087	7	24	effective	effective	ADJ
ejpam-6087	7	25	in	in	ADP
ejpam-6087	7	26	this	this	DET
ejpam-6087	7	27	context	context	NOUN
ejpam-6087	7	28	.	.	PUNCT
ejpam-6087	8	1	this	this	DET
ejpam-6087	8	2	study	study	NOUN
ejpam-6087	8	3	highlights	highlight	VERB
ejpam-6087	8	4	the	the	DET
ejpam-6087	8	5	importance	importance	NOUN
ejpam-6087	8	6	of	of	ADP
ejpam-6087	8	7	integrating	integrate	VERB
ejpam-6087	8	8	robust	robust	ADJ
ejpam-6087	8	9	classifiers	classifier	NOUN
ejpam-6087	8	10	with	with	ADP
ejpam-6087	8	11	optimal	optimal	ADJ
ejpam-6087	8	12	feature	feature	NOUN
ejpam-6087	8	13	selection	selection	NOUN
ejpam-6087	8	14	methods	method	NOUN
ejpam-6087	8	15	to	to	PART
ejpam-6087	8	16	improve	improve	VERB
ejpam-6087	8	17	the	the	DET
ejpam-6087	8	18	efficacy	efficacy	NOUN
ejpam-6087	8	19	of	of	ADP
ejpam-6087	8	20	fake	fake	ADJ
ejpam-6087	8	21	news	news	NOUN
ejpam-6087	8	22	detection	detection	NOUN
ejpam-6087	8	23	systems	system	NOUN
ejpam-6087	8	24	.	.	PUNCT
ejpam-6087	9	1	2020	2020	NUM
ejpam-6087	9	2	mathematics	mathematic	NOUN
ejpam-6087	9	3	subject	subject	NOUN
ejpam-6087	9	4	classifications	classification	NOUN
ejpam-6087	9	5	:	:	PUNCT
ejpam-6087	9	6	68t01	68t01	NUM
ejpam-6087	9	7	key	key	ADJ
ejpam-6087	9	8	words	word	NOUN
ejpam-6087	9	9	and	and	CCONJ
ejpam-6087	9	10	phrases	phrase	NOUN
ejpam-6087	9	11	:	:	PUNCT
ejpam-6087	9	12	fake	fake	ADJ
ejpam-6087	9	13	news	news	NOUN
ejpam-6087	9	14	detection	detection	NOUN
ejpam-6087	9	15	,	,	PUNCT
ejpam-6087	9	16	machine	machine	NOUN
ejpam-6087	9	17	learning	learning	NOUN
ejpam-6087	9	18	,	,	PUNCT
ejpam-6087	9	19	feature	feature	NOUN
ejpam-6087	9	20	selection	selection	NOUN
ejpam-6087	9	21	,	,	PUNCT
ejpam-6087	9	22	truthseeker	truthseeker	NOUN
ejpam-6087	9	23	∗corresponding	∗corresponde	VERB
ejpam-6087	9	24	author	author	NOUN
ejpam-6087	9	25	.	.	PUNCT
ejpam-6087	10	1	doi	doi	NOUN
ejpam-6087	10	2	:	:	PUNCT
ejpam-6087	10	3	https://doi.org/10.29020/nybg.ejpam.v18i2.6087	https://doi.org/10.29020/nybg.ejpam.v18i2.6087	DET
ejpam-6087	10	4	email	email	NOUN
ejpam-6087	10	5	addresses	address	VERB
ejpam-6087	10	6	:	:	PUNCT
ejpam-6087	10	7	mutaz.al-tarawneh@aum.edu.kw	mutaz.al-tarawneh@aum.edu.kw	CCONJ
ejpam-6087	10	8	(	(	PUNCT
ejpam-6087	10	9	m.	m.	PROPN
ejpam-6087	10	10	al	al	PROPN
ejpam-6087	10	11	-	-	PUNCT
ejpam-6087	10	12	tarawneh	tarawneh	PROPN
ejpam-6087	10	13	)	)	PUNCT
ejpam-6087	10	14	,	,	PUNCT
ejpam-6087	11	1	khashraf@ttu.edu.jo	khashraf@ttu.edu.jo	PROPN
ejpam-6087	11	2	(	(	PUNCT
ejpam-6087	11	3	a.	a.	PROPN
ejpam-6087	11	4	al	al	PROPN
ejpam-6087	11	5	-	-	PUNCT
ejpam-6087	11	6	khresheh	khresheh	PROPN
ejpam-6087	11	7	)	)	PUNCT
ejpam-6087	11	8	,	,	PUNCT
ejpam-6087	11	9	omar.alirr@aum.edu.kw	omar.alirr@aum.edu.kw	PROPN
ejpam-6087	11	10	(	(	PUNCT
ejpam-6087	11	11	o.	o.	NOUN
ejpam-6087	11	12	alirr	alirr	PROPN
ejpam-6087	11	13	)	)	PUNCT
ejpam-6087	11	14	,	,	PUNCT
ejpam-6087	11	15	ajla.kulaglic@aum.edu.kw	ajla.kulaglic@aum.edu.kw	PROPN
ejpam-6087	11	16	(	(	PUNCT
ejpam-6087	11	17	a.	a.	NOUN
ejpam-6087	11	18	kulaglic	kulaglic	PROPN
ejpam-6087	11	19	)	)	PUNCT
ejpam-6087	11	20	,	,	PUNCT
ejpam-6087	11	21	kassem.danach@iul.edu.lb	kassem.danach@iul.edu.lb	PROPN
ejpam-6087	11	22	(	(	PUNCT
ejpam-6087	11	23	k.	k.	NOUN
ejpam-6087	11	24	danach	danach	PROPN
ejpam-6087	11	25	)	)	PUNCT
ejpam-6087	11	26	,	,	PUNCT
ejpam-6087	11	27	hassan.kanj@aum.edu.kw	hassan.kanj@aum.edu.kw	PROPN
ejpam-6087	11	28	(	(	PUNCT
ejpam-6087	11	29	h.	h.	PROPN
ejpam-6087	11	30	kanj	kanj	PROPN
ejpam-6087	11	31	)	)	PUNCT
ejpam-6087	11	32	,	,	PUNCT
ejpam-6087	11	33	ghayth.mahadin@mutah.edu.jo	ghayth.mahadin@mutah.edu.jo	PROPN
ejpam-6087	11	34	(	(	PUNCT
ejpam-6087	11	35	g.	g.	NOUN
ejpam-6087	11	36	almahadin	almahadin	NOUN
ejpam-6087	11	37	)	)	PUNCT
ejpam-6087	11	38	https://www.ejpam.com	https://www.ejpam.com	NOUN
ejpam-6087	11	39	1	1	NUM
ejpam-6087	11	40	copyright	copyright	NOUN
ejpam-6087	11	41	:	:	PUNCT
ejpam-6087	11	42	©	©	PROPN
ejpam-6087	11	43	2025	2025	NUM
ejpam-6087	11	44	the	the	DET
ejpam-6087	11	45	author(s	author(s	NOUN
ejpam-6087	11	46	)	)	PUNCT
ejpam-6087	11	47	.	.	PUNCT
ejpam-6087	12	1	(	(	PUNCT
ejpam-6087	12	2	cc	cc	NOUN
ejpam-6087	12	3	by	by	ADP
ejpam-6087	12	4	-	-	PUNCT
ejpam-6087	12	5	nc	nc	PROPN
ejpam-6087	12	6	4.0	4.0	NUM
ejpam-6087	12	7	)	)	PUNCT
ejpam-6087	12	8	m.	m.	NOUN
ejpam-6087	12	9	al	al	PROPN
ejpam-6087	12	10	-	-	PUNCT
ejpam-6087	12	11	tarawneh	tarawneh	PROPN
ejpam-6087	12	12	et	et	PROPN
ejpam-6087	12	13	al	al	PROPN
ejpam-6087	12	14	.	.	PUNCT
ejpam-6087	12	15	/	/	SYM
ejpam-6087	12	16	eur	eur	PROPN
ejpam-6087	12	17	.	.	PUNCT
ejpam-6087	13	1	j.	j.	PROPN
ejpam-6087	13	2	pure	pure	PROPN
ejpam-6087	13	3	appl	appl	PROPN
ejpam-6087	13	4	.	.	PROPN
ejpam-6087	13	5	math	math	PROPN
ejpam-6087	13	6	,	,	PUNCT
ejpam-6087	13	7	18	18	NUM
ejpam-6087	13	8	(	(	PUNCT
ejpam-6087	13	9	2	2	NUM
ejpam-6087	13	10	)	)	PUNCT
ejpam-6087	13	11	(	(	PUNCT
ejpam-6087	13	12	2025	2025	NUM
ejpam-6087	13	13	)	)	PUNCT
ejpam-6087	13	14	,	,	PUNCT
ejpam-6087	13	15	6087	6087	NUM
ejpam-6087	13	16	2	2	NUM
ejpam-6087	13	17	of	of	ADP
ejpam-6087	13	18	39	39	NUM
ejpam-6087	13	19	1	1	NUM
ejpam-6087	13	20	.	.	PUNCT
ejpam-6087	13	21	introduction	introduction	NOUN
ejpam-6087	13	22	the	the	DET
ejpam-6087	13	23	advent	advent	NOUN
ejpam-6087	13	24	of	of	ADP
ejpam-6087	13	25	the	the	DET
ejpam-6087	13	26	digital	digital	ADJ
ejpam-6087	13	27	age	age	NOUN
ejpam-6087	13	28	has	have	AUX
ejpam-6087	13	29	transformed	transform	VERB
ejpam-6087	13	30	the	the	DET
ejpam-6087	13	31	way	way	NOUN
ejpam-6087	13	32	information	information	NOUN
ejpam-6087	13	33	is	be	AUX
ejpam-6087	13	34	accessed	access	VERB
ejpam-6087	13	35	and	and	CCONJ
ejpam-6087	13	36	disseminated	disseminate	VERB
ejpam-6087	13	37	,	,	PUNCT
ejpam-6087	13	38	leading	lead	VERB
ejpam-6087	13	39	to	to	ADP
ejpam-6087	13	40	unprecedented	unprecedented	ADJ
ejpam-6087	13	41	opportunities	opportunity	NOUN
ejpam-6087	13	42	for	for	ADP
ejpam-6087	13	43	knowledge	knowledge	NOUN
ejpam-6087	13	44	sharing	sharing	NOUN
ejpam-6087	13	45	.	.	PUNCT
ejpam-6087	14	1	however	however	ADV
ejpam-6087	14	2	,	,	PUNCT
ejpam-6087	14	3	this	this	DET
ejpam-6087	14	4	increased	increase	VERB
ejpam-6087	14	5	accessibility	accessibility	NOUN
ejpam-6087	14	6	has	have	AUX
ejpam-6087	14	7	also	also	ADV
ejpam-6087	14	8	facilitated	facilitate	VERB
ejpam-6087	14	9	the	the	DET
ejpam-6087	14	10	widespread	widespread	ADJ
ejpam-6087	14	11	circulation	circulation	NOUN
ejpam-6087	14	12	of	of	ADP
ejpam-6087	14	13	misinformation	misinformation	NOUN
ejpam-6087	14	14	and	and	CCONJ
ejpam-6087	14	15	fake	fake	ADJ
ejpam-6087	14	16	news	news	NOUN
ejpam-6087	14	17	.	.	PUNCT
ejpam-6087	15	1	fake	fake	ADJ
ejpam-6087	15	2	news	news	NOUN
ejpam-6087	15	3	,	,	PUNCT
ejpam-6087	15	4	characterized	characterize	VERB
ejpam-6087	15	5	as	as	ADP
ejpam-6087	15	6	deliberately	deliberately	ADV
ejpam-6087	15	7	false	false	ADJ
ejpam-6087	15	8	or	or	CCONJ
ejpam-6087	15	9	misleading	misleading	ADJ
ejpam-6087	15	10	content	content	NOUN
ejpam-6087	15	11	presented	present	VERB
ejpam-6087	15	12	in	in	ADP
ejpam-6087	15	13	the	the	DET
ejpam-6087	15	14	guise	guise	NOUN
ejpam-6087	15	15	of	of	ADP
ejpam-6087	15	16	authentic	authentic	ADJ
ejpam-6087	15	17	news	news	NOUN
ejpam-6087	15	18	,	,	PUNCT
ejpam-6087	15	19	has	have	AUX
ejpam-6087	15	20	emerged	emerge	VERB
ejpam-6087	15	21	as	as	ADP
ejpam-6087	15	22	a	a	DET
ejpam-6087	15	23	critical	critical	ADJ
ejpam-6087	15	24	challenge	challenge	NOUN
ejpam-6087	15	25	with	with	ADP
ejpam-6087	15	26	far	far	ADV
ejpam-6087	15	27	-	-	PUNCT
ejpam-6087	15	28	reaching	reach	VERB
ejpam-6087	15	29	consequences	consequence	NOUN
ejpam-6087	15	30	.	.	PUNCT
ejpam-6087	16	1	it	it	PRON
ejpam-6087	16	2	influences	influence	VERB
ejpam-6087	16	3	public	public	ADJ
ejpam-6087	16	4	perception	perception	NOUN
ejpam-6087	16	5	,	,	PUNCT
ejpam-6087	16	6	shapes	shape	VERB
ejpam-6087	16	7	political	political	ADJ
ejpam-6087	16	8	decisions	decision	NOUN
ejpam-6087	16	9	,	,	PUNCT
ejpam-6087	16	10	and	and	CCONJ
ejpam-6087	16	11	can	can	AUX
ejpam-6087	16	12	even	even	ADV
ejpam-6087	16	13	lead	lead	VERB
ejpam-6087	16	14	to	to	ADP
ejpam-6087	16	15	social	social	ADJ
ejpam-6087	16	16	instability	instability	NOUN
ejpam-6087	16	17	.	.	PUNCT
ejpam-6087	17	1	the	the	DET
ejpam-6087	17	2	difficulty	difficulty	NOUN
ejpam-6087	17	3	of	of	ADP
ejpam-6087	17	4	distinguishing	distinguish	VERB
ejpam-6087	17	5	genuine	genuine	ADJ
ejpam-6087	17	6	news	news	NOUN
ejpam-6087	17	7	from	from	ADP
ejpam-6087	17	8	fabricated	fabricate	VERB
ejpam-6087	17	9	stories	story	NOUN
ejpam-6087	17	10	is	be	AUX
ejpam-6087	17	11	exacerbated	exacerbate	VERB
ejpam-6087	17	12	by	by	ADP
ejpam-6087	17	13	the	the	DET
ejpam-6087	17	14	rapid	rapid	ADJ
ejpam-6087	17	15	and	and	CCONJ
ejpam-6087	17	16	extensive	extensive	ADJ
ejpam-6087	17	17	distribution	distribution	NOUN
ejpam-6087	17	18	of	of	ADP
ejpam-6087	17	19	information	information	NOUN
ejpam-6087	17	20	through	through	ADP
ejpam-6087	17	21	social	social	ADJ
ejpam-6087	17	22	media	medium	NOUN
ejpam-6087	17	23	and	and	CCONJ
ejpam-6087	17	24	digital	digital	ADJ
ejpam-6087	17	25	platforms	platform	NOUN
ejpam-6087	18	1	[	[	X
ejpam-6087	18	2	1	1	NUM
ejpam-6087	18	3	]	]	PUNCT
ejpam-6087	18	4	.	.	PUNCT
ejpam-6087	19	1	the	the	DET
ejpam-6087	19	2	repercussions	repercussion	NOUN
ejpam-6087	19	3	of	of	ADP
ejpam-6087	19	4	fake	fake	ADJ
ejpam-6087	19	5	news	news	NOUN
ejpam-6087	19	6	extend	extend	NOUN
ejpam-6087	19	7	into	into	ADP
ejpam-6087	19	8	the	the	DET
ejpam-6087	19	9	physical	physical	ADJ
ejpam-6087	19	10	world	world	NOUN
ejpam-6087	19	11	,	,	PUNCT
ejpam-6087	19	12	eroding	erode	VERB
ejpam-6087	19	13	trust	trust	NOUN
ejpam-6087	19	14	in	in	ADP
ejpam-6087	19	15	traditional	traditional	ADJ
ejpam-6087	19	16	media	medium	NOUN
ejpam-6087	19	17	and	and	CCONJ
ejpam-6087	19	18	impacting	impact	VERB
ejpam-6087	19	19	real	real	ADJ
ejpam-6087	19	20	-	-	PUNCT
ejpam-6087	19	21	world	world	NOUN
ejpam-6087	19	22	events	event	NOUN
ejpam-6087	19	23	.	.	PUNCT
ejpam-6087	20	1	for	for	ADP
ejpam-6087	20	2	example	example	NOUN
ejpam-6087	20	3	,	,	PUNCT
ejpam-6087	20	4	the	the	DET
ejpam-6087	20	5	2016	2016	NUM
ejpam-6087	20	6	us	us	PROPN
ejpam-6087	20	7	presidential	presidential	ADJ
ejpam-6087	20	8	election	election	NOUN
ejpam-6087	20	9	saw	see	VERB
ejpam-6087	20	10	a	a	DET
ejpam-6087	20	11	proliferation	proliferation	NOUN
ejpam-6087	20	12	of	of	ADP
ejpam-6087	20	13	misleading	mislead	VERB
ejpam-6087	20	14	stories	story	NOUN
ejpam-6087	20	15	that	that	PRON
ejpam-6087	20	16	may	may	AUX
ejpam-6087	20	17	have	have	AUX
ejpam-6087	20	18	affected	affect	VERB
ejpam-6087	20	19	voter	voter	NOUN
ejpam-6087	20	20	behavior	behavior	NOUN
ejpam-6087	20	21	and	and	CCONJ
ejpam-6087	20	22	the	the	DET
ejpam-6087	20	23	final	final	ADJ
ejpam-6087	20	24	outcome	outcome	NOUN
ejpam-6087	20	25	[	[	X
ejpam-6087	20	26	2	2	NUM
ejpam-6087	20	27	]	]	PUNCT
ejpam-6087	20	28	.	.	PUNCT
ejpam-6087	21	1	similarly	similarly	ADV
ejpam-6087	21	2	,	,	PUNCT
ejpam-6087	21	3	during	during	ADP
ejpam-6087	21	4	the	the	DET
ejpam-6087	21	5	2019	2019	NUM
ejpam-6087	21	6	indian	indian	ADJ
ejpam-6087	21	7	general	general	ADJ
ejpam-6087	21	8	election	election	NOUN
ejpam-6087	21	9	,	,	PUNCT
ejpam-6087	21	10	targeted	target	VERB
ejpam-6087	21	11	misinformation	misinformation	NOUN
ejpam-6087	21	12	campaigns	campaign	NOUN
ejpam-6087	21	13	spread	spread	VERB
ejpam-6087	21	14	false	false	ADJ
ejpam-6087	21	15	narratives	narrative	NOUN
ejpam-6087	21	16	and	and	CCONJ
ejpam-6087	21	17	incited	incited	ADJ
ejpam-6087	21	18	division	division	NOUN
ejpam-6087	21	19	among	among	ADP
ejpam-6087	21	20	voters	voter	NOUN
ejpam-6087	21	21	[	[	X
ejpam-6087	21	22	3	3	NUM
ejpam-6087	21	23	]	]	PUNCT
ejpam-6087	21	24	.	.	PUNCT
ejpam-6087	22	1	the	the	DET
ejpam-6087	22	2	covid-19	covid-19	PROPN
ejpam-6087	22	3	pandemic	pandemic	NOUN
ejpam-6087	22	4	further	far	ADV
ejpam-6087	22	5	highlighted	highlight	VERB
ejpam-6087	22	6	the	the	DET
ejpam-6087	22	7	dangers	danger	NOUN
ejpam-6087	22	8	of	of	ADP
ejpam-6087	22	9	false	false	ADJ
ejpam-6087	22	10	news	news	NOUN
ejpam-6087	22	11	,	,	PUNCT
ejpam-6087	22	12	with	with	ADP
ejpam-6087	22	13	widespread	widespread	ADJ
ejpam-6087	22	14	misinformation	misinformation	NOUN
ejpam-6087	22	15	about	about	ADP
ejpam-6087	22	16	the	the	DET
ejpam-6087	22	17	virus	virus	NOUN
ejpam-6087	22	18	,	,	PUNCT
ejpam-6087	22	19	treatments	treatment	NOUN
ejpam-6087	22	20	,	,	PUNCT
ejpam-6087	22	21	and	and	CCONJ
ejpam-6087	22	22	vaccines	vaccine	NOUN
ejpam-6087	22	23	creating	create	VERB
ejpam-6087	22	24	significant	significant	ADJ
ejpam-6087	22	25	public	public	ADJ
ejpam-6087	22	26	health	health	NOUN
ejpam-6087	22	27	risks	risk	NOUN
ejpam-6087	22	28	[	[	X
ejpam-6087	22	29	4	4	NUM
ejpam-6087	22	30	]	]	PUNCT
ejpam-6087	22	31	.	.	PUNCT
ejpam-6087	23	1	these	these	DET
ejpam-6087	23	2	instances	instance	NOUN
ejpam-6087	23	3	underscore	underscore	VERB
ejpam-6087	23	4	the	the	DET
ejpam-6087	23	5	pressing	press	VERB
ejpam-6087	23	6	need	need	NOUN
ejpam-6087	23	7	for	for	ADP
ejpam-6087	23	8	robust	robust	ADJ
ejpam-6087	23	9	strategies	strategy	NOUN
ejpam-6087	23	10	to	to	PART
ejpam-6087	23	11	identify	identify	VERB
ejpam-6087	23	12	and	and	CCONJ
ejpam-6087	23	13	mitigate	mitigate	VERB
ejpam-6087	23	14	fake	fake	ADJ
ejpam-6087	23	15	news	news	NOUN
ejpam-6087	23	16	.	.	PUNCT
ejpam-6087	24	1	machine	machine	NOUN
ejpam-6087	24	2	learning	learning	NOUN
ejpam-6087	24	3	(	(	PUNCT
ejpam-6087	24	4	ml	ml	NOUN
ejpam-6087	24	5	)	)	PUNCT
ejpam-6087	24	6	has	have	AUX
ejpam-6087	24	7	become	become	VERB
ejpam-6087	24	8	a	a	DET
ejpam-6087	24	9	cornerstone	cornerstone	NOUN
ejpam-6087	24	10	in	in	ADP
ejpam-6087	24	11	addressing	address	VERB
ejpam-6087	24	12	this	this	DET
ejpam-6087	24	13	challenge	challenge	NOUN
ejpam-6087	24	14	,	,	PUNCT
ejpam-6087	24	15	offering	offer	VERB
ejpam-6087	24	16	advanced	advanced	ADJ
ejpam-6087	24	17	methods	method	NOUN
ejpam-6087	24	18	to	to	PART
ejpam-6087	24	19	identify	identify	VERB
ejpam-6087	24	20	and	and	CCONJ
ejpam-6087	24	21	classify	classify	VERB
ejpam-6087	24	22	fake	fake	ADJ
ejpam-6087	24	23	news	news	NOUN
ejpam-6087	24	24	.	.	PUNCT
ejpam-6087	25	1	by	by	ADP
ejpam-6087	25	2	analyzing	analyze	VERB
ejpam-6087	25	3	extensive	extensive	ADJ
ejpam-6087	25	4	datasets	dataset	NOUN
ejpam-6087	25	5	with	with	ADP
ejpam-6087	25	6	sophisticated	sophisticated	ADJ
ejpam-6087	25	7	algorithms	algorithm	NOUN
ejpam-6087	25	8	,	,	PUNCT
ejpam-6087	25	9	ml	ml	ADP
ejpam-6087	25	10	models	model	NOUN
ejpam-6087	25	11	can	can	AUX
ejpam-6087	25	12	detect	detect	VERB
ejpam-6087	25	13	patterns	pattern	NOUN
ejpam-6087	25	14	and	and	CCONJ
ejpam-6087	25	15	attributes	attribute	VERB
ejpam-6087	25	16	indicative	indicative	ADJ
ejpam-6087	25	17	of	of	ADP
ejpam-6087	25	18	misinformation	misinformation	NOUN
ejpam-6087	25	19	[	[	X
ejpam-6087	25	20	5	5	NUM
ejpam-6087	25	21	]	]	PUNCT
ejpam-6087	25	22	.	.	PUNCT
ejpam-6087	26	1	the	the	DET
ejpam-6087	26	2	fake	fake	ADJ
ejpam-6087	26	3	news	news	NOUN
ejpam-6087	26	4	detection	detection	NOUN
ejpam-6087	26	5	process	process	NOUN
ejpam-6087	26	6	typically	typically	ADV
ejpam-6087	26	7	involves	involve	VERB
ejpam-6087	26	8	several	several	ADJ
ejpam-6087	26	9	stages	stage	NOUN
ejpam-6087	26	10	:	:	PUNCT
ejpam-6087	26	11	data	data	NOUN
ejpam-6087	26	12	collection	collection	NOUN
ejpam-6087	26	13	,	,	PUNCT
ejpam-6087	26	14	feature	feature	NOUN
ejpam-6087	26	15	extraction	extraction	NOUN
ejpam-6087	26	16	,	,	PUNCT
ejpam-6087	26	17	model	model	NOUN
ejpam-6087	26	18	training	training	NOUN
ejpam-6087	26	19	,	,	PUNCT
ejpam-6087	26	20	and	and	CCONJ
ejpam-6087	26	21	evaluation	evaluation	NOUN
ejpam-6087	26	22	.	.	PUNCT
ejpam-6087	27	1	among	among	ADP
ejpam-6087	27	2	these	these	PRON
ejpam-6087	27	3	,	,	PUNCT
ejpam-6087	27	4	feature	feature	NOUN
ejpam-6087	27	5	selection	selection	NOUN
ejpam-6087	27	6	plays	play	VERB
ejpam-6087	27	7	a	a	DET
ejpam-6087	27	8	pivotal	pivotal	ADJ
ejpam-6087	27	9	role	role	NOUN
ejpam-6087	27	10	in	in	ADP
ejpam-6087	27	11	pinpointing	pinpoint	VERB
ejpam-6087	27	12	the	the	DET
ejpam-6087	27	13	most	most	ADV
ejpam-6087	27	14	relevant	relevant	ADJ
ejpam-6087	27	15	attributes	attribute	NOUN
ejpam-6087	27	16	in	in	ADP
ejpam-6087	27	17	the	the	DET
ejpam-6087	27	18	dataset	dataset	NOUN
ejpam-6087	27	19	,	,	PUNCT
ejpam-6087	27	20	thereby	thereby	ADV
ejpam-6087	27	21	enhancing	enhance	VERB
ejpam-6087	27	22	model	model	NOUN
ejpam-6087	27	23	accuracy	accuracy	NOUN
ejpam-6087	27	24	and	and	CCONJ
ejpam-6087	27	25	efficiency	efficiency	NOUN
ejpam-6087	27	26	.	.	PUNCT
ejpam-6087	28	1	feature	feature	NOUN
ejpam-6087	28	2	selection	selection	NOUN
ejpam-6087	28	3	is	be	AUX
ejpam-6087	28	4	crucial	crucial	ADJ
ejpam-6087	28	5	to	to	PART
ejpam-6087	28	6	improve	improve	VERB
ejpam-6087	28	7	the	the	DET
ejpam-6087	28	8	performance	performance	NOUN
ejpam-6087	28	9	of	of	ADP
ejpam-6087	28	10	the	the	DET
ejpam-6087	28	11	ml	ml	PROPN
ejpam-6087	28	12	model	model	NOUN
ejpam-6087	28	13	,	,	PUNCT
ejpam-6087	28	14	as	as	SCONJ
ejpam-6087	28	15	it	it	PRON
ejpam-6087	28	16	reduces	reduce	VERB
ejpam-6087	28	17	complexity	complexity	NOUN
ejpam-6087	28	18	and	and	CCONJ
ejpam-6087	28	19	increases	increase	VERB
ejpam-6087	28	20	interpretability	interpretability	NOUN
ejpam-6087	28	21	[	[	X
ejpam-6087	28	22	6	6	NUM
ejpam-6087	28	23	]	]	PUNCT
ejpam-6087	28	24	.	.	PUNCT
ejpam-6087	29	1	several	several	ADJ
ejpam-6087	29	2	techniques	technique	NOUN
ejpam-6087	29	3	have	have	AUX
ejpam-6087	29	4	been	be	AUX
ejpam-6087	29	5	developed	develop	VERB
ejpam-6087	29	6	for	for	ADP
ejpam-6087	29	7	this	this	DET
ejpam-6087	29	8	purpose	purpose	NOUN
ejpam-6087	29	9	.	.	PUNCT
ejpam-6087	30	1	recursive	recursive	ADJ
ejpam-6087	30	2	feature	feature	NOUN
ejpam-6087	30	3	elimination	elimination	NOUN
ejpam-6087	30	4	(	(	PUNCT
ejpam-6087	30	5	rfe	rfe	NOUN
ejpam-6087	30	6	)	)	PUNCT
ejpam-6087	30	7	incrementally	incrementally	ADV
ejpam-6087	30	8	removes	remove	VERB
ejpam-6087	30	9	less	less	ADV
ejpam-6087	30	10	relevant	relevant	ADJ
ejpam-6087	30	11	features	feature	NOUN
ejpam-6087	30	12	based	base	VERB
ejpam-6087	30	13	on	on	ADP
ejpam-6087	30	14	model	model	NOUN
ejpam-6087	30	15	weights	weight	NOUN
ejpam-6087	30	16	to	to	PART
ejpam-6087	30	17	refine	refine	VERB
ejpam-6087	30	18	the	the	DET
ejpam-6087	30	19	set	set	NOUN
ejpam-6087	30	20	of	of	ADP
ejpam-6087	30	21	features	feature	NOUN
ejpam-6087	30	22	[	[	X
ejpam-6087	30	23	7	7	NUM
ejpam-6087	30	24	]	]	PUNCT
ejpam-6087	30	25	.	.	PUNCT
ejpam-6087	31	1	selectkbest	selectkb	ADJ
ejpam-6087	31	2	applies	apply	VERB
ejpam-6087	31	3	statistical	statistical	ADJ
ejpam-6087	31	4	tests	test	NOUN
ejpam-6087	31	5	to	to	PART
ejpam-6087	31	6	identify	identify	VERB
ejpam-6087	31	7	the	the	DET
ejpam-6087	31	8	key	key	ADJ
ejpam-6087	31	9	features	feature	NOUN
ejpam-6087	31	10	that	that	PRON
ejpam-6087	31	11	most	most	ADV
ejpam-6087	31	12	closely	closely	ADV
ejpam-6087	31	13	correlated	correlate	VERB
ejpam-6087	31	14	with	with	ADP
ejpam-6087	31	15	the	the	DET
ejpam-6087	31	16	target	target	NOUN
ejpam-6087	31	17	variable	variable	NOUN
ejpam-6087	31	18	[	[	X
ejpam-6087	31	19	8	8	NUM
ejpam-6087	31	20	]	]	PUNCT
ejpam-6087	31	21	.	.	PUNCT
ejpam-6087	32	1	principal	principal	ADJ
ejpam-6087	32	2	component	component	NOUN
ejpam-6087	32	3	analysis	analysis	NOUN
ejpam-6087	32	4	(	(	PUNCT
ejpam-6087	32	5	pca	pca	NOUN
ejpam-6087	32	6	)	)	PUNCT
ejpam-6087	32	7	transforms	transform	VERB
ejpam-6087	32	8	existing	exist	VERB
ejpam-6087	32	9	features	feature	NOUN
ejpam-6087	32	10	into	into	ADP
ejpam-6087	32	11	a	a	DET
ejpam-6087	32	12	smaller	small	ADJ
ejpam-6087	32	13	set	set	NOUN
ejpam-6087	32	14	of	of	ADP
ejpam-6087	32	15	components	component	NOUN
ejpam-6087	32	16	,	,	PUNCT
ejpam-6087	32	17	retaining	retain	VERB
ejpam-6087	32	18	maximum	maximum	ADJ
ejpam-6087	32	19	variance	variance	NOUN
ejpam-6087	32	20	[	[	X
ejpam-6087	32	21	9	9	NUM
ejpam-6087	32	22	]	]	PUNCT
ejpam-6087	32	23	.	.	PUNCT
ejpam-6087	33	1	furthermore	furthermore	ADV
ejpam-6087	33	2	,	,	PUNCT
ejpam-6087	33	3	genetic	genetic	ADJ
ejpam-6087	33	4	algorithms	algorithm	NOUN
ejpam-6087	33	5	(	(	PUNCT
ejpam-6087	33	6	gas	gas	NOUN
ejpam-6087	33	7	)	)	PUNCT
ejpam-6087	33	8	offer	offer	VERB
ejpam-6087	33	9	a	a	DET
ejpam-6087	33	10	global	global	ADJ
ejpam-6087	33	11	optimization	optimization	NOUN
ejpam-6087	33	12	approach	approach	NOUN
ejpam-6087	33	13	by	by	ADP
ejpam-6087	33	14	mimicking	mimic	VERB
ejpam-6087	33	15	natural	natural	ADJ
ejpam-6087	33	16	selection	selection	NOUN
ejpam-6087	33	17	processes	process	NOUN
ejpam-6087	33	18	to	to	PART
ejpam-6087	33	19	identify	identify	VERB
ejpam-6087	33	20	the	the	DET
ejpam-6087	33	21	optimal	optimal	ADJ
ejpam-6087	33	22	set	set	NOUN
ejpam-6087	33	23	of	of	ADP
ejpam-6087	33	24	characteristics	characteristic	NOUN
ejpam-6087	33	25	in	in	ADP
ejpam-6087	33	26	complex	complex	ADJ
ejpam-6087	33	27	search	search	NOUN
ejpam-6087	33	28	spaces	space	NOUN
ejpam-6087	33	29	[	[	X
ejpam-6087	33	30	10	10	NUM
ejpam-6087	33	31	]	]	PUNCT
ejpam-6087	33	32	.	.	PUNCT
ejpam-6087	34	1	this	this	DET
ejpam-6087	34	2	research	research	NOUN
ejpam-6087	34	3	presents	present	VERB
ejpam-6087	34	4	a	a	DET
ejpam-6087	34	5	comprehensive	comprehensive	ADJ
ejpam-6087	34	6	evaluation	evaluation	NOUN
ejpam-6087	34	7	of	of	ADP
ejpam-6087	34	8	machine	machine	NOUN
ejpam-6087	34	9	learning	learning	NOUN
ejpam-6087	34	10	methods	method	NOUN
ejpam-6087	34	11	for	for	ADP
ejpam-6087	34	12	detecting	detect	VERB
ejpam-6087	34	13	fake	fake	ADJ
ejpam-6087	34	14	news	news	NOUN
ejpam-6087	34	15	.	.	PUNCT
ejpam-6087	35	1	explores	explore	VERB
ejpam-6087	35	2	the	the	DET
ejpam-6087	35	3	performance	performance	NOUN
ejpam-6087	35	4	of	of	ADP
ejpam-6087	35	5	several	several	ADJ
ejpam-6087	35	6	individual	individual	ADJ
ejpam-6087	35	7	classifiers	classifier	NOUN
ejpam-6087	35	8	,	,	PUNCT
ejpam-6087	35	9	including	include	VERB
ejpam-6087	35	10	decision	decision	NOUN
ejpam-6087	35	11	trees	tree	NOUN
ejpam-6087	35	12	(	(	PUNCT
ejpam-6087	35	13	dt	dt	PROPN
ejpam-6087	35	14	)	)	PUNCT
ejpam-6087	35	15	,	,	PUNCT
ejpam-6087	35	16	näıve	näıve	NUM
ejpam-6087	35	17	bayes	bayes	PROPN
ejpam-6087	35	18	(	(	PUNCT
ejpam-6087	35	19	nb	nb	PROPN
ejpam-6087	35	20	)	)	PUNCT
ejpam-6087	35	21	,	,	PUNCT
ejpam-6087	35	22	support	support	NOUN
ejpam-6087	35	23	vector	vector	NOUN
ejpam-6087	35	24	machine	machine	NOUN
ejpam-6087	35	25	(	(	PUNCT
ejpam-6087	35	26	svm	svm	PROPN
ejpam-6087	35	27	)	)	PUNCT
ejpam-6087	35	28	,	,	PUNCT
ejpam-6087	35	29	k	k	X
ejpam-6087	35	30	-	-	PUNCT
ejpam-6087	35	31	nearest	near	ADJ
ejpam-6087	35	32	neighbors	neighbor	NOUN
ejpam-6087	35	33	(	(	PUNCT
ejpam-6087	35	34	knn	knn	PROPN
ejpam-6087	35	35	)	)	PUNCT
ejpam-6087	35	36	,	,	PUNCT
ejpam-6087	35	37	random	random	ADJ
ejpam-6087	35	38	forest	forest	NOUN
ejpam-6087	35	39	(	(	PUNCT
ejpam-6087	35	40	rf	rf	NOUN
ejpam-6087	35	41	)	)	PUNCT
ejpam-6087	35	42	,	,	PUNCT
ejpam-6087	35	43	and	and	CCONJ
ejpam-6087	35	44	multilayer	multilayer	ADJ
ejpam-6087	35	45	perceptron	perceptron	PROPN
ejpam-6087	35	46	(	(	PUNCT
ejpam-6087	35	47	mlp	mlp	PROPN
ejpam-6087	35	48	)	)	PUNCT
ejpam-6087	35	49	,	,	PUNCT
ejpam-6087	35	50	along	along	ADP
ejpam-6087	35	51	with	with	ADP
ejpam-6087	35	52	ensemble	ensemble	ADJ
ejpam-6087	35	53	approaches	approach	NOUN
ejpam-6087	35	54	like	like	ADP
ejpam-6087	35	55	bagging	bagging	NOUN
ejpam-6087	35	56	,	,	PUNCT
ejpam-6087	35	57	boosting	boost	VERB
ejpam-6087	35	58	,	,	PUNCT
ejpam-6087	35	59	and	and	CCONJ
ejpam-6087	35	60	stacking	stacking	NOUN
ejpam-6087	35	61	.	.	PUNCT
ejpam-6087	36	1	an	an	DET
ejpam-6087	36	2	emphasis	emphasis	NOUN
ejpam-6087	36	3	is	be	AUX
ejpam-6087	36	4	placed	place	VERB
ejpam-6087	36	5	on	on	ADP
ejpam-6087	36	6	the	the	DET
ejpam-6087	36	7	influence	influence	NOUN
ejpam-6087	36	8	of	of	ADP
ejpam-6087	36	9	various	various	ADJ
ejpam-6087	36	10	feature	feature	NOUN
ejpam-6087	36	11	selection	selection	NOUN
ejpam-6087	36	12	methods	method	NOUN
ejpam-6087	36	13	on	on	ADP
ejpam-6087	36	14	model	model	NOUN
ejpam-6087	36	15	effectiveness	effectiveness	NOUN
ejpam-6087	36	16	.	.	PUNCT
ejpam-6087	37	1	a	a	DET
ejpam-6087	37	2	robust	robust	ADJ
ejpam-6087	37	3	evaluation	evaluation	NOUN
ejpam-6087	37	4	framework	framework	NOUN
ejpam-6087	37	5	that	that	PRON
ejpam-6087	37	6	employs	employ	VERB
ejpam-6087	37	7	metrics	metric	NOUN
ejpam-6087	37	8	such	such	ADJ
ejpam-6087	37	9	as	as	ADP
ejpam-6087	37	10	accuracy	accuracy	NOUN
ejpam-6087	37	11	,	,	PUNCT
ejpam-6087	37	12	precision	precision	NOUN
ejpam-6087	37	13	,	,	PUNCT
ejpam-6087	37	14	recall	recall	NOUN
ejpam-6087	37	15	,	,	PUNCT
ejpam-6087	37	16	f1	f1	NOUN
ejpam-6087	37	17	score	score	NOUN
ejpam-6087	37	18	,	,	PUNCT
ejpam-6087	37	19	and	and	CCONJ
ejpam-6087	37	20	m.	m.	PROPN
ejpam-6087	37	21	al	al	PROPN
ejpam-6087	37	22	-	-	PUNCT
ejpam-6087	37	23	tarawneh	tarawneh	PROPN
ejpam-6087	37	24	et	et	PROPN
ejpam-6087	37	25	al	al	PROPN
ejpam-6087	37	26	.	.	PUNCT
ejpam-6087	37	27	/	/	SYM
ejpam-6087	37	28	eur	eur	PROPN
ejpam-6087	37	29	.	.	PUNCT
ejpam-6087	38	1	j.	j.	PROPN
ejpam-6087	38	2	pure	pure	PROPN
ejpam-6087	38	3	appl	appl	PROPN
ejpam-6087	38	4	.	.	PROPN
ejpam-6087	38	5	math	math	PROPN
ejpam-6087	38	6	,	,	PUNCT
ejpam-6087	38	7	18	18	NUM
ejpam-6087	38	8	(	(	PUNCT
ejpam-6087	38	9	2	2	NUM
ejpam-6087	38	10	)	)	PUNCT
ejpam-6087	38	11	(	(	PUNCT
ejpam-6087	38	12	2025	2025	NUM
ejpam-6087	38	13	)	)	PUNCT
ejpam-6087	38	14	,	,	PUNCT
ejpam-6087	38	15	6087	6087	NUM
ejpam-6087	38	16	3	3	NUM
ejpam-6087	38	17	of	of	ADP
ejpam-6087	38	18	39	39	NUM
ejpam-6087	38	19	auc	auc	NOUN
ejpam-6087	38	20	-	-	PUNCT
ejpam-6087	38	21	roc	roc	NOUN
ejpam-6087	38	22	ensures	ensure	VERB
ejpam-6087	38	23	a	a	DET
ejpam-6087	38	24	thorough	thorough	ADJ
ejpam-6087	38	25	analysis	analysis	NOUN
ejpam-6087	38	26	of	of	ADP
ejpam-6087	38	27	the	the	DET
ejpam-6087	38	28	model	model	NOUN
ejpam-6087	38	29	capabilities	capability	NOUN
ejpam-6087	38	30	.	.	PUNCT
ejpam-6087	39	1	a	a	DET
ejpam-6087	39	2	key	key	ADJ
ejpam-6087	39	3	asset	asset	NOUN
ejpam-6087	39	4	in	in	ADP
ejpam-6087	39	5	this	this	DET
ejpam-6087	39	6	research	research	NOUN
ejpam-6087	39	7	is	be	AUX
ejpam-6087	39	8	the	the	DET
ejpam-6087	39	9	truthseeker	truthseeker	NOUN
ejpam-6087	39	10	dataset	dataset	NOUN
ejpam-6087	39	11	,	,	PUNCT
ejpam-6087	39	12	which	which	PRON
ejpam-6087	39	13	encompasses	encompass	VERB
ejpam-6087	39	14	a	a	DET
ejpam-6087	39	15	decade	decade	NOUN
ejpam-6087	39	16	worth	worth	NOUN
ejpam-6087	39	17	of	of	ADP
ejpam-6087	39	18	news	news	NOUN
ejpam-6087	39	19	articles	article	NOUN
ejpam-6087	39	20	,	,	PUNCT
ejpam-6087	39	21	tweets	tweet	NOUN
ejpam-6087	39	22	,	,	PUNCT
ejpam-6087	39	23	and	and	CCONJ
ejpam-6087	39	24	social	social	ADJ
ejpam-6087	39	25	media	medium	NOUN
ejpam-6087	39	26	content	content	NOUN
ejpam-6087	39	27	.	.	PUNCT
ejpam-6087	40	1	this	this	DET
ejpam-6087	40	2	diverse	diverse	ADJ
ejpam-6087	40	3	dataset	dataset	NOUN
ejpam-6087	40	4	supports	support	VERB
ejpam-6087	40	5	a	a	DET
ejpam-6087	40	6	detailed	detailed	ADJ
ejpam-6087	40	7	analysis	analysis	NOUN
ejpam-6087	40	8	by	by	ADP
ejpam-6087	40	9	providing	provide	VERB
ejpam-6087	40	10	textual	textual	ADJ
ejpam-6087	40	11	,	,	PUNCT
ejpam-6087	40	12	lexical	lexical	ADJ
ejpam-6087	40	13	,	,	PUNCT
ejpam-6087	40	14	and	and	CCONJ
ejpam-6087	40	15	metadata	metadata	NOUN
ejpam-6087	40	16	features	feature	NOUN
ejpam-6087	40	17	.	.	PUNCT
ejpam-6087	41	1	inclusion	inclusion	NOUN
ejpam-6087	41	2	of	of	ADP
ejpam-6087	41	3	metadata	metadata	NOUN
ejpam-6087	41	4	,	,	PUNCT
ejpam-6087	41	5	such	such	ADJ
ejpam-6087	41	6	as	as	ADP
ejpam-6087	41	7	user	user	NOUN
ejpam-6087	41	8	engagement	engagement	NOUN
ejpam-6087	41	9	statistics	statistic	NOUN
ejpam-6087	41	10	,	,	PUNCT
ejpam-6087	41	11	allows	allow	VERB
ejpam-6087	41	12	the	the	DET
ejpam-6087	41	13	detection	detection	NOUN
ejpam-6087	41	14	of	of	ADP
ejpam-6087	41	15	subtle	subtle	ADJ
ejpam-6087	41	16	patterns	pattern	NOUN
ejpam-6087	41	17	that	that	PRON
ejpam-6087	41	18	textual	textual	ADJ
ejpam-6087	41	19	analysis	analysis	NOUN
ejpam-6087	41	20	alone	alone	ADV
ejpam-6087	41	21	may	may	AUX
ejpam-6087	41	22	miss	miss	VERB
ejpam-6087	41	23	.	.	PUNCT
ejpam-6087	42	1	the	the	DET
ejpam-6087	42	2	temporal	temporal	ADJ
ejpam-6087	42	3	and	and	CCONJ
ejpam-6087	42	4	content	content	NOUN
ejpam-6087	42	5	diversity	diversity	NOUN
ejpam-6087	42	6	of	of	ADP
ejpam-6087	42	7	the	the	DET
ejpam-6087	42	8	dataset	dataset	NOUN
ejpam-6087	42	9	ensures	ensure	VERB
ejpam-6087	42	10	that	that	SCONJ
ejpam-6087	42	11	the	the	DET
ejpam-6087	42	12	models	model	NOUN
ejpam-6087	42	13	are	be	AUX
ejpam-6087	42	14	evaluated	evaluate	VERB
ejpam-6087	42	15	across	across	ADP
ejpam-6087	42	16	a	a	DET
ejpam-6087	42	17	wide	wide	ADJ
ejpam-6087	42	18	range	range	NOUN
ejpam-6087	42	19	of	of	ADP
ejpam-6087	42	20	real	real	ADJ
ejpam-6087	42	21	-	-	PUNCT
ejpam-6087	42	22	world	world	NOUN
ejpam-6087	42	23	scenarios	scenario	NOUN
ejpam-6087	42	24	.	.	PUNCT
ejpam-6087	43	1	by	by	ADP
ejpam-6087	43	2	systematically	systematically	ADV
ejpam-6087	43	3	comparing	compare	VERB
ejpam-6087	43	4	individual	individual	ADJ
ejpam-6087	43	5	classifiers	classifier	NOUN
ejpam-6087	43	6	and	and	CCONJ
ejpam-6087	43	7	ensemble	ensemble	ADJ
ejpam-6087	43	8	methods	method	NOUN
ejpam-6087	43	9	alongside	alongside	ADP
ejpam-6087	43	10	various	various	ADJ
ejpam-6087	43	11	feature	feature	NOUN
ejpam-6087	43	12	selection	selection	NOUN
ejpam-6087	43	13	techniques	technique	NOUN
ejpam-6087	43	14	,	,	PUNCT
ejpam-6087	43	15	this	this	DET
ejpam-6087	43	16	study	study	NOUN
ejpam-6087	43	17	sheds	shed	VERB
ejpam-6087	43	18	light	light	NOUN
ejpam-6087	43	19	on	on	ADP
ejpam-6087	43	20	the	the	DET
ejpam-6087	43	21	optimal	optimal	ADJ
ejpam-6087	43	22	strategies	strategy	NOUN
ejpam-6087	43	23	for	for	ADP
ejpam-6087	43	24	effective	effective	ADJ
ejpam-6087	43	25	fake	fake	ADJ
ejpam-6087	43	26	news	news	NOUN
ejpam-6087	43	27	detection	detection	NOUN
ejpam-6087	43	28	.	.	PUNCT
ejpam-6087	44	1	the	the	DET
ejpam-6087	44	2	contributions	contribution	NOUN
ejpam-6087	44	3	of	of	ADP
ejpam-6087	44	4	this	this	DET
ejpam-6087	44	5	work	work	NOUN
ejpam-6087	44	6	include	include	VERB
ejpam-6087	44	7	the	the	DET
ejpam-6087	44	8	following	following	NOUN
ejpam-6087	44	9	.	.	PUNCT
ejpam-6087	45	1	•	•	NUM
ejpam-6087	45	2	comprehensive	comprehensive	ADJ
ejpam-6087	45	3	evaluation	evaluation	NOUN
ejpam-6087	45	4	:	:	PUNCT
ejpam-6087	45	5	a	a	DET
ejpam-6087	45	6	detailed	detailed	ADJ
ejpam-6087	45	7	assessment	assessment	NOUN
ejpam-6087	45	8	of	of	ADP
ejpam-6087	45	9	multiple	multiple	ADJ
ejpam-6087	45	10	machine	machine	NOUN
ejpam-6087	45	11	learning	learning	NOUN
ejpam-6087	45	12	models	model	NOUN
ejpam-6087	45	13	for	for	ADP
ejpam-6087	45	14	the	the	DET
ejpam-6087	45	15	detection	detection	NOUN
ejpam-6087	45	16	of	of	ADP
ejpam-6087	45	17	fake	fake	ADJ
ejpam-6087	45	18	news	news	NOUN
ejpam-6087	45	19	using	use	VERB
ejpam-6087	45	20	various	various	ADJ
ejpam-6087	45	21	performance	performance	NOUN
ejpam-6087	45	22	metrics	metric	NOUN
ejpam-6087	45	23	such	such	ADJ
ejpam-6087	45	24	as	as	ADP
ejpam-6087	45	25	accuracy	accuracy	NOUN
ejpam-6087	45	26	,	,	PUNCT
ejpam-6087	45	27	precision	precision	NOUN
ejpam-6087	45	28	,	,	PUNCT
ejpam-6087	45	29	recall	recall	NOUN
ejpam-6087	45	30	,	,	PUNCT
ejpam-6087	45	31	f1	f1	NOUN
ejpam-6087	45	32	score	score	NOUN
ejpam-6087	45	33	and	and	CCONJ
ejpam-6087	45	34	auc	auc	NOUN
ejpam-6087	45	35	-	-	PUNCT
ejpam-6087	45	36	roc	roc	NOUN
ejpam-6087	45	37	.	.	PROPN
ejpam-6087	45	38	•	•	ADP
ejpam-6087	45	39	impactful	impactful	ADJ
ejpam-6087	45	40	feature	feature	NOUN
ejpam-6087	45	41	selection	selection	NOUN
ejpam-6087	45	42	:	:	PUNCT
ejpam-6087	45	43	analysis	analysis	NOUN
ejpam-6087	45	44	of	of	ADP
ejpam-6087	45	45	the	the	DET
ejpam-6087	45	46	effects	effect	NOUN
ejpam-6087	45	47	of	of	ADP
ejpam-6087	45	48	feature	feature	NOUN
ejpam-6087	45	49	selection	selection	NOUN
ejpam-6087	45	50	methods	method	NOUN
ejpam-6087	45	51	,	,	PUNCT
ejpam-6087	45	52	including	include	VERB
ejpam-6087	45	53	rfe	rfe	NOUN
ejpam-6087	45	54	,	,	PUNCT
ejpam-6087	45	55	selectkbest	selectkb	ADJ
ejpam-6087	45	56	,	,	PUNCT
ejpam-6087	45	57	pca	pca	PROPN
ejpam-6087	45	58	,	,	PUNCT
ejpam-6087	45	59	and	and	CCONJ
ejpam-6087	45	60	ga	ga	PROPN
ejpam-6087	45	61	,	,	PUNCT
ejpam-6087	45	62	on	on	ADP
ejpam-6087	45	63	model	model	NOUN
ejpam-6087	45	64	performance	performance	NOUN
ejpam-6087	45	65	.	.	PUNCT
ejpam-6087	46	1	•	•	NUM
ejpam-6087	46	2	rich	rich	ADJ
ejpam-6087	46	3	dataset	dataset	NOUN
ejpam-6087	46	4	utilization	utilization	NOUN
ejpam-6087	46	5	:	:	PUNCT
ejpam-6087	46	6	utilization	utilization	NOUN
ejpam-6087	46	7	of	of	ADP
ejpam-6087	46	8	the	the	DET
ejpam-6087	46	9	truthseeker	truthseeker	NOUN
ejpam-6087	46	10	dataset	dataset	NOUN
ejpam-6087	46	11	,	,	PUNCT
ejpam-6087	46	12	offering	offer	VERB
ejpam-6087	46	13	diverse	diverse	ADJ
ejpam-6087	46	14	and	and	CCONJ
ejpam-6087	46	15	temporally	temporally	ADV
ejpam-6087	46	16	extensive	extensive	ADJ
ejpam-6087	46	17	data	datum	NOUN
ejpam-6087	46	18	to	to	PART
ejpam-6087	46	19	ensure	ensure	VERB
ejpam-6087	46	20	robust	robust	ADJ
ejpam-6087	46	21	model	model	NOUN
ejpam-6087	46	22	evaluation	evaluation	NOUN
ejpam-6087	46	23	.	.	PUNCT
ejpam-6087	47	1	•	•	NUM
ejpam-6087	47	2	holistic	holistic	ADJ
ejpam-6087	47	3	analysis	analysis	NOUN
ejpam-6087	47	4	:	:	PUNCT
ejpam-6087	47	5	examination	examination	NOUN
ejpam-6087	47	6	of	of	ADP
ejpam-6087	47	7	textual	textual	ADJ
ejpam-6087	47	8	,	,	PUNCT
ejpam-6087	47	9	lexical	lexical	ADJ
ejpam-6087	47	10	,	,	PUNCT
ejpam-6087	47	11	and	and	CCONJ
ejpam-6087	47	12	metadata	metadata	VERB
ejpam-6087	47	13	features	feature	NOUN
ejpam-6087	47	14	to	to	PART
ejpam-6087	47	15	capture	capture	VERB
ejpam-6087	47	16	nuanced	nuanced	ADJ
ejpam-6087	47	17	patterns	pattern	NOUN
ejpam-6087	47	18	in	in	ADP
ejpam-6087	47	19	the	the	DET
ejpam-6087	47	20	detection	detection	NOUN
ejpam-6087	47	21	of	of	ADP
ejpam-6087	47	22	fake	fake	ADJ
ejpam-6087	47	23	news	news	NOUN
ejpam-6087	47	24	.	.	PUNCT
ejpam-6087	48	1	•	•	NUM
ejpam-6087	48	2	comparative	comparative	ADJ
ejpam-6087	48	3	analysis	analysis	NOUN
ejpam-6087	48	4	:	:	PUNCT
ejpam-6087	48	5	insights	insight	NOUN
ejpam-6087	48	6	into	into	ADP
ejpam-6087	48	7	the	the	DET
ejpam-6087	48	8	performance	performance	NOUN
ejpam-6087	48	9	of	of	ADP
ejpam-6087	48	10	individual	individual	ADJ
ejpam-6087	48	11	classifiers	classifier	NOUN
ejpam-6087	48	12	versus	versus	ADP
ejpam-6087	48	13	ensemble	ensemble	ADJ
ejpam-6087	48	14	methods	method	NOUN
ejpam-6087	48	15	to	to	PART
ejpam-6087	48	16	identify	identify	VERB
ejpam-6087	48	17	optimal	optimal	ADJ
ejpam-6087	48	18	detection	detection	NOUN
ejpam-6087	48	19	strategies	strategy	NOUN
ejpam-6087	48	20	.	.	PUNCT
ejpam-6087	49	1	the	the	DET
ejpam-6087	49	2	remainder	remainder	NOUN
ejpam-6087	49	3	of	of	ADP
ejpam-6087	49	4	this	this	DET
ejpam-6087	49	5	paper	paper	NOUN
ejpam-6087	49	6	is	be	AUX
ejpam-6087	49	7	structured	structure	VERB
ejpam-6087	49	8	as	as	SCONJ
ejpam-6087	49	9	follows	follow	VERB
ejpam-6087	49	10	.	.	PUNCT
ejpam-6087	50	1	section	section	NOUN
ejpam-6087	50	2	2	2	NUM
ejpam-6087	50	3	reviews	review	NOUN
ejpam-6087	50	4	related	relate	VERB
ejpam-6087	50	5	work	work	NOUN
ejpam-6087	50	6	,	,	PUNCT
ejpam-6087	50	7	section	section	NOUN
ejpam-6087	50	8	3	3	NUM
ejpam-6087	50	9	details	detail	NOUN
ejpam-6087	50	10	the	the	DET
ejpam-6087	50	11	methodology	methodology	NOUN
ejpam-6087	50	12	,	,	PUNCT
ejpam-6087	50	13	including	include	VERB
ejpam-6087	50	14	data	data	NOUN
ejpam-6087	50	15	collection	collection	NOUN
ejpam-6087	50	16	,	,	PUNCT
ejpam-6087	50	17	machine	machine	NOUN
ejpam-6087	50	18	learning	learn	VERB
ejpam-6087	50	19	techniques	technique	NOUN
ejpam-6087	50	20	,	,	PUNCT
ejpam-6087	50	21	feature	feature	NOUN
ejpam-6087	50	22	selection	selection	NOUN
ejpam-6087	50	23	,	,	PUNCT
ejpam-6087	50	24	and	and	CCONJ
ejpam-6087	50	25	evaluation	evaluation	NOUN
ejpam-6087	50	26	.	.	PUNCT
ejpam-6087	51	1	section	section	NOUN
ejpam-6087	51	2	4	4	NUM
ejpam-6087	51	3	discusses	discuss	VERB
ejpam-6087	51	4	the	the	DET
ejpam-6087	51	5	results	result	NOUN
ejpam-6087	51	6	and	and	CCONJ
ejpam-6087	51	7	findings	finding	NOUN
ejpam-6087	51	8	,	,	PUNCT
ejpam-6087	51	9	and	and	CCONJ
ejpam-6087	51	10	section	section	NOUN
ejpam-6087	51	11	5	5	NUM
ejpam-6087	51	12	concludes	conclude	VERB
ejpam-6087	51	13	the	the	DET
ejpam-6087	51	14	study	study	NOUN
ejpam-6087	51	15	with	with	ADP
ejpam-6087	51	16	key	key	ADJ
ejpam-6087	51	17	findings	finding	NOUN
ejpam-6087	51	18	.	.	PUNCT
ejpam-6087	52	1	2	2	X
ejpam-6087	52	2	.	.	X
ejpam-6087	52	3	literature	literature	NOUN
ejpam-6087	52	4	review	review	VERB
ejpam-6087	52	5	fake	fake	ADJ
ejpam-6087	52	6	news	news	NOUN
ejpam-6087	52	7	detection	detection	NOUN
ejpam-6087	52	8	has	have	AUX
ejpam-6087	52	9	emerged	emerge	VERB
ejpam-6087	52	10	as	as	ADP
ejpam-6087	52	11	a	a	DET
ejpam-6087	52	12	critical	critical	ADJ
ejpam-6087	52	13	research	research	NOUN
ejpam-6087	52	14	area	area	NOUN
ejpam-6087	52	15	due	due	ADP
ejpam-6087	52	16	to	to	ADP
ejpam-6087	52	17	the	the	DET
ejpam-6087	52	18	pervasive	pervasive	ADJ
ejpam-6087	52	19	spread	spread	NOUN
ejpam-6087	52	20	of	of	ADP
ejpam-6087	52	21	misinformation	misinformation	NOUN
ejpam-6087	52	22	across	across	ADP
ejpam-6087	52	23	digital	digital	ADJ
ejpam-6087	52	24	platforms	platform	NOUN
ejpam-6087	52	25	.	.	PUNCT
ejpam-6087	53	1	in	in	ADP
ejpam-6087	53	2	recent	recent	ADJ
ejpam-6087	53	3	years	year	NOUN
ejpam-6087	53	4	,	,	PUNCT
ejpam-6087	53	5	several	several	ADJ
ejpam-6087	53	6	studies	study	NOUN
ejpam-6087	53	7	have	have	AUX
ejpam-6087	53	8	explored	explore	VERB
ejpam-6087	53	9	ml	ml	ADP
ejpam-6087	53	10	techniques	technique	NOUN
ejpam-6087	53	11	to	to	PART
ejpam-6087	53	12	tackle	tackle	VERB
ejpam-6087	53	13	this	this	DET
ejpam-6087	53	14	pressing	press	VERB
ejpam-6087	53	15	issue	issue	NOUN
ejpam-6087	53	16	.	.	PUNCT
ejpam-6087	54	1	this	this	DET
ejpam-6087	54	2	review	review	NOUN
ejpam-6087	54	3	of	of	ADP
ejpam-6087	54	4	the	the	DET
ejpam-6087	54	5	literature	literature	NOUN
ejpam-6087	54	6	discusses	discuss	VERB
ejpam-6087	54	7	notable	notable	ADJ
ejpam-6087	54	8	work	work	NOUN
ejpam-6087	54	9	in	in	ADP
ejpam-6087	54	10	the	the	DET
ejpam-6087	54	11	field	field	NOUN
ejpam-6087	54	12	,	,	PUNCT
ejpam-6087	54	13	highlighting	highlight	VERB
ejpam-6087	54	14	their	their	PRON
ejpam-6087	54	15	methodologies	methodology	NOUN
ejpam-6087	54	16	and	and	CCONJ
ejpam-6087	54	17	comparing	compare	VERB
ejpam-6087	54	18	them	they	PRON
ejpam-6087	54	19	with	with	ADP
ejpam-6087	54	20	current	current	ADJ
ejpam-6087	54	21	research	research	NOUN
ejpam-6087	54	22	.	.	PUNCT
ejpam-6087	55	1	the	the	DET
ejpam-6087	55	2	study	study	NOUN
ejpam-6087	55	3	in	in	ADP
ejpam-6087	55	4	[	[	X
ejpam-6087	55	5	11	11	NUM
ejpam-6087	55	6	]	]	PUNCT
ejpam-6087	55	7	presented	present	VERB
ejpam-6087	55	8	bil	bil	PROPN
ejpam-6087	55	9	-	-	PROPN
ejpam-6087	55	10	fand	fand	PROPN
ejpam-6087	55	11	,	,	PUNCT
ejpam-6087	55	12	an	an	DET
ejpam-6087	55	13	ensemble	ensemble	ADJ
ejpam-6087	55	14	-	-	PUNCT
ejpam-6087	55	15	based	base	VERB
ejpam-6087	55	16	system	system	NOUN
ejpam-6087	55	17	for	for	ADP
ejpam-6087	55	18	detecting	detect	VERB
ejpam-6087	55	19	fake	fake	ADJ
ejpam-6087	55	20	news	news	NOUN
ejpam-6087	55	21	in	in	ADP
ejpam-6087	55	22	english	english	PROPN
ejpam-6087	55	23	and	and	CCONJ
ejpam-6087	55	24	urdu	urdu	PROPN
ejpam-6087	55	25	.	.	PUNCT
ejpam-6087	56	1	combining	combine	VERB
ejpam-6087	56	2	multilingual	multilingual	ADJ
ejpam-6087	56	3	bert	bert	NOUN
ejpam-6087	56	4	,	,	PUNCT
ejpam-6087	56	5	long	long	ADJ
ejpam-6087	56	6	-	-	PUNCT
ejpam-6087	56	7	short	short	ADJ
ejpam-6087	56	8	-	-	PUNCT
ejpam-6087	56	9	term	term	NOUN
ejpam-6087	56	10	memory	memory	NOUN
ejpam-6087	56	11	(	(	PUNCT
ejpam-6087	56	12	lstm	lstm	NOUN
ejpam-6087	56	13	)	)	PUNCT
ejpam-6087	56	14	models	model	NOUN
ejpam-6087	56	15	,	,	PUNCT
ejpam-6087	56	16	and	and	CCONJ
ejpam-6087	56	17	a	a	DET
ejpam-6087	56	18	resnet-101	resnet-101	NOUN
ejpam-6087	56	19	with	with	ADP
ejpam-6087	56	20	gru	gru	NOUN
ejpam-6087	56	21	for	for	ADP
ejpam-6087	56	22	multimedia	multimedia	NOUN
ejpam-6087	56	23	analysis	analysis	NOUN
ejpam-6087	56	24	,	,	PUNCT
ejpam-6087	56	25	the	the	DET
ejpam-6087	56	26	system	system	NOUN
ejpam-6087	56	27	achieved	achieve	VERB
ejpam-6087	56	28	an	an	DET
ejpam-6087	56	29	overall	overall	ADJ
ejpam-6087	56	30	notable	notable	ADJ
ejpam-6087	56	31	accuracy	accuracy	NOUN
ejpam-6087	56	32	of	of	ADP
ejpam-6087	56	33	92.07	92.07	NUM
ejpam-6087	56	34	%	%	NOUN
ejpam-6087	56	35	.	.	PUNCT
ejpam-6087	57	1	the	the	DET
ejpam-6087	57	2	approach	approach	NOUN
ejpam-6087	57	3	demonstrated	demonstrate	VERB
ejpam-6087	57	4	strong	strong	ADJ
ejpam-6087	57	5	performance	performance	NOUN
ejpam-6087	57	6	in	in	ADP
ejpam-6087	57	7	linguistic	linguistic	ADJ
ejpam-6087	57	8	analysis	analysis	NOUN
ejpam-6087	57	9	,	,	PUNCT
ejpam-6087	57	10	pattern	pattern	NOUN
ejpam-6087	57	11	recognition	recognition	NOUN
ejpam-6087	57	12	,	,	PUNCT
ejpam-6087	57	13	and	and	CCONJ
ejpam-6087	57	14	multimedia	multimedia	NOUN
ejpam-6087	57	15	content	content	NOUN
ejpam-6087	57	16	evaluation	evaluation	NOUN
ejpam-6087	57	17	.	.	PUNCT
ejpam-6087	58	1	however	however	ADV
ejpam-6087	58	2	,	,	PUNCT
ejpam-6087	58	3	the	the	DET
ejpam-6087	58	4	study	study	NOUN
ejpam-6087	58	5	does	do	AUX
ejpam-6087	58	6	not	not	PART
ejpam-6087	58	7	explore	explore	VERB
ejpam-6087	58	8	feature	feature	NOUN
ejpam-6087	58	9	selection	selection	NOUN
ejpam-6087	58	10	methods	method	NOUN
ejpam-6087	58	11	.	.	PUNCT
ejpam-6087	59	1	m.	m.	NOUN
ejpam-6087	59	2	al	al	PROPN
ejpam-6087	59	3	-	-	PUNCT
ejpam-6087	59	4	tarawneh	tarawneh	PROPN
ejpam-6087	59	5	et	et	PROPN
ejpam-6087	59	6	al	al	PROPN
ejpam-6087	59	7	.	.	PUNCT
ejpam-6087	59	8	/	/	SYM
ejpam-6087	59	9	eur	eur	PROPN
ejpam-6087	59	10	.	.	PUNCT
ejpam-6087	60	1	j.	j.	PROPN
ejpam-6087	60	2	pure	pure	PROPN
ejpam-6087	60	3	appl	appl	PROPN
ejpam-6087	60	4	.	.	PROPN
ejpam-6087	60	5	math	math	PROPN
ejpam-6087	60	6	,	,	PUNCT
ejpam-6087	60	7	18	18	NUM
ejpam-6087	60	8	(	(	PUNCT
ejpam-6087	60	9	2	2	NUM
ejpam-6087	60	10	)	)	PUNCT
ejpam-6087	60	11	(	(	PUNCT
ejpam-6087	60	12	2025	2025	NUM
ejpam-6087	60	13	)	)	PUNCT
ejpam-6087	60	14	,	,	PUNCT
ejpam-6087	60	15	6087	6087	NUM
ejpam-6087	60	16	4	4	NUM
ejpam-6087	60	17	of	of	ADP
ejpam-6087	60	18	39	39	NUM
ejpam-6087	60	19	the	the	DET
ejpam-6087	60	20	authors	author	NOUN
ejpam-6087	60	21	of	of	ADP
ejpam-6087	60	22	[	[	X
ejpam-6087	60	23	12	12	NUM
ejpam-6087	60	24	]	]	PUNCT
ejpam-6087	60	25	developed	develop	VERB
ejpam-6087	60	26	a	a	DET
ejpam-6087	60	27	robust	robust	ADJ
ejpam-6087	60	28	fake	fake	ADJ
ejpam-6087	60	29	news	news	NOUN
ejpam-6087	60	30	detection	detection	NOUN
ejpam-6087	60	31	approach	approach	NOUN
ejpam-6087	60	32	using	use	VERB
ejpam-6087	60	33	welfake	welfake	NOUN
ejpam-6087	60	34	,	,	PUNCT
ejpam-6087	60	35	fakenewsnet	fakenewsnet	NOUN
ejpam-6087	60	36	,	,	PUNCT
ejpam-6087	60	37	and	and	CCONJ
ejpam-6087	60	38	fakenewsprediction	fakenewsprediction	NOUN
ejpam-6087	60	39	datasets	dataset	NOUN
ejpam-6087	60	40	.	.	PUNCT
ejpam-6087	61	1	fasttext	fasttext	PROPN
ejpam-6087	61	2	embeddings	embedding	NOUN
ejpam-6087	61	3	were	be	AUX
ejpam-6087	61	4	combined	combine	VERB
ejpam-6087	61	5	with	with	ADP
ejpam-6087	61	6	ml	ml	ADV
ejpam-6087	61	7	and	and	CCONJ
ejpam-6087	61	8	deep	deep	ADJ
ejpam-6087	61	9	learning	learning	NOUN
ejpam-6087	61	10	(	(	PUNCT
ejpam-6087	61	11	dl	dl	NOUN
ejpam-6087	61	12	)	)	PUNCT
ejpam-6087	61	13	methods	method	NOUN
ejpam-6087	61	14	,	,	PUNCT
ejpam-6087	61	15	including	include	VERB
ejpam-6087	61	16	a	a	DET
ejpam-6087	61	17	hybrid	hybrid	ADJ
ejpam-6087	61	18	cnn	cnn	PROPN
ejpam-6087	61	19	-	-	PUNCT
ejpam-6087	61	20	lstm	lstm	ADJ
ejpam-6087	61	21	model	model	NOUN
ejpam-6087	61	22	that	that	PRON
ejpam-6087	61	23	achieved	achieve	VERB
ejpam-6087	61	24	high	high	ADJ
ejpam-6087	61	25	performance	performance	NOUN
ejpam-6087	61	26	(	(	PUNCT
ejpam-6087	61	27	accuracy	accuracy	NOUN
ejpam-6087	61	28	and	and	CCONJ
ejpam-6087	61	29	f1	f1	NOUN
ejpam-6087	61	30	scores	score	NOUN
ejpam-6087	61	31	of	of	ADP
ejpam-6087	61	32	99	99	NUM
ejpam-6087	61	33	%	%	NOUN
ejpam-6087	61	34	,	,	PUNCT
ejpam-6087	61	35	and	and	CCONJ
ejpam-6087	61	36	97	97	NUM
ejpam-6087	61	37	%	%	NOUN
ejpam-6087	61	38	)	)	PUNCT
ejpam-6087	61	39	.	.	PUNCT
ejpam-6087	62	1	transformer	transformer	NOUN
ejpam-6087	62	2	-	-	PUNCT
ejpam-6087	62	3	based	base	VERB
ejpam-6087	62	4	models	model	NOUN
ejpam-6087	62	5	(	(	PUNCT
ejpam-6087	62	6	bert	bert	PROPN
ejpam-6087	62	7	,	,	PUNCT
ejpam-6087	62	8	xlnet	xlnet	PROPN
ejpam-6087	62	9	,	,	PUNCT
ejpam-6087	62	10	roberta	roberta	PROPN
ejpam-6087	62	11	)	)	PUNCT
ejpam-6087	62	12	were	be	AUX
ejpam-6087	62	13	also	also	ADV
ejpam-6087	62	14	optimized	optimize	VERB
ejpam-6087	62	15	for	for	ADP
ejpam-6087	62	16	superior	superior	ADJ
ejpam-6087	62	17	syntactic	syntactic	ADJ
ejpam-6087	62	18	management	management	NOUN
ejpam-6087	62	19	.	.	PUNCT
ejpam-6087	63	1	explainable	explainable	ADJ
ejpam-6087	63	2	ai	ai	PROPN
ejpam-6087	63	3	techniques	technique	NOUN
ejpam-6087	63	4	(	(	PUNCT
ejpam-6087	63	5	lime	lime	NOUN
ejpam-6087	63	6	,	,	PUNCT
ejpam-6087	63	7	lda	lda	PROPN
ejpam-6087	63	8	)	)	PUNCT
ejpam-6087	63	9	provided	provide	VERB
ejpam-6087	63	10	insight	insight	NOUN
ejpam-6087	63	11	into	into	ADP
ejpam-6087	63	12	model	model	NOUN
ejpam-6087	63	13	decisions	decision	NOUN
ejpam-6087	63	14	.	.	PUNCT
ejpam-6087	64	1	however	however	ADV
ejpam-6087	64	2	,	,	PUNCT
ejpam-6087	64	3	the	the	DET
ejpam-6087	64	4	study	study	NOUN
ejpam-6087	64	5	does	do	AUX
ejpam-6087	64	6	not	not	PART
ejpam-6087	64	7	explore	explore	VERB
ejpam-6087	64	8	feature	feature	NOUN
ejpam-6087	64	9	selection	selection	NOUN
ejpam-6087	64	10	methods	method	NOUN
ejpam-6087	64	11	.	.	PUNCT
ejpam-6087	65	1	the	the	DET
ejpam-6087	65	2	work	work	NOUN
ejpam-6087	65	3	in	in	ADP
ejpam-6087	65	4	[	[	X
ejpam-6087	65	5	13	13	NUM
ejpam-6087	65	6	]	]	PUNCT
ejpam-6087	65	7	addressed	address	VERB
ejpam-6087	65	8	the	the	DET
ejpam-6087	65	9	spread	spread	NOUN
ejpam-6087	65	10	of	of	ADP
ejpam-6087	65	11	covid-19	covid-19	PROPN
ejpam-6087	65	12	misinformation	misinformation	NOUN
ejpam-6087	65	13	by	by	ADP
ejpam-6087	65	14	developing	develop	VERB
ejpam-6087	65	15	a	a	DET
ejpam-6087	65	16	convolutional	convolutional	ADJ
ejpam-6087	65	17	neural	neural	ADJ
ejpam-6087	65	18	network	network	NOUN
ejpam-6087	65	19	(	(	PUNCT
ejpam-6087	65	20	cnn)-based	cnn)-base	VERB
ejpam-6087	65	21	deep	deep	ADJ
ejpam-6087	65	22	learning	learning	NOUN
ejpam-6087	65	23	model	model	NOUN
ejpam-6087	65	24	using	use	VERB
ejpam-6087	65	25	word	word	NOUN
ejpam-6087	65	26	embeddings	embedding	NOUN
ejpam-6087	65	27	.	.	PUNCT
ejpam-6087	66	1	the	the	DET
ejpam-6087	66	2	optimal	optimal	ADJ
ejpam-6087	66	3	cnn	cnn	PROPN
ejpam-6087	66	4	architecture	architecture	NOUN
ejpam-6087	66	5	was	be	AUX
ejpam-6087	66	6	determined	determine	VERB
ejpam-6087	66	7	through	through	ADP
ejpam-6087	66	8	grid	grid	NOUN
ejpam-6087	66	9	search	search	NOUN
ejpam-6087	66	10	.	.	PUNCT
ejpam-6087	67	1	the	the	DET
ejpam-6087	67	2	effectiveness	effectiveness	NOUN
ejpam-6087	67	3	of	of	ADP
ejpam-6087	67	4	the	the	DET
ejpam-6087	67	5	model	model	NOUN
ejpam-6087	67	6	was	be	AUX
ejpam-6087	67	7	evaluated	evaluate	VERB
ejpam-6087	67	8	against	against	ADP
ejpam-6087	67	9	various	various	ADJ
ejpam-6087	67	10	state	state	NOUN
ejpam-6087	67	11	-	-	PUNCT
ejpam-6087	67	12	of	of	ADP
ejpam-6087	67	13	-	-	PUNCT
ejpam-6087	67	14	the	the	DET
ejpam-6087	67	15	-	-	PUNCT
ejpam-6087	67	16	art	art	NOUN
ejpam-6087	67	17	ml	ml	NOUN
ejpam-6087	67	18	algorithms	algorithm	NOUN
ejpam-6087	67	19	,	,	PUNCT
ejpam-6087	67	20	with	with	ADP
ejpam-6087	67	21	cnn	cnn	PROPN
ejpam-6087	67	22	achieving	achieve	VERB
ejpam-6087	67	23	the	the	DET
ejpam-6087	67	24	highest	high	ADJ
ejpam-6087	67	25	performance	performance	NOUN
ejpam-6087	67	26	:	:	PUNCT
ejpam-6087	67	27	96.19	96.19	NUM
ejpam-6087	67	28	%	%	NOUN
ejpam-6087	67	29	mean	mean	NOUN
ejpam-6087	67	30	accuracy	accuracy	NOUN
ejpam-6087	67	31	,	,	PUNCT
ejpam-6087	67	32	95	95	NUM
ejpam-6087	67	33	%	%	NOUN
ejpam-6087	67	34	mean	mean	NOUN
ejpam-6087	67	35	f1	f1	NOUN
ejpam-6087	67	36	score	score	NOUN
ejpam-6087	67	37	,	,	PUNCT
ejpam-6087	67	38	and	and	CCONJ
ejpam-6087	67	39	0.985	0.985	NUM
ejpam-6087	67	40	auc	auc	NOUN
ejpam-6087	67	41	.	.	PUNCT
ejpam-6087	68	1	however	however	ADV
ejpam-6087	68	2	,	,	PUNCT
ejpam-6087	68	3	they	they	PRON
ejpam-6087	68	4	did	do	AUX
ejpam-6087	68	5	not	not	PART
ejpam-6087	68	6	explore	explore	VERB
ejpam-6087	68	7	feature	feature	NOUN
ejpam-6087	68	8	selection	selection	NOUN
ejpam-6087	68	9	methods	method	NOUN
ejpam-6087	68	10	.	.	PUNCT
ejpam-6087	69	1	the	the	DET
ejpam-6087	69	2	study	study	NOUN
ejpam-6087	69	3	in	in	ADP
ejpam-6087	69	4	[	[	X
ejpam-6087	69	5	14	14	NUM
ejpam-6087	69	6	]	]	PUNCT
ejpam-6087	69	7	examined	examine	VERB
ejpam-6087	69	8	the	the	DET
ejpam-6087	69	9	impact	impact	NOUN
ejpam-6087	69	10	of	of	ADP
ejpam-6087	69	11	fake	fake	ADJ
ejpam-6087	69	12	news	news	NOUN
ejpam-6087	69	13	on	on	ADP
ejpam-6087	69	14	society	society	NOUN
ejpam-6087	69	15	and	and	CCONJ
ejpam-6087	69	16	presented	present	VERB
ejpam-6087	69	17	a	a	DET
ejpam-6087	69	18	comparative	comparative	ADJ
ejpam-6087	69	19	analysis	analysis	NOUN
ejpam-6087	69	20	of	of	ADP
ejpam-6087	69	21	ml	ml	ADP
ejpam-6087	69	22	methods	method	NOUN
ejpam-6087	69	23	to	to	PART
ejpam-6087	69	24	detect	detect	VERB
ejpam-6087	69	25	fake	fake	ADJ
ejpam-6087	69	26	news	news	NOUN
ejpam-6087	69	27	.	.	PUNCT
ejpam-6087	70	1	the	the	DET
ejpam-6087	70	2	models	model	NOUN
ejpam-6087	70	3	evaluated	evaluate	VERB
ejpam-6087	70	4	included	include	VERB
ejpam-6087	70	5	nb	nb	PROPN
ejpam-6087	70	6	,	,	PUNCT
ejpam-6087	70	7	dt	dt	PROPN
ejpam-6087	70	8	classifier	classifier	NOUN
ejpam-6087	70	9	,	,	PUNCT
ejpam-6087	70	10	rf	rf	ADJ
ejpam-6087	70	11	,	,	PUNCT
ejpam-6087	70	12	and	and	CCONJ
ejpam-6087	70	13	logistic	logistic	ADJ
ejpam-6087	70	14	regression	regression	NOUN
ejpam-6087	70	15	(	(	PUNCT
ejpam-6087	70	16	lr	lr	NOUN
ejpam-6087	70	17	)	)	PUNCT
ejpam-6087	70	18	.	.	PUNCT
ejpam-6087	71	1	the	the	DET
ejpam-6087	71	2	results	result	NOUN
ejpam-6087	71	3	indicated	indicate	VERB
ejpam-6087	71	4	that	that	SCONJ
ejpam-6087	71	5	the	the	DET
ejpam-6087	71	6	dt	dt	PROPN
ejpam-6087	71	7	classifier	classifier	NOUN
ejpam-6087	71	8	achieved	achieve	VERB
ejpam-6087	71	9	the	the	DET
ejpam-6087	71	10	highest	high	ADJ
ejpam-6087	71	11	accuracy	accuracy	NOUN
ejpam-6087	71	12	at	at	ADP
ejpam-6087	71	13	99.56	99.56	NUM
ejpam-6087	71	14	%	%	NOUN
ejpam-6087	71	15	,	,	PUNCT
ejpam-6087	71	16	followed	follow	VERB
ejpam-6087	71	17	by	by	ADP
ejpam-6087	71	18	rf	rf	ADJ
ejpam-6087	71	19	(	(	PUNCT
ejpam-6087	71	20	99	99	NUM
ejpam-6087	71	21	.	.	NOUN
ejpam-6087	71	22	35	35	NUM
ejpam-6087	71	23	%	%	NOUN
ejpam-6087	71	24	)	)	PUNCT
ejpam-6087	71	25	,	,	PUNCT
ejpam-6087	71	26	lr	lr	X
ejpam-6087	71	27	(	(	PUNCT
ejpam-6087	71	28	98	98	NUM
ejpam-6087	71	29	.	.	NOUN
ejpam-6087	71	30	91	91	NUM
ejpam-6087	71	31	%	%	NOUN
ejpam-6087	71	32	)	)	PUNCT
ejpam-6087	71	33	and	and	CCONJ
ejpam-6087	71	34	nb	nb	INTJ
ejpam-6087	71	35	(	(	PUNCT
ejpam-6087	71	36	94	94	NUM
ejpam-6087	71	37	.	.	SYM
ejpam-6087	71	38	89	89	NUM
ejpam-6087	71	39	%	%	NOUN
ejpam-6087	71	40	)	)	PUNCT
ejpam-6087	71	41	.	.	PUNCT
ejpam-6087	72	1	recall	recall	PROPN
ejpam-6087	72	2	results	result	NOUN
ejpam-6087	72	3	were	be	AUX
ejpam-6087	72	4	also	also	ADV
ejpam-6087	72	5	better	well	ADJ
ejpam-6087	72	6	for	for	ADP
ejpam-6087	72	7	dt	dt	PROPN
ejpam-6087	72	8	and	and	CCONJ
ejpam-6087	72	9	rf	rf	PRON
ejpam-6087	72	10	compared	compare	VERB
ejpam-6087	72	11	to	to	ADP
ejpam-6087	72	12	lr	lr	PROPN
ejpam-6087	72	13	and	and	CCONJ
ejpam-6087	72	14	nb	nb	INTJ
ejpam-6087	72	15	.	.	PUNCT
ejpam-6087	73	1	in	in	ADP
ejpam-6087	73	2	particular	particular	ADJ
ejpam-6087	73	3	,	,	PUNCT
ejpam-6087	73	4	the	the	DET
ejpam-6087	73	5	study	study	NOUN
ejpam-6087	73	6	focused	focus	VERB
ejpam-6087	73	7	on	on	ADP
ejpam-6087	73	8	model	model	NOUN
ejpam-6087	73	9	comparison	comparison	NOUN
ejpam-6087	73	10	rather	rather	ADV
ejpam-6087	73	11	than	than	ADP
ejpam-6087	73	12	exploring	explore	VERB
ejpam-6087	73	13	feature	feature	NOUN
ejpam-6087	73	14	selection	selection	NOUN
ejpam-6087	73	15	methods	method	NOUN
ejpam-6087	73	16	.	.	PUNCT
ejpam-6087	74	1	the	the	DET
ejpam-6087	74	2	research	research	NOUN
ejpam-6087	74	3	proposed	propose	VERB
ejpam-6087	74	4	in	in	ADP
ejpam-6087	74	5	[	[	X
ejpam-6087	74	6	15	15	NUM
ejpam-6087	74	7	]	]	PUNCT
ejpam-6087	74	8	introduced	introduce	VERB
ejpam-6087	74	9	an	an	DET
ejpam-6087	74	10	innovative	innovative	ADJ
ejpam-6087	74	11	approach	approach	NOUN
ejpam-6087	74	12	to	to	ADP
ejpam-6087	74	13	detecting	detect	VERB
ejpam-6087	74	14	fake	fake	ADJ
ejpam-6087	74	15	news	news	NOUN
ejpam-6087	74	16	by	by	ADP
ejpam-6087	74	17	analyzing	analyze	VERB
ejpam-6087	74	18	semantic	semantic	ADJ
ejpam-6087	74	19	discrepancies	discrepancy	NOUN
ejpam-6087	74	20	between	between	ADP
ejpam-6087	74	21	the	the	DET
ejpam-6087	74	22	titles	title	NOUN
ejpam-6087	74	23	and	and	CCONJ
ejpam-6087	74	24	content	content	NOUN
ejpam-6087	74	25	of	of	ADP
ejpam-6087	74	26	the	the	DET
ejpam-6087	74	27	articles	article	NOUN
ejpam-6087	74	28	.	.	PUNCT
ejpam-6087	75	1	using	use	VERB
ejpam-6087	75	2	feature	feature	NOUN
ejpam-6087	75	3	selection	selection	NOUN
ejpam-6087	75	4	,	,	PUNCT
ejpam-6087	75	5	the	the	DET
ejpam-6087	75	6	contents	content	NOUN
ejpam-6087	75	7	of	of	ADP
ejpam-6087	75	8	the	the	DET
ejpam-6087	75	9	articles	article	NOUN
ejpam-6087	75	10	are	be	AUX
ejpam-6087	75	11	summarized	summarize	VERB
ejpam-6087	75	12	,	,	PUNCT
ejpam-6087	75	13	and	and	CCONJ
ejpam-6087	75	14	the	the	DET
ejpam-6087	75	15	vector	vector	NOUN
ejpam-6087	75	16	distances	distance	NOUN
ejpam-6087	75	17	(	(	PUNCT
ejpam-6087	75	18	cosine	cosine	NOUN
ejpam-6087	75	19	similarity	similarity	NOUN
ejpam-6087	75	20	,	,	PUNCT
ejpam-6087	75	21	jaccard	jaccard	ADJ
ejpam-6087	75	22	distance	distance	NOUN
ejpam-6087	75	23	,	,	PUNCT
ejpam-6087	75	24	euclidean	euclidean	ADJ
ejpam-6087	75	25	distance	distance	NOUN
ejpam-6087	75	26	)	)	PUNCT
ejpam-6087	75	27	between	between	ADP
ejpam-6087	75	28	the	the	DET
ejpam-6087	75	29	titles	title	NOUN
ejpam-6087	75	30	and	and	CCONJ
ejpam-6087	75	31	the	the	DET
ejpam-6087	75	32	contents	content	NOUN
ejpam-6087	75	33	are	be	AUX
ejpam-6087	75	34	calculated	calculate	VERB
ejpam-6087	75	35	.	.	PUNCT
ejpam-6087	76	1	these	these	DET
ejpam-6087	76	2	distances	distance	NOUN
ejpam-6087	76	3	are	be	AUX
ejpam-6087	76	4	combined	combine	VERB
ejpam-6087	76	5	as	as	ADP
ejpam-6087	76	6	features	feature	NOUN
ejpam-6087	76	7	to	to	PART
ejpam-6087	76	8	train	train	VERB
ejpam-6087	76	9	the	the	DET
ejpam-6087	76	10	ml	ml	NOUN
ejpam-6087	76	11	and	and	CCONJ
ejpam-6087	76	12	dl	dl	PROPN
ejpam-6087	76	13	models	model	NOUN
ejpam-6087	76	14	,	,	PUNCT
ejpam-6087	76	15	achieving	achieve	VERB
ejpam-6087	76	16	almost	almost	ADV
ejpam-6087	76	17	99.9	99.9	NUM
ejpam-6087	76	18	%	%	NOUN
ejpam-6087	76	19	accuracy	accuracy	NOUN
ejpam-6087	76	20	in	in	ADP
ejpam-6087	76	21	three	three	NUM
ejpam-6087	76	22	data	datum	NOUN
ejpam-6087	76	23	sets	set	NOUN
ejpam-6087	76	24	.	.	PUNCT
ejpam-6087	77	1	this	this	DET
ejpam-6087	77	2	method	method	NOUN
ejpam-6087	77	3	demonstrates	demonstrate	VERB
ejpam-6087	77	4	the	the	DET
ejpam-6087	77	5	effectiveness	effectiveness	NOUN
ejpam-6087	77	6	of	of	ADP
ejpam-6087	77	7	exploiting	exploit	VERB
ejpam-6087	77	8	semantic	semantic	ADJ
ejpam-6087	77	9	differences	difference	NOUN
ejpam-6087	77	10	and	and	CCONJ
ejpam-6087	77	11	underscores	underscore	VERB
ejpam-6087	77	12	the	the	DET
ejpam-6087	77	13	importance	importance	NOUN
ejpam-6087	77	14	of	of	ADP
ejpam-6087	77	15	feature	feature	NOUN
ejpam-6087	77	16	selection	selection	NOUN
ejpam-6087	77	17	in	in	ADP
ejpam-6087	77	18	enhancing	enhance	VERB
ejpam-6087	77	19	detection	detection	NOUN
ejpam-6087	77	20	performance	performance	NOUN
ejpam-6087	77	21	.	.	PUNCT
ejpam-6087	78	1	the	the	DET
ejpam-6087	78	2	approach	approach	NOUN
ejpam-6087	78	3	of	of	ADP
ejpam-6087	78	4	[	[	X
ejpam-6087	78	5	16	16	NUM
ejpam-6087	78	6	]	]	PUNCT
ejpam-6087	78	7	addressed	address	VERB
ejpam-6087	78	8	the	the	DET
ejpam-6087	78	9	rapid	rapid	ADJ
ejpam-6087	78	10	spread	spread	NOUN
ejpam-6087	78	11	of	of	ADP
ejpam-6087	78	12	fake	fake	ADJ
ejpam-6087	78	13	news	news	NOUN
ejpam-6087	78	14	by	by	ADP
ejpam-6087	78	15	investigating	investigate	VERB
ejpam-6087	78	16	stylebased	stylebase	VERB
ejpam-6087	78	17	detection	detection	NOUN
ejpam-6087	78	18	using	use	VERB
ejpam-6087	78	19	text	text	NOUN
ejpam-6087	78	20	analysis	analysis	NOUN
ejpam-6087	78	21	with	with	ADP
ejpam-6087	78	22	natural	natural	ADJ
ejpam-6087	78	23	language	language	NOUN
ejpam-6087	78	24	processing	processing	NOUN
ejpam-6087	78	25	techniques	technique	NOUN
ejpam-6087	78	26	.	.	PUNCT
ejpam-6087	79	1	various	various	ADJ
ejpam-6087	79	2	models	model	NOUN
ejpam-6087	79	3	were	be	AUX
ejpam-6087	79	4	built	build	VERB
ejpam-6087	79	5	using	use	VERB
ejpam-6087	79	6	word	word	NOUN
ejpam-6087	79	7	representations	representation	NOUN
ejpam-6087	79	8	(	(	PUNCT
ejpam-6087	79	9	tf	tf	NOUN
ejpam-6087	79	10	-	-	PUNCT
ejpam-6087	79	11	idf	idf	PROPN
ejpam-6087	79	12	,	,	PUNCT
ejpam-6087	79	13	word2vec	word2vec	PROPN
ejpam-6087	79	14	)	)	PUNCT
ejpam-6087	79	15	and	and	CCONJ
ejpam-6087	79	16	ml	ml	INTJ
ejpam-6087	79	17	(	(	PUNCT
ejpam-6087	79	18	e.g.	e.g.	ADV
ejpam-6087	79	19	,	,	PUNCT
ejpam-6087	79	20	knn	knn	PROPN
ejpam-6087	79	21	,	,	PUNCT
ejpam-6087	79	22	nb	nb	PROPN
ejpam-6087	79	23	,	,	PUNCT
ejpam-6087	79	24	lr	lr	PROPN
ejpam-6087	79	25	)	)	PUNCT
ejpam-6087	79	26	and	and	CCONJ
ejpam-6087	79	27	dl	dl	X
ejpam-6087	79	28	(	(	PUNCT
ejpam-6087	79	29	e.g.	e.g.	ADJ
ejpam-6087	79	30	,	,	PUNCT
ejpam-6087	79	31	lstm	lstm	ADJ
ejpam-6087	79	32	)	)	PUNCT
ejpam-6087	79	33	methods	method	NOUN
ejpam-6087	79	34	on	on	ADP
ejpam-6087	79	35	the	the	DET
ejpam-6087	79	36	isot	isot	NOUN
ejpam-6087	79	37	dataset	dataset	NOUN
ejpam-6087	79	38	.	.	PUNCT
ejpam-6087	80	1	performance	performance	NOUN
ejpam-6087	80	2	was	be	AUX
ejpam-6087	80	3	evaluated	evaluate	VERB
ejpam-6087	80	4	using	use	VERB
ejpam-6087	80	5	accuracy	accuracy	NOUN
ejpam-6087	80	6	,	,	PUNCT
ejpam-6087	80	7	precision	precision	NOUN
ejpam-6087	80	8	,	,	PUNCT
ejpam-6087	80	9	recall	recall	NOUN
ejpam-6087	80	10	,	,	PUNCT
ejpam-6087	80	11	and	and	CCONJ
ejpam-6087	80	12	the	the	DET
ejpam-6087	80	13	f1	f1	PROPN
ejpam-6087	80	14	score	score	NOUN
ejpam-6087	80	15	,	,	PUNCT
ejpam-6087	80	16	with	with	ADP
ejpam-6087	80	17	the	the	DET
ejpam-6087	80	18	lstm	lstm	PROPN
ejpam-6087	80	19	model	model	NOUN
ejpam-6087	80	20	achieving	achieve	VERB
ejpam-6087	80	21	the	the	DET
ejpam-6087	80	22	highest	high	ADJ
ejpam-6087	80	23	accuracy	accuracy	NOUN
ejpam-6087	80	24	at	at	ADP
ejpam-6087	80	25	99.2	99.2	NUM
ejpam-6087	80	26	%	%	NOUN
ejpam-6087	80	27	.	.	PUNCT
ejpam-6087	81	1	although	although	SCONJ
ejpam-6087	81	2	the	the	DET
ejpam-6087	81	3	study	study	NOUN
ejpam-6087	81	4	focused	focus	VERB
ejpam-6087	81	5	on	on	ADP
ejpam-6087	81	6	improving	improve	VERB
ejpam-6087	81	7	word	word	NOUN
ejpam-6087	81	8	representations	representation	NOUN
ejpam-6087	81	9	and	and	CCONJ
ejpam-6087	81	10	classification	classification	NOUN
ejpam-6087	81	11	methods	method	NOUN
ejpam-6087	81	12	,	,	PUNCT
ejpam-6087	81	13	it	it	PRON
ejpam-6087	81	14	did	do	AUX
ejpam-6087	81	15	not	not	PART
ejpam-6087	81	16	investigate	investigate	VERB
ejpam-6087	81	17	feature	feature	NOUN
ejpam-6087	81	18	selection	selection	NOUN
ejpam-6087	81	19	techniques	technique	NOUN
ejpam-6087	81	20	,	,	PUNCT
ejpam-6087	81	21	which	which	PRON
ejpam-6087	81	22	are	be	AUX
ejpam-6087	81	23	also	also	ADV
ejpam-6087	81	24	important	important	ADJ
ejpam-6087	81	25	to	to	PART
ejpam-6087	81	26	optimize	optimize	VERB
ejpam-6087	81	27	model	model	NOUN
ejpam-6087	81	28	performance	performance	NOUN
ejpam-6087	81	29	.	.	PUNCT
ejpam-6087	82	1	the	the	DET
ejpam-6087	82	2	study	study	NOUN
ejpam-6087	82	3	in	in	ADP
ejpam-6087	82	4	[	[	X
ejpam-6087	82	5	17	17	NUM
ejpam-6087	82	6	]	]	PUNCT
ejpam-6087	82	7	explored	explore	VERB
ejpam-6087	82	8	various	various	ADJ
ejpam-6087	82	9	ml	ml	NOUN
ejpam-6087	82	10	algorithms	algorithm	NOUN
ejpam-6087	82	11	to	to	PART
ejpam-6087	82	12	identify	identify	VERB
ejpam-6087	82	13	fake	fake	ADJ
ejpam-6087	82	14	news	news	NOUN
ejpam-6087	82	15	,	,	PUNCT
ejpam-6087	82	16	with	with	ADP
ejpam-6087	82	17	the	the	DET
ejpam-6087	82	18	aim	aim	NOUN
ejpam-6087	82	19	of	of	ADP
ejpam-6087	82	20	determining	determine	VERB
ejpam-6087	82	21	which	which	DET
ejpam-6087	82	22	algorithm	algorithm	NOUN
ejpam-6087	82	23	achieved	achieve	VERB
ejpam-6087	82	24	the	the	DET
ejpam-6087	82	25	highest	high	ADJ
ejpam-6087	82	26	precision	precision	NOUN
ejpam-6087	82	27	.	.	PUNCT
ejpam-6087	83	1	the	the	DET
ejpam-6087	83	2	study	study	NOUN
ejpam-6087	83	3	found	find	VERB
ejpam-6087	83	4	that	that	SCONJ
ejpam-6087	83	5	the	the	DET
ejpam-6087	83	6	dt	dt	PROPN
ejpam-6087	83	7	algorithm	algorithm	PROPN
ejpam-6087	83	8	performed	perform	VERB
ejpam-6087	83	9	the	the	DET
ejpam-6087	83	10	best	good	ADJ
ejpam-6087	83	11	,	,	PUNCT
ejpam-6087	83	12	achieving	achieve	VERB
ejpam-6087	83	13	an	an	DET
ejpam-6087	83	14	accuracy	accuracy	NOUN
ejpam-6087	83	15	of	of	ADP
ejpam-6087	83	16	99.97	99.97	NUM
ejpam-6087	83	17	%	%	NOUN
ejpam-6087	83	18	compared	compare	VERB
ejpam-6087	83	19	to	to	ADP
ejpam-6087	83	20	other	other	ADJ
ejpam-6087	83	21	methods	method	NOUN
ejpam-6087	83	22	.	.	PUNCT
ejpam-6087	84	1	however	however	ADV
ejpam-6087	84	2	,	,	PUNCT
ejpam-6087	84	3	the	the	DET
ejpam-6087	84	4	research	research	NOUN
ejpam-6087	84	5	did	do	AUX
ejpam-6087	84	6	not	not	PART
ejpam-6087	84	7	investigate	investigate	VERB
ejpam-6087	84	8	feature	feature	NOUN
ejpam-6087	84	9	selection	selection	NOUN
ejpam-6087	84	10	techniques	technique	NOUN
ejpam-6087	84	11	,	,	PUNCT
ejpam-6087	84	12	which	which	PRON
ejpam-6087	84	13	are	be	AUX
ejpam-6087	84	14	also	also	ADV
ejpam-6087	84	15	crucial	crucial	ADJ
ejpam-6087	84	16	to	to	ADP
ejpam-6087	84	17	enhancing	enhance	VERB
ejpam-6087	84	18	model	model	NOUN
ejpam-6087	84	19	performance	performance	NOUN
ejpam-6087	84	20	.	.	PUNCT
ejpam-6087	85	1	the	the	DET
ejpam-6087	85	2	study	study	NOUN
ejpam-6087	85	3	in	in	ADP
ejpam-6087	85	4	[	[	X
ejpam-6087	85	5	18	18	NUM
ejpam-6087	85	6	]	]	PUNCT
ejpam-6087	85	7	explored	explore	VERB
ejpam-6087	85	8	various	various	ADJ
ejpam-6087	85	9	ml	ml	NOUN
ejpam-6087	85	10	models	model	NOUN
ejpam-6087	85	11	to	to	PART
ejpam-6087	85	12	classify	classify	VERB
ejpam-6087	85	13	the	the	DET
ejpam-6087	85	14	veracity	veracity	NOUN
ejpam-6087	85	15	of	of	ADP
ejpam-6087	85	16	information	information	NOUN
ejpam-6087	85	17	and	and	CCONJ
ejpam-6087	85	18	detect	detect	VERB
ejpam-6087	85	19	fake	fake	ADJ
ejpam-6087	85	20	news	news	NOUN
ejpam-6087	85	21	on	on	ADP
ejpam-6087	85	22	social	social	ADJ
ejpam-6087	85	23	networks	network	NOUN
ejpam-6087	85	24	.	.	PUNCT
ejpam-6087	86	1	the	the	DET
ejpam-6087	86	2	analysis	analysis	NOUN
ejpam-6087	86	3	used	use	VERB
ejpam-6087	86	4	a	a	DET
ejpam-6087	86	5	dataset	dataset	NOUN
ejpam-6087	86	6	of	of	ADP
ejpam-6087	86	7	approximately	approximately	ADV
ejpam-6087	86	8	40,000	40,000	NUM
ejpam-6087	86	9	m.	m.	NOUN
ejpam-6087	86	10	al	al	PROPN
ejpam-6087	86	11	-	-	PUNCT
ejpam-6087	86	12	tarawneh	tarawneh	PROPN
ejpam-6087	86	13	et	et	PROPN
ejpam-6087	86	14	al	al	PROPN
ejpam-6087	86	15	.	.	PUNCT
ejpam-6087	86	16	/	/	SYM
ejpam-6087	86	17	eur	eur	PROPN
ejpam-6087	86	18	.	.	PUNCT
ejpam-6087	87	1	j.	j.	PROPN
ejpam-6087	87	2	pure	pure	PROPN
ejpam-6087	87	3	appl	appl	PROPN
ejpam-6087	87	4	.	.	PROPN
ejpam-6087	87	5	math	math	PROPN
ejpam-6087	87	6	,	,	PUNCT
ejpam-6087	87	7	18	18	NUM
ejpam-6087	87	8	(	(	PUNCT
ejpam-6087	87	9	2	2	NUM
ejpam-6087	87	10	)	)	PUNCT
ejpam-6087	87	11	(	(	PUNCT
ejpam-6087	87	12	2025	2025	NUM
ejpam-6087	87	13	)	)	PUNCT
ejpam-6087	87	14	,	,	PUNCT
ejpam-6087	87	15	6087	6087	NUM
ejpam-6087	87	16	5	5	NUM
ejpam-6087	87	17	of	of	ADP
ejpam-6087	87	18	39	39	NUM
ejpam-6087	87	19	items	item	NOUN
ejpam-6087	87	20	(	(	PUNCT
ejpam-6087	87	21	20,000	20,000	NUM
ejpam-6087	87	22	fake	fake	ADJ
ejpam-6087	87	23	and	and	CCONJ
ejpam-6087	87	24	20,000	20,000	NUM
ejpam-6087	87	25	real	real	ADJ
ejpam-6087	87	26	news	news	NOUN
ejpam-6087	87	27	)	)	PUNCT
ejpam-6087	87	28	and	and	CCONJ
ejpam-6087	87	29	used	use	VERB
ejpam-6087	87	30	a	a	DET
ejpam-6087	87	31	variety	variety	NOUN
ejpam-6087	87	32	of	of	ADP
ejpam-6087	87	33	models	model	NOUN
ejpam-6087	87	34	,	,	PUNCT
ejpam-6087	87	35	including	include	VERB
ejpam-6087	87	36	svm	svm	PROPN
ejpam-6087	87	37	,	,	PUNCT
ejpam-6087	87	38	lr	lr	NOUN
ejpam-6087	87	39	,	,	PUNCT
ejpam-6087	87	40	catboost	catboost	PROPN
ejpam-6087	87	41	,	,	PUNCT
ejpam-6087	87	42	xgboost	xgboost	ADV
ejpam-6087	87	43	,	,	PUNCT
ejpam-6087	87	44	multinomial	multinomial	PROPN
ejpam-6087	87	45	nb	nb	PROPN
ejpam-6087	87	46	,	,	PUNCT
ejpam-6087	87	47	and	and	CCONJ
ejpam-6087	87	48	rf	rf	NOUN
ejpam-6087	87	49	.	.	NOUN
ejpam-6087	87	50	performance	performance	NOUN
ejpam-6087	87	51	was	be	AUX
ejpam-6087	87	52	evaluated	evaluate	VERB
ejpam-6087	87	53	using	use	VERB
ejpam-6087	87	54	metrics	metric	NOUN
ejpam-6087	87	55	such	such	ADJ
ejpam-6087	87	56	as	as	ADP
ejpam-6087	87	57	accuracy	accuracy	NOUN
ejpam-6087	87	58	,	,	PUNCT
ejpam-6087	87	59	precision	precision	NOUN
ejpam-6087	87	60	,	,	PUNCT
ejpam-6087	87	61	recall	recall	NOUN
ejpam-6087	87	62	,	,	PUNCT
ejpam-6087	87	63	f1	f1	NOUN
ejpam-6087	87	64	score	score	NOUN
ejpam-6087	87	65	,	,	PUNCT
ejpam-6087	87	66	and	and	CCONJ
ejpam-6087	87	67	more	more	ADJ
ejpam-6087	87	68	.	.	PUNCT
ejpam-6087	88	1	the	the	DET
ejpam-6087	88	2	deep	deep	ADJ
ejpam-6087	88	3	auto	auto	NOUN
ejpam-6087	88	4	viml	viml	NOUN
ejpam-6087	88	5	model	model	NOUN
ejpam-6087	88	6	achieved	achieve	VERB
ejpam-6087	88	7	the	the	DET
ejpam-6087	88	8	highest	high	ADJ
ejpam-6087	88	9	accuracy	accuracy	NOUN
ejpam-6087	88	10	,	,	PUNCT
ejpam-6087	88	11	precision	precision	NOUN
ejpam-6087	88	12	,	,	PUNCT
ejpam-6087	88	13	recall	recall	NOUN
ejpam-6087	88	14	and	and	CCONJ
ejpam-6087	88	15	f1	f1	ADJ
ejpam-6087	88	16	score	score	NOUN
ejpam-6087	88	17	at	at	ADP
ejpam-6087	88	18	99	99	NUM
ejpam-6087	88	19	%	%	NOUN
ejpam-6087	88	20	,	,	PUNCT
ejpam-6087	88	21	while	while	SCONJ
ejpam-6087	88	22	the	the	DET
ejpam-6087	88	23	hybrid	hybrid	ADJ
ejpam-6087	88	24	learning	learning	NOUN
ejpam-6087	88	25	model	model	NOUN
ejpam-6087	88	26	had	have	VERB
ejpam-6087	88	27	the	the	DET
ejpam-6087	88	28	best	good	ADJ
ejpam-6087	88	29	false	false	ADJ
ejpam-6087	88	30	rejection	rejection	NOUN
ejpam-6087	88	31	rate	rate	NOUN
ejpam-6087	88	32	at	at	ADP
ejpam-6087	88	33	71	71	NUM
ejpam-6087	88	34	%	%	NOUN
ejpam-6087	88	35	.	.	PUNCT
ejpam-6087	89	1	the	the	DET
ejpam-6087	89	2	svm	svm	PROPN
ejpam-6087	89	3	was	be	AUX
ejpam-6087	89	4	noted	note	VERB
ejpam-6087	89	5	for	for	ADP
ejpam-6087	89	6	its	its	PRON
ejpam-6087	89	7	computational	computational	ADJ
ejpam-6087	89	8	efficiency	efficiency	NOUN
ejpam-6087	89	9	,	,	PUNCT
ejpam-6087	89	10	taking	take	VERB
ejpam-6087	89	11	only	only	ADV
ejpam-6087	89	12	0.245	0.245	NUM
ejpam-6087	89	13	ms	ms	NOUN
ejpam-6087	89	14	for	for	ADP
ejpam-6087	89	15	computation	computation	NOUN
ejpam-6087	89	16	.	.	PUNCT
ejpam-6087	90	1	however	however	ADV
ejpam-6087	90	2	,	,	PUNCT
ejpam-6087	90	3	the	the	DET
ejpam-6087	90	4	study	study	NOUN
ejpam-6087	90	5	did	do	AUX
ejpam-6087	90	6	not	not	PART
ejpam-6087	90	7	investigate	investigate	VERB
ejpam-6087	90	8	feature	feature	NOUN
ejpam-6087	90	9	selection	selection	NOUN
ejpam-6087	90	10	techniques	technique	NOUN
ejpam-6087	90	11	,	,	PUNCT
ejpam-6087	90	12	which	which	PRON
ejpam-6087	90	13	are	be	AUX
ejpam-6087	90	14	also	also	ADV
ejpam-6087	90	15	crucial	crucial	ADJ
ejpam-6087	90	16	to	to	ADP
ejpam-6087	90	17	enhancing	enhance	VERB
ejpam-6087	90	18	model	model	NOUN
ejpam-6087	90	19	performance	performance	NOUN
ejpam-6087	90	20	.	.	PUNCT
ejpam-6087	91	1	the	the	DET
ejpam-6087	91	2	authors	author	NOUN
ejpam-6087	91	3	of	of	ADP
ejpam-6087	91	4	[	[	X
ejpam-6087	91	5	19	19	NUM
ejpam-6087	91	6	]	]	PUNCT
ejpam-6087	91	7	proposed	propose	VERB
ejpam-6087	91	8	a	a	DET
ejpam-6087	91	9	method	method	NOUN
ejpam-6087	91	10	that	that	PRON
ejpam-6087	91	11	uses	use	VERB
ejpam-6087	91	12	ml	ml	PART
ejpam-6087	91	13	classification	classification	NOUN
ejpam-6087	91	14	to	to	PART
ejpam-6087	91	15	distinguish	distinguish	VERB
ejpam-6087	91	16	between	between	ADP
ejpam-6087	91	17	genuine	genuine	ADJ
ejpam-6087	91	18	and	and	CCONJ
ejpam-6087	91	19	fake	fake	ADJ
ejpam-6087	91	20	content	content	NOUN
ejpam-6087	91	21	.	.	PUNCT
ejpam-6087	92	1	the	the	DET
ejpam-6087	92	2	approach	approach	NOUN
ejpam-6087	92	3	used	use	VERB
ejpam-6087	92	4	natural	natural	ADJ
ejpam-6087	92	5	language	language	NOUN
ejpam-6087	92	6	processing	processing	NOUN
ejpam-6087	92	7	techniques	technique	NOUN
ejpam-6087	92	8	,	,	PUNCT
ejpam-6087	92	9	including	include	VERB
ejpam-6087	92	10	the	the	DET
ejpam-6087	92	11	tf	tf	PROPN
ejpam-6087	92	12	-	-	PUNCT
ejpam-6087	92	13	idf	idf	PROPN
ejpam-6087	92	14	and	and	CCONJ
ejpam-6087	92	15	word2vec	word2vec	NOUN
ejpam-6087	92	16	features	feature	NOUN
ejpam-6087	92	17	,	,	PUNCT
ejpam-6087	92	18	which	which	PRON
ejpam-6087	92	19	were	be	AUX
ejpam-6087	92	20	optimized	optimize	VERB
ejpam-6087	92	21	using	use	VERB
ejpam-6087	92	22	the	the	DET
ejpam-6087	92	23	rfe	rfe	NOUN
ejpam-6087	92	24	algorithm	algorithm	NOUN
ejpam-6087	92	25	for	for	ADP
ejpam-6087	92	26	feature	feature	NOUN
ejpam-6087	92	27	selection	selection	NOUN
ejpam-6087	92	28	.	.	PUNCT
ejpam-6087	93	1	training	training	NOUN
ejpam-6087	93	2	and	and	CCONJ
ejpam-6087	93	3	testing	testing	NOUN
ejpam-6087	93	4	were	be	AUX
ejpam-6087	93	5	performed	perform	VERB
ejpam-6087	93	6	on	on	ADP
ejpam-6087	93	7	the	the	DET
ejpam-6087	93	8	kaggle	kaggle	ADJ
ejpam-6087	93	9	’	'	PUNCT
ejpam-6087	93	10	fake	fake	ADJ
ejpam-6087	93	11	news	news	NOUN
ejpam-6087	93	12	detection	detection	NOUN
ejpam-6087	93	13	’	'	PUNCT
ejpam-6087	93	14	dataset	dataset	NOUN
ejpam-6087	93	15	,	,	PUNCT
ejpam-6087	93	16	and	and	CCONJ
ejpam-6087	93	17	the	the	DET
ejpam-6087	93	18	results	result	NOUN
ejpam-6087	93	19	showed	show	VERB
ejpam-6087	93	20	that	that	SCONJ
ejpam-6087	93	21	the	the	DET
ejpam-6087	93	22	boosting	boost	VERB
ejpam-6087	93	23	ensemble	ensemble	ADJ
ejpam-6087	93	24	methods	method	NOUN
ejpam-6087	93	25	outperformed	outperform	VERB
ejpam-6087	93	26	other	other	ADJ
ejpam-6087	93	27	techniques	technique	NOUN
ejpam-6087	93	28	.	.	PUNCT
ejpam-6087	94	1	the	the	DET
ejpam-6087	94	2	study	study	NOUN
ejpam-6087	94	3	highlighted	highlight	VERB
ejpam-6087	94	4	the	the	DET
ejpam-6087	94	5	effectiveness	effectiveness	NOUN
ejpam-6087	94	6	of	of	ADP
ejpam-6087	94	7	feature	feature	NOUN
ejpam-6087	94	8	selection	selection	NOUN
ejpam-6087	94	9	through	through	ADP
ejpam-6087	94	10	rfe	rfe	NOUN
ejpam-6087	94	11	,	,	PUNCT
ejpam-6087	94	12	but	but	CCONJ
ejpam-6087	94	13	did	do	AUX
ejpam-6087	94	14	not	not	PART
ejpam-6087	94	15	explore	explore	VERB
ejpam-6087	94	16	other	other	ADJ
ejpam-6087	94	17	feature	feature	NOUN
ejpam-6087	94	18	selection	selection	NOUN
ejpam-6087	94	19	methods	method	NOUN
ejpam-6087	94	20	.	.	PUNCT
ejpam-6087	95	1	the	the	DET
ejpam-6087	95	2	study	study	NOUN
ejpam-6087	95	3	in	in	ADP
ejpam-6087	95	4	[	[	X
ejpam-6087	95	5	20	20	NUM
ejpam-6087	95	6	]	]	PUNCT
ejpam-6087	95	7	aimed	aim	VERB
ejpam-6087	95	8	to	to	PART
ejpam-6087	95	9	build	build	VERB
ejpam-6087	95	10	a	a	DET
ejpam-6087	95	11	model	model	NOUN
ejpam-6087	95	12	to	to	PART
ejpam-6087	95	13	detect	detect	VERB
ejpam-6087	95	14	fake	fake	ADJ
ejpam-6087	95	15	news	news	NOUN
ejpam-6087	95	16	by	by	ADP
ejpam-6087	95	17	analyzing	analyze	VERB
ejpam-6087	95	18	the	the	DET
ejpam-6087	95	19	characteristics	characteristic	NOUN
ejpam-6087	95	20	of	of	ADP
ejpam-6087	95	21	the	the	DET
ejpam-6087	95	22	text	text	NOUN
ejpam-6087	95	23	.	.	PUNCT
ejpam-6087	96	1	tf	tf	ADJ
ejpam-6087	96	2	-	-	PUNCT
ejpam-6087	96	3	idf	idf	PROPN
ejpam-6087	96	4	technology	technology	NOUN
ejpam-6087	96	5	was	be	AUX
ejpam-6087	96	6	used	use	VERB
ejpam-6087	96	7	to	to	PART
ejpam-6087	96	8	convert	convert	VERB
ejpam-6087	96	9	words	word	NOUN
ejpam-6087	96	10	into	into	ADP
ejpam-6087	96	11	features	feature	NOUN
ejpam-6087	96	12	and	and	CCONJ
ejpam-6087	96	13	identify	identify	VERB
ejpam-6087	96	14	the	the	DET
ejpam-6087	96	15	highest	highest	ADV
ejpam-6087	96	16	-	-	PUNCT
ejpam-6087	96	17	ranking	rank	VERB
ejpam-6087	96	18	features	feature	NOUN
ejpam-6087	96	19	to	to	PART
ejpam-6087	96	20	distinguish	distinguish	VERB
ejpam-6087	96	21	between	between	ADP
ejpam-6087	96	22	real	real	ADJ
ejpam-6087	96	23	and	and	CCONJ
ejpam-6087	96	24	fake	fake	ADJ
ejpam-6087	96	25	news	news	NOUN
ejpam-6087	96	26	.	.	PUNCT
ejpam-6087	97	1	the	the	DET
ejpam-6087	97	2	study	study	NOUN
ejpam-6087	97	3	then	then	ADV
ejpam-6087	97	4	adapted	adapt	VERB
ejpam-6087	97	5	ml	ml	NOUN
ejpam-6087	97	6	techniques	technique	NOUN
ejpam-6087	97	7	including	include	VERB
ejpam-6087	97	8	lr	lr	NOUN
ejpam-6087	97	9	,	,	PUNCT
ejpam-6087	97	10	dt	dt	ADJ
ejpam-6087	97	11	,	,	PUNCT
ejpam-6087	97	12	gradient	gradient	ADJ
ejpam-6087	97	13	boosting	boosting	NOUN
ejpam-6087	97	14	and	and	CCONJ
ejpam-6087	97	15	rf	rf	VERB
ejpam-6087	97	16	,	,	PUNCT
ejpam-6087	97	17	with	with	ADP
ejpam-6087	97	18	dt	dt	NOUN
ejpam-6087	97	19	and	and	CCONJ
ejpam-6087	97	20	gradient	gradient	ADJ
ejpam-6087	97	21	boosting	boost	VERB
ejpam-6087	97	22	achieving	achieve	VERB
ejpam-6087	97	23	the	the	DET
ejpam-6087	97	24	highest	high	ADJ
ejpam-6087	97	25	precision	precision	NOUN
ejpam-6087	97	26	at	at	ADP
ejpam-6087	97	27	99.4	99.4	NUM
ejpam-6087	97	28	%	%	NOUN
ejpam-6087	97	29	and	and	CCONJ
ejpam-6087	97	30	99.49	99.49	NUM
ejpam-6087	97	31	%	%	NOUN
ejpam-6087	97	32	,	,	PUNCT
ejpam-6087	97	33	respectively	respectively	ADV
ejpam-6087	97	34	.	.	PUNCT
ejpam-6087	98	1	although	although	SCONJ
ejpam-6087	98	2	the	the	DET
ejpam-6087	98	3	study	study	NOUN
ejpam-6087	98	4	focused	focus	VERB
ejpam-6087	98	5	on	on	ADP
ejpam-6087	98	6	feature	feature	NOUN
ejpam-6087	98	7	extraction	extraction	NOUN
ejpam-6087	98	8	through	through	ADP
ejpam-6087	98	9	tf	tf	PROPN
ejpam-6087	98	10	-	-	PUNCT
ejpam-6087	98	11	idf	idf	PROPN
ejpam-6087	98	12	,	,	PUNCT
ejpam-6087	98	13	it	it	PRON
ejpam-6087	98	14	did	do	AUX
ejpam-6087	98	15	not	not	PART
ejpam-6087	98	16	explore	explore	VERB
ejpam-6087	98	17	other	other	ADJ
ejpam-6087	98	18	feature	feature	NOUN
ejpam-6087	98	19	selection	selection	NOUN
ejpam-6087	98	20	methods	method	NOUN
ejpam-6087	98	21	,	,	PUNCT
ejpam-6087	98	22	which	which	PRON
ejpam-6087	98	23	could	could	AUX
ejpam-6087	98	24	have	have	VERB
ejpam-6087	98	25	further	far	ADV
ejpam-6087	98	26	enhanced	enhance	VERB
ejpam-6087	98	27	model	model	NOUN
ejpam-6087	98	28	performance	performance	NOUN
ejpam-6087	98	29	.	.	PUNCT
ejpam-6087	99	1	the	the	DET
ejpam-6087	99	2	research	research	NOUN
ejpam-6087	99	3	presented	present	VERB
ejpam-6087	99	4	in	in	ADP
ejpam-6087	99	5	[	[	X
ejpam-6087	99	6	21	21	NUM
ejpam-6087	99	7	]	]	PUNCT
ejpam-6087	99	8	evaluated	evaluate	VERB
ejpam-6087	99	9	the	the	DET
ejpam-6087	99	10	effectiveness	effectiveness	NOUN
ejpam-6087	99	11	of	of	ADP
ejpam-6087	99	12	three	three	NUM
ejpam-6087	99	13	ml	ml	NOUN
ejpam-6087	99	14	algorithmsmultinomial	algorithmsmultinomial	PROPN
ejpam-6087	99	15	nb	nb	PROPN
ejpam-6087	99	16	,	,	PUNCT
ejpam-6087	99	17	passive	passive	ADJ
ejpam-6087	99	18	aggressive	aggressive	ADJ
ejpam-6087	99	19	classifier	classifier	NOUN
ejpam-6087	99	20	,	,	PUNCT
ejpam-6087	99	21	and	and	CCONJ
ejpam-6087	99	22	lr	lr	NOUN
ejpam-6087	99	23	-	-	PUNCT
ejpam-6087	99	24	to	to	PART
ejpam-6087	99	25	distinguish	distinguish	VERB
ejpam-6087	99	26	between	between	ADP
ejpam-6087	99	27	fake	fake	ADJ
ejpam-6087	99	28	and	and	CCONJ
ejpam-6087	99	29	genuine	genuine	ADJ
ejpam-6087	99	30	news	news	NOUN
ejpam-6087	99	31	articles	article	NOUN
ejpam-6087	99	32	.	.	PUNCT
ejpam-6087	100	1	using	use	VERB
ejpam-6087	100	2	a	a	DET
ejpam-6087	100	3	balanced	balanced	ADJ
ejpam-6087	100	4	dataset	dataset	NOUN
ejpam-6087	100	5	,	,	PUNCT
ejpam-6087	100	6	the	the	DET
ejpam-6087	100	7	study	study	NOUN
ejpam-6087	100	8	processed	process	VERB
ejpam-6087	100	9	and	and	CCONJ
ejpam-6087	100	10	vectorized	vectorize	VERB
ejpam-6087	100	11	the	the	DET
ejpam-6087	100	12	text	text	NOUN
ejpam-6087	100	13	with	with	ADP
ejpam-6087	100	14	tf	tf	PROPN
ejpam-6087	100	15	-	-	PUNCT
ejpam-6087	100	16	idf	idf	PROPN
ejpam-6087	100	17	and	and	CCONJ
ejpam-6087	100	18	assessed	assess	VERB
ejpam-6087	100	19	algorithms	algorithm	NOUN
ejpam-6087	100	20	based	base	VERB
ejpam-6087	100	21	on	on	ADP
ejpam-6087	100	22	precision	precision	NOUN
ejpam-6087	100	23	,	,	PUNCT
ejpam-6087	100	24	recall	recall	NOUN
ejpam-6087	100	25	and	and	CCONJ
ejpam-6087	100	26	f1	f1	PROPN
ejpam-6087	100	27	score	score	NOUN
ejpam-6087	100	28	.	.	PUNCT
ejpam-6087	101	1	the	the	DET
ejpam-6087	101	2	passive	passive	ADJ
ejpam-6087	101	3	aggressive	aggressive	ADJ
ejpam-6087	101	4	classifier	classifier	NOUN
ejpam-6087	101	5	achieved	achieve	VERB
ejpam-6087	101	6	the	the	DET
ejpam-6087	101	7	highest	high	ADJ
ejpam-6087	101	8	performance	performance	NOUN
ejpam-6087	101	9	with	with	ADP
ejpam-6087	101	10	a	a	DET
ejpam-6087	101	11	precision	precision	NOUN
ejpam-6087	101	12	and	and	CCONJ
ejpam-6087	101	13	recall	recall	NOUN
ejpam-6087	101	14	of	of	ADP
ejpam-6087	101	15	0.99	0.99	NUM
ejpam-6087	101	16	,	,	PUNCT
ejpam-6087	101	17	while	while	SCONJ
ejpam-6087	101	18	lr	lr	NOUN
ejpam-6087	101	19	had	have	VERB
ejpam-6087	101	20	an	an	DET
ejpam-6087	101	21	accuracy	accuracy	NOUN
ejpam-6087	101	22	of	of	ADP
ejpam-6087	101	23	0.98	0.98	NUM
ejpam-6087	101	24	and	and	CCONJ
ejpam-6087	101	25	multinomial	multinomial	ADJ
ejpam-6087	101	26	nb	nb	PROPN
ejpam-6087	101	27	exhibited	exhibit	VERB
ejpam-6087	101	28	a	a	DET
ejpam-6087	101	29	robust	robust	ADJ
ejpam-6087	101	30	recall	recall	NOUN
ejpam-6087	101	31	of	of	ADP
ejpam-6087	101	32	100	100	NUM
ejpam-6087	101	33	%	%	NOUN
ejpam-6087	101	34	but	but	CCONJ
ejpam-6087	101	35	lower	low	ADJ
ejpam-6087	101	36	precision	precision	NOUN
ejpam-6087	101	37	at	at	ADP
ejpam-6087	101	38	91	91	NUM
ejpam-6087	101	39	%	%	NOUN
ejpam-6087	101	40	,	,	PUNCT
ejpam-6087	101	41	resulting	result	VERB
ejpam-6087	101	42	in	in	ADP
ejpam-6087	101	43	an	an	DET
ejpam-6087	101	44	accuracy	accuracy	NOUN
ejpam-6087	101	45	of	of	ADP
ejpam-6087	101	46	95	95	NUM
ejpam-6087	101	47	%	%	NOUN
ejpam-6087	101	48	.	.	PUNCT
ejpam-6087	102	1	the	the	DET
ejpam-6087	102	2	study	study	NOUN
ejpam-6087	102	3	did	do	AUX
ejpam-6087	102	4	not	not	PART
ejpam-6087	102	5	explore	explore	VERB
ejpam-6087	102	6	advanced	advanced	ADJ
ejpam-6087	102	7	feature	feature	NOUN
ejpam-6087	102	8	selection	selection	NOUN
ejpam-6087	102	9	methods	method	NOUN
ejpam-6087	102	10	beyond	beyond	ADP
ejpam-6087	102	11	tf	tf	PROPN
ejpam-6087	102	12	-	-	PUNCT
ejpam-6087	102	13	idf	idf	PROPN
ejpam-6087	102	14	,	,	PUNCT
ejpam-6087	102	15	which	which	PRON
ejpam-6087	102	16	could	could	AUX
ejpam-6087	102	17	have	have	AUX
ejpam-6087	102	18	further	far	ADV
ejpam-6087	102	19	improved	improve	VERB
ejpam-6087	102	20	the	the	DET
ejpam-6087	102	21	models	model	NOUN
ejpam-6087	102	22	’	'	PUNCT
ejpam-6087	102	23	performance	performance	NOUN
ejpam-6087	102	24	.	.	PUNCT
ejpam-6087	103	1	the	the	DET
ejpam-6087	103	2	paper	paper	NOUN
ejpam-6087	103	3	in	in	ADP
ejpam-6087	103	4	[	[	X
ejpam-6087	103	5	22	22	NUM
ejpam-6087	103	6	]	]	PUNCT
ejpam-6087	103	7	addressed	address	VERB
ejpam-6087	103	8	the	the	DET
ejpam-6087	103	9	challenge	challenge	NOUN
ejpam-6087	103	10	of	of	ADP
ejpam-6087	103	11	fake	fake	ADJ
ejpam-6087	103	12	news	news	NOUN
ejpam-6087	103	13	detection	detection	NOUN
ejpam-6087	103	14	by	by	ADP
ejpam-6087	103	15	proposing	propose	VERB
ejpam-6087	103	16	the	the	DET
ejpam-6087	103	17	use	use	NOUN
ejpam-6087	103	18	of	of	ADP
ejpam-6087	103	19	various	various	ADJ
ejpam-6087	103	20	ai	ai	NOUN
ejpam-6087	103	21	and	and	CCONJ
ejpam-6087	103	22	ml	ml	ADP
ejpam-6087	103	23	methods	method	NOUN
ejpam-6087	103	24	,	,	PUNCT
ejpam-6087	103	25	combining	combine	VERB
ejpam-6087	103	26	both	both	CCONJ
ejpam-6087	103	27	individual	individual	ADJ
ejpam-6087	103	28	and	and	CCONJ
ejpam-6087	103	29	ensemble	ensemble	ADJ
ejpam-6087	103	30	approaches	approach	NOUN
ejpam-6087	103	31	for	for	ADP
ejpam-6087	103	32	automated	automate	VERB
ejpam-6087	103	33	news	news	NOUN
ejpam-6087	103	34	classification	classification	NOUN
ejpam-6087	103	35	.	.	PUNCT
ejpam-6087	104	1	the	the	DET
ejpam-6087	104	2	study	study	NOUN
ejpam-6087	104	3	explored	explore	VERB
ejpam-6087	104	4	different	different	ADJ
ejpam-6087	104	5	textual	textual	ADJ
ejpam-6087	104	6	properties	property	NOUN
ejpam-6087	104	7	to	to	PART
ejpam-6087	104	8	distinguish	distinguish	VERB
ejpam-6087	104	9	fake	fake	ADJ
ejpam-6087	104	10	news	news	NOUN
ejpam-6087	104	11	from	from	ADP
ejpam-6087	104	12	real	real	ADJ
ejpam-6087	104	13	news	news	NOUN
ejpam-6087	104	14	and	and	CCONJ
ejpam-6087	104	15	applied	apply	VERB
ejpam-6087	104	16	multiple	multiple	ADJ
ejpam-6087	104	17	ml	ml	NOUN
ejpam-6087	104	18	algorithms	algorithm	NOUN
ejpam-6087	104	19	to	to	ADP
ejpam-6087	104	20	a	a	DET
ejpam-6087	104	21	real	real	ADJ
ejpam-6087	104	22	-	-	PUNCT
ejpam-6087	104	23	world	world	NOUN
ejpam-6087	104	24	dataset	dataset	NOUN
ejpam-6087	104	25	for	for	ADP
ejpam-6087	104	26	this	this	DET
ejpam-6087	104	27	classification	classification	NOUN
ejpam-6087	104	28	task	task	NOUN
ejpam-6087	104	29	.	.	PUNCT
ejpam-6087	105	1	the	the	DET
ejpam-6087	105	2	performance	performance	NOUN
ejpam-6087	105	3	of	of	ADP
ejpam-6087	105	4	these	these	DET
ejpam-6087	105	5	models	model	NOUN
ejpam-6087	105	6	was	be	AUX
ejpam-6087	105	7	evaluated	evaluate	VERB
ejpam-6087	105	8	on	on	ADP
ejpam-6087	105	9	the	the	DET
ejpam-6087	105	10	basis	basis	NOUN
ejpam-6087	105	11	of	of	ADP
ejpam-6087	105	12	their	their	PRON
ejpam-6087	105	13	accuracy	accuracy	NOUN
ejpam-6087	105	14	in	in	ADP
ejpam-6087	105	15	identifying	identify	VERB
ejpam-6087	105	16	fake	fake	ADJ
ejpam-6087	105	17	news	news	NOUN
ejpam-6087	105	18	.	.	PUNCT
ejpam-6087	106	1	however	however	ADV
ejpam-6087	106	2	,	,	PUNCT
ejpam-6087	106	3	the	the	DET
ejpam-6087	106	4	study	study	NOUN
ejpam-6087	106	5	did	do	AUX
ejpam-6087	106	6	not	not	PART
ejpam-6087	106	7	discuss	discuss	VERB
ejpam-6087	106	8	advanced	advanced	ADJ
ejpam-6087	106	9	feature	feature	NOUN
ejpam-6087	106	10	selection	selection	NOUN
ejpam-6087	106	11	techniques	technique	NOUN
ejpam-6087	106	12	beyond	beyond	ADP
ejpam-6087	106	13	basic	basic	ADJ
ejpam-6087	106	14	textual	textual	ADJ
ejpam-6087	106	15	properties	property	NOUN
ejpam-6087	106	16	.	.	PUNCT
ejpam-6087	107	1	the	the	DET
ejpam-6087	107	2	research	research	NOUN
ejpam-6087	107	3	in	in	ADP
ejpam-6087	107	4	[	[	X
ejpam-6087	107	5	23	23	NUM
ejpam-6087	107	6	]	]	PUNCT
ejpam-6087	107	7	addressed	address	VERB
ejpam-6087	107	8	the	the	DET
ejpam-6087	107	9	challenge	challenge	NOUN
ejpam-6087	107	10	of	of	ADP
ejpam-6087	107	11	fake	fake	ADJ
ejpam-6087	107	12	news	news	NOUN
ejpam-6087	107	13	detection	detection	NOUN
ejpam-6087	107	14	in	in	ADP
ejpam-6087	107	15	pakistan	pakistan	PROPN
ejpam-6087	107	16	by	by	ADP
ejpam-6087	107	17	creating	create	VERB
ejpam-6087	107	18	a	a	DET
ejpam-6087	107	19	comprehensive	comprehensive	ADJ
ejpam-6087	107	20	dataset	dataset	NOUN
ejpam-6087	107	21	from	from	ADP
ejpam-6087	107	22	multiple	multiple	ADJ
ejpam-6087	107	23	fact	fact	NOUN
ejpam-6087	107	24	-	-	PUNCT
ejpam-6087	107	25	checked	check	VERB
ejpam-6087	107	26	news	news	NOUN
ejpam-6087	107	27	apis	apis	ADJ
ejpam-6087	107	28	and	and	CCONJ
ejpam-6087	107	29	evaluating	evaluate	VERB
ejpam-6087	107	30	it	it	PRON
ejpam-6087	107	31	using	use	VERB
ejpam-6087	107	32	a	a	DET
ejpam-6087	107	33	range	range	NOUN
ejpam-6087	107	34	of	of	ADP
ejpam-6087	107	35	ml	ml	NOUN
ejpam-6087	107	36	and	and	CCONJ
ejpam-6087	107	37	dl	dl	PROPN
ejpam-6087	107	38	techniques	technique	NOUN
ejpam-6087	107	39	.	.	PUNCT
ejpam-6087	108	1	the	the	DET
ejpam-6087	108	2	study	study	NOUN
ejpam-6087	108	3	applied	apply	VERB
ejpam-6087	108	4	five	five	NUM
ejpam-6087	108	5	ml	ml	NOUN
ejpam-6087	108	6	algorithms	algorithm	NOUN
ejpam-6087	108	7	:	:	PUNCT
ejpam-6087	108	8	nb	nb	PROPN
ejpam-6087	108	9	,	,	PUNCT
ejpam-6087	108	10	knn	knn	PROPN
ejpam-6087	108	11	,	,	PUNCT
ejpam-6087	108	12	lr	lr	PROPN
ejpam-6087	108	13	,	,	PUNCT
ejpam-6087	108	14	svm	svm	ADJ
ejpam-6087	108	15	,	,	PUNCT
ejpam-6087	108	16	and	and	CCONJ
ejpam-6087	108	17	dt	dt	X
ejpam-6087	108	18	,	,	PUNCT
ejpam-6087	108	19	as	as	ADV
ejpam-6087	108	20	well	well	ADV
ejpam-6087	108	21	as	as	ADP
ejpam-6087	108	22	two	two	NUM
ejpam-6087	108	23	dl	dl	PROPN
ejpam-6087	108	24	models	model	NOUN
ejpam-6087	108	25	:	:	PUNCT
ejpam-6087	108	26	cnn	cnn	PROPN
ejpam-6087	108	27	and	and	CCONJ
ejpam-6087	108	28	lstm	lstm	NOUN
ejpam-6087	108	29	using	use	VERB
ejpam-6087	108	30	glove	glove	NOUN
ejpam-6087	108	31	and	and	CCONJ
ejpam-6087	108	32	bert	bert	NOUN
ejpam-6087	108	33	embeddings	embedding	NOUN
ejpam-6087	108	34	for	for	ADP
ejpam-6087	108	35	text	text	NOUN
ejpam-6087	108	36	representation	representation	NOUN
ejpam-6087	108	37	.	.	PUNCT
ejpam-6087	109	1	the	the	DET
ejpam-6087	109	2	models	model	NOUN
ejpam-6087	109	3	were	be	AUX
ejpam-6087	109	4	assessed	assess	VERB
ejpam-6087	109	5	based	base	VERB
ejpam-6087	109	6	on	on	ADP
ejpam-6087	109	7	precision	precision	NOUN
ejpam-6087	109	8	,	,	PUNCT
ejpam-6087	109	9	f1	f1	PROPN
ejpam-6087	109	10	m.	m.	PROPN
ejpam-6087	109	11	al	al	PROPN
ejpam-6087	109	12	-	-	PUNCT
ejpam-6087	109	13	tarawneh	tarawneh	PROPN
ejpam-6087	109	14	et	et	PROPN
ejpam-6087	109	15	al	al	PROPN
ejpam-6087	109	16	.	.	PUNCT
ejpam-6087	109	17	/	/	SYM
ejpam-6087	109	18	eur	eur	PROPN
ejpam-6087	109	19	.	.	PUNCT
ejpam-6087	110	1	j.	j.	PROPN
ejpam-6087	110	2	pure	pure	PROPN
ejpam-6087	110	3	appl	appl	PROPN
ejpam-6087	110	4	.	.	PROPN
ejpam-6087	110	5	math	math	PROPN
ejpam-6087	110	6	,	,	PUNCT
ejpam-6087	110	7	18	18	NUM
ejpam-6087	110	8	(	(	PUNCT
ejpam-6087	110	9	2	2	NUM
ejpam-6087	110	10	)	)	PUNCT
ejpam-6087	110	11	(	(	PUNCT
ejpam-6087	110	12	2025	2025	NUM
ejpam-6087	110	13	)	)	PUNCT
ejpam-6087	110	14	,	,	PUNCT
ejpam-6087	110	15	6087	6087	NUM
ejpam-6087	110	16	6	6	NUM
ejpam-6087	110	17	of	of	ADP
ejpam-6087	110	18	39	39	NUM
ejpam-6087	110	19	score	score	NOUN
ejpam-6087	110	20	,	,	PUNCT
ejpam-6087	110	21	accuracy	accuracy	NOUN
ejpam-6087	110	22	,	,	PUNCT
ejpam-6087	110	23	and	and	CCONJ
ejpam-6087	110	24	recall	recall	NOUN
ejpam-6087	110	25	,	,	PUNCT
ejpam-6087	110	26	with	with	ADP
ejpam-6087	110	27	the	the	DET
ejpam-6087	110	28	lstm	lstm	PROPN
ejpam-6087	110	29	model	model	NOUN
ejpam-6087	110	30	using	use	VERB
ejpam-6087	110	31	glove	glove	NOUN
ejpam-6087	110	32	embeddings	embedding	NOUN
ejpam-6087	110	33	achieving	achieve	VERB
ejpam-6087	110	34	the	the	DET
ejpam-6087	110	35	highest	high	ADJ
ejpam-6087	110	36	performance	performance	NOUN
ejpam-6087	110	37	.	.	PUNCT
ejpam-6087	111	1	however	however	ADV
ejpam-6087	111	2	,	,	PUNCT
ejpam-6087	111	3	the	the	DET
ejpam-6087	111	4	study	study	NOUN
ejpam-6087	111	5	did	do	AUX
ejpam-6087	111	6	not	not	PART
ejpam-6087	111	7	investigate	investigate	VERB
ejpam-6087	111	8	feature	feature	NOUN
ejpam-6087	111	9	selection	selection	NOUN
ejpam-6087	111	10	methods	method	NOUN
ejpam-6087	111	11	but	but	CCONJ
ejpam-6087	111	12	focused	focus	VERB
ejpam-6087	111	13	on	on	ADP
ejpam-6087	111	14	different	different	ADJ
ejpam-6087	111	15	embeddings	embedding	NOUN
ejpam-6087	111	16	and	and	CCONJ
ejpam-6087	111	17	algorithms	algorithm	NOUN
ejpam-6087	111	18	for	for	ADP
ejpam-6087	111	19	classification	classification	NOUN
ejpam-6087	111	20	.	.	PUNCT
ejpam-6087	112	1	in	in	ADP
ejpam-6087	112	2	addition	addition	NOUN
ejpam-6087	112	3	,	,	PUNCT
ejpam-6087	112	4	the	the	DET
ejpam-6087	112	5	research	research	NOUN
ejpam-6087	112	6	included	include	VERB
ejpam-6087	112	7	an	an	DET
ejpam-6087	112	8	analysis	analysis	NOUN
ejpam-6087	112	9	of	of	ADP
ejpam-6087	112	10	misclassified	misclassifie	VERB
ejpam-6087	112	11	samples	sample	NOUN
ejpam-6087	112	12	by	by	ADP
ejpam-6087	112	13	comparing	compare	VERB
ejpam-6087	112	14	them	they	PRON
ejpam-6087	112	15	with	with	ADP
ejpam-6087	112	16	human	human	ADJ
ejpam-6087	112	17	judgments	judgment	NOUN
ejpam-6087	112	18	.	.	PUNCT
ejpam-6087	113	1	the	the	DET
ejpam-6087	113	2	study	study	NOUN
ejpam-6087	113	3	in	in	ADP
ejpam-6087	113	4	[	[	X
ejpam-6087	113	5	24	24	NUM
ejpam-6087	113	6	]	]	PUNCT
ejpam-6087	113	7	addressed	address	VERB
ejpam-6087	113	8	the	the	DET
ejpam-6087	113	9	challenge	challenge	NOUN
ejpam-6087	113	10	of	of	ADP
ejpam-6087	113	11	fake	fake	ADJ
ejpam-6087	113	12	news	news	NOUN
ejpam-6087	113	13	detection	detection	NOUN
ejpam-6087	113	14	by	by	ADP
ejpam-6087	113	15	developing	develop	VERB
ejpam-6087	113	16	a	a	DET
ejpam-6087	113	17	model	model	NOUN
ejpam-6087	113	18	that	that	PRON
ejpam-6087	113	19	incorporates	incorporate	VERB
ejpam-6087	113	20	user	user	NOUN
ejpam-6087	113	21	characteristics	characteristic	NOUN
ejpam-6087	113	22	,	,	PUNCT
ejpam-6087	113	23	news	news	NOUN
ejpam-6087	113	24	content	content	NOUN
ejpam-6087	113	25	,	,	PUNCT
ejpam-6087	113	26	and	and	CCONJ
ejpam-6087	113	27	social	social	ADJ
ejpam-6087	113	28	network	network	NOUN
ejpam-6087	113	29	features	feature	NOUN
ejpam-6087	113	30	based	base	VERB
ejpam-6087	113	31	on	on	ADP
ejpam-6087	113	32	social	social	ADJ
ejpam-6087	113	33	capital	capital	NOUN
ejpam-6087	113	34	.	.	PUNCT
ejpam-6087	114	1	the	the	DET
ejpam-6087	114	2	xgboost	xgboost	PROPN
ejpam-6087	114	3	model	model	NOUN
ejpam-6087	114	4	was	be	AUX
ejpam-6087	114	5	used	use	VERB
ejpam-6087	114	6	to	to	PART
ejpam-6087	114	7	estimate	estimate	VERB
ejpam-6087	114	8	the	the	DET
ejpam-6087	114	9	importance	importance	NOUN
ejpam-6087	114	10	of	of	ADP
ejpam-6087	114	11	the	the	DET
ejpam-6087	114	12	features	feature	NOUN
ejpam-6087	114	13	and	and	CCONJ
ejpam-6087	114	14	identify	identify	VERB
ejpam-6087	114	15	key	key	ADJ
ejpam-6087	114	16	variables	variable	NOUN
ejpam-6087	114	17	for	for	ADP
ejpam-6087	114	18	the	the	DET
ejpam-6087	114	19	detection	detection	NOUN
ejpam-6087	114	20	of	of	ADP
ejpam-6087	114	21	fake	fake	ADJ
ejpam-6087	114	22	news	news	NOUN
ejpam-6087	114	23	.	.	PUNCT
ejpam-6087	115	1	several	several	ADJ
ejpam-6087	115	2	ml	ml	NOUN
ejpam-6087	115	3	algorithms	algorithm	NOUN
ejpam-6087	115	4	,	,	PUNCT
ejpam-6087	115	5	including	include	VERB
ejpam-6087	115	6	svm	svm	PROPN
ejpam-6087	115	7	,	,	PUNCT
ejpam-6087	115	8	rf	rf	ADJ
ejpam-6087	115	9	,	,	PUNCT
ejpam-6087	115	10	lr	lr	NOUN
ejpam-6087	115	11	,	,	PUNCT
ejpam-6087	115	12	classification	classification	NOUN
ejpam-6087	115	13	and	and	CCONJ
ejpam-6087	115	14	regression	regression	NOUN
ejpam-6087	115	15	trees	tree	NOUN
ejpam-6087	115	16	(	(	PUNCT
ejpam-6087	115	17	cart	cart	NOUN
ejpam-6087	115	18	)	)	PUNCT
ejpam-6087	115	19	and	and	CCONJ
ejpam-6087	115	20	neural	neural	ADJ
ejpam-6087	115	21	networks	network	NOUN
ejpam-6087	115	22	(	(	PUNCT
ejpam-6087	115	23	nn	nn	NOUN
ejpam-6087	115	24	)	)	PUNCT
ejpam-6087	115	25	,	,	PUNCT
ejpam-6087	115	26	were	be	AUX
ejpam-6087	115	27	applied	apply	VERB
ejpam-6087	115	28	to	to	PART
ejpam-6087	115	29	classify	classify	VERB
ejpam-6087	115	30	news	news	NOUN
ejpam-6087	115	31	articles	article	NOUN
ejpam-6087	115	32	.	.	PUNCT
ejpam-6087	116	1	a	a	DET
ejpam-6087	116	2	cross	cross	ADJ
ejpam-6087	116	3	-	-	ADJ
ejpam-6087	116	4	validation	validation	ADJ
ejpam-6087	116	5	step	step	NOUN
ejpam-6087	116	6	was	be	AUX
ejpam-6087	116	7	performed	perform	VERB
ejpam-6087	116	8	to	to	PART
ejpam-6087	116	9	generalize	generalize	VERB
ejpam-6087	116	10	the	the	DET
ejpam-6087	116	11	models	model	NOUN
ejpam-6087	116	12	and	and	CCONJ
ejpam-6087	116	13	prevent	prevent	VERB
ejpam-6087	116	14	overfitting	overfitting	NOUN
ejpam-6087	116	15	or	or	CCONJ
ejpam-6087	116	16	underfitting	underfitting	NOUN
ejpam-6087	116	17	.	.	PUNCT
ejpam-6087	117	1	the	the	DET
ejpam-6087	117	2	rf	rf	NOUN
ejpam-6087	117	3	model	model	NOUN
ejpam-6087	117	4	achieved	achieve	VERB
ejpam-6087	117	5	the	the	DET
ejpam-6087	117	6	highest	high	ADJ
ejpam-6087	117	7	prediction	prediction	NOUN
ejpam-6087	117	8	rate	rate	NOUN
ejpam-6087	117	9	of	of	ADP
ejpam-6087	117	10	approximately	approximately	ADV
ejpam-6087	117	11	94	94	NUM
ejpam-6087	117	12	%	%	NOUN
ejpam-6087	117	13	,	,	PUNCT
ejpam-6087	117	14	while	while	SCONJ
ejpam-6087	117	15	the	the	DET
ejpam-6087	117	16	nn	nn	PROPN
ejpam-6087	117	17	model	model	NOUN
ejpam-6087	117	18	had	have	VERB
ejpam-6087	117	19	the	the	DET
ejpam-6087	117	20	lowest	low	ADJ
ejpam-6087	117	21	performance	performance	NOUN
ejpam-6087	117	22	of	of	ADP
ejpam-6087	117	23	about	about	ADV
ejpam-6087	117	24	92.1	92.1	NUM
ejpam-6087	117	25	%	%	NOUN
ejpam-6087	117	26	.	.	PUNCT
ejpam-6087	118	1	the	the	DET
ejpam-6087	118	2	study	study	NOUN
ejpam-6087	118	3	did	do	AUX
ejpam-6087	118	4	not	not	PART
ejpam-6087	118	5	explore	explore	VERB
ejpam-6087	118	6	traditional	traditional	ADJ
ejpam-6087	118	7	feature	feature	NOUN
ejpam-6087	118	8	selection	selection	NOUN
ejpam-6087	118	9	methods	method	NOUN
ejpam-6087	118	10	but	but	CCONJ
ejpam-6087	118	11	focused	focus	VERB
ejpam-6087	118	12	on	on	ADP
ejpam-6087	118	13	derived	derived	ADJ
ejpam-6087	118	14	features	feature	NOUN
ejpam-6087	118	15	based	base	VERB
ejpam-6087	118	16	on	on	ADP
ejpam-6087	118	17	social	social	ADJ
ejpam-6087	118	18	capital	capital	NOUN
ejpam-6087	118	19	and	and	CCONJ
ejpam-6087	118	20	estimated	estimate	VERB
ejpam-6087	118	21	their	their	PRON
ejpam-6087	118	22	importance	importance	NOUN
ejpam-6087	118	23	to	to	PART
ejpam-6087	118	24	detect	detect	VERB
ejpam-6087	118	25	fake	fake	ADJ
ejpam-6087	118	26	news	news	NOUN
ejpam-6087	118	27	.	.	PUNCT
ejpam-6087	119	1	the	the	DET
ejpam-6087	119	2	work	work	NOUN
ejpam-6087	119	3	in	in	ADP
ejpam-6087	119	4	[	[	X
ejpam-6087	119	5	25	25	NUM
ejpam-6087	119	6	]	]	PUNCT
ejpam-6087	119	7	provided	provide	VERB
ejpam-6087	119	8	valuable	valuable	ADJ
ejpam-6087	119	9	information	information	NOUN
ejpam-6087	119	10	on	on	ADP
ejpam-6087	119	11	the	the	DET
ejpam-6087	119	12	impact	impact	NOUN
ejpam-6087	119	13	of	of	ADP
ejpam-6087	119	14	different	different	ADJ
ejpam-6087	119	15	word	word	NOUN
ejpam-6087	119	16	embedding	embed	VERB
ejpam-6087	119	17	techniques	technique	NOUN
ejpam-6087	119	18	on	on	ADP
ejpam-6087	119	19	the	the	DET
ejpam-6087	119	20	performance	performance	NOUN
ejpam-6087	119	21	of	of	ADP
ejpam-6087	119	22	machine	machine	NOUN
ejpam-6087	119	23	learning	learning	NOUN
ejpam-6087	119	24	and	and	CCONJ
ejpam-6087	119	25	deep	deep	ADJ
ejpam-6087	119	26	learning	learning	NOUN
ejpam-6087	119	27	models	model	NOUN
ejpam-6087	119	28	for	for	ADP
ejpam-6087	119	29	the	the	DET
ejpam-6087	119	30	detection	detection	NOUN
ejpam-6087	119	31	of	of	ADP
ejpam-6087	119	32	fake	fake	ADJ
ejpam-6087	119	33	news	news	NOUN
ejpam-6087	119	34	.	.	PUNCT
ejpam-6087	120	1	although	although	SCONJ
ejpam-6087	120	2	that	that	DET
ejpam-6087	120	3	study	study	NOUN
ejpam-6087	120	4	focused	focus	VERB
ejpam-6087	120	5	on	on	ADP
ejpam-6087	120	6	embedding	embed	VERB
ejpam-6087	120	7	-	-	PUNCT
ejpam-6087	120	8	based	base	VERB
ejpam-6087	120	9	feature	feature	NOUN
ejpam-6087	120	10	representations	representation	NOUN
ejpam-6087	120	11	and	and	CCONJ
ejpam-6087	120	12	compared	compare	VERB
ejpam-6087	120	13	individual	individual	ADJ
ejpam-6087	120	14	classifiers	classifier	NOUN
ejpam-6087	120	15	and	and	CCONJ
ejpam-6087	120	16	cnn	cnn	PROPN
ejpam-6087	120	17	-	-	PUNCT
ejpam-6087	120	18	based	base	VERB
ejpam-6087	120	19	architectures	architecture	NOUN
ejpam-6087	120	20	,	,	PUNCT
ejpam-6087	120	21	it	it	PRON
ejpam-6087	120	22	did	do	AUX
ejpam-6087	120	23	not	not	PART
ejpam-6087	120	24	consider	consider	VERB
ejpam-6087	120	25	ensemble	ensemble	ADJ
ejpam-6087	120	26	learning	learning	NOUN
ejpam-6087	120	27	techniques	technique	NOUN
ejpam-6087	120	28	or	or	CCONJ
ejpam-6087	120	29	feature	feature	NOUN
ejpam-6087	120	30	selection	selection	NOUN
ejpam-6087	120	31	strategies	strategy	NOUN
ejpam-6087	120	32	beyond	beyond	ADP
ejpam-6087	120	33	embeddings	embedding	NOUN
ejpam-6087	120	34	.	.	PUNCT
ejpam-6087	121	1	in	in	ADP
ejpam-6087	121	2	contrast	contrast	NOUN
ejpam-6087	121	3	,	,	PUNCT
ejpam-6087	121	4	the	the	DET
ejpam-6087	121	5	current	current	ADJ
ejpam-6087	121	6	study	study	NOUN
ejpam-6087	121	7	adopts	adopt	VERB
ejpam-6087	121	8	a	a	DET
ejpam-6087	121	9	different	different	ADJ
ejpam-6087	121	10	research	research	NOUN
ejpam-6087	121	11	direction	direction	NOUN
ejpam-6087	121	12	by	by	ADP
ejpam-6087	121	13	conducting	conduct	VERB
ejpam-6087	121	14	a	a	DET
ejpam-6087	121	15	comprehensive	comprehensive	ADJ
ejpam-6087	121	16	evaluation	evaluation	NOUN
ejpam-6087	121	17	of	of	ADP
ejpam-6087	121	18	traditional	traditional	ADJ
ejpam-6087	121	19	and	and	CCONJ
ejpam-6087	121	20	ensemble	ensemble	ADJ
ejpam-6087	121	21	classifiers	classifier	NOUN
ejpam-6087	121	22	in	in	ADP
ejpam-6087	121	23	combination	combination	NOUN
ejpam-6087	121	24	with	with	ADP
ejpam-6087	121	25	multiple	multiple	ADJ
ejpam-6087	121	26	feature	feature	NOUN
ejpam-6087	121	27	selection	selection	NOUN
ejpam-6087	121	28	techniques	technique	NOUN
ejpam-6087	121	29	,	,	PUNCT
ejpam-6087	121	30	including	include	VERB
ejpam-6087	121	31	recursive	recursive	ADJ
ejpam-6087	121	32	feature	feature	NOUN
ejpam-6087	121	33	elimination	elimination	NOUN
ejpam-6087	121	34	(	(	PUNCT
ejpam-6087	121	35	rfe	rfe	NOUN
ejpam-6087	121	36	)	)	PUNCT
ejpam-6087	121	37	,	,	PUNCT
ejpam-6087	121	38	selectkbest	selectkb	ADJ
ejpam-6087	121	39	,	,	PUNCT
ejpam-6087	121	40	principal	principal	ADJ
ejpam-6087	121	41	component	component	NOUN
ejpam-6087	121	42	analysis	analysis	NOUN
ejpam-6087	121	43	(	(	PUNCT
ejpam-6087	121	44	pca	pca	NOUN
ejpam-6087	121	45	)	)	PUNCT
ejpam-6087	121	46	,	,	PUNCT
ejpam-6087	121	47	and	and	CCONJ
ejpam-6087	121	48	genetic	genetic	ADJ
ejpam-6087	121	49	algorithms	algorithm	NOUN
ejpam-6087	121	50	(	(	PUNCT
ejpam-6087	121	51	ga	ga	NOUN
ejpam-6087	121	52	)	)	PUNCT
ejpam-6087	121	53	.	.	PUNCT
ejpam-6087	122	1	this	this	DET
ejpam-6087	122	2	approach	approach	NOUN
ejpam-6087	122	3	enables	enable	VERB
ejpam-6087	122	4	a	a	DET
ejpam-6087	122	5	deeper	deep	ADJ
ejpam-6087	122	6	investigation	investigation	NOUN
ejpam-6087	122	7	into	into	ADP
ejpam-6087	122	8	the	the	DET
ejpam-6087	122	9	relevance	relevance	NOUN
ejpam-6087	122	10	and	and	CCONJ
ejpam-6087	122	11	interpretability	interpretability	NOUN
ejpam-6087	122	12	of	of	ADP
ejpam-6087	122	13	features	feature	NOUN
ejpam-6087	122	14	and	and	CCONJ
ejpam-6087	122	15	offers	offer	VERB
ejpam-6087	122	16	comparative	comparative	ADJ
ejpam-6087	122	17	information	information	NOUN
ejpam-6087	122	18	on	on	ADP
ejpam-6087	122	19	how	how	SCONJ
ejpam-6087	122	20	various	various	ADJ
ejpam-6087	122	21	selection	selection	NOUN
ejpam-6087	122	22	methods	method	NOUN
ejpam-6087	122	23	influence	influence	VERB
ejpam-6087	122	24	classification	classification	NOUN
ejpam-6087	122	25	performance	performance	NOUN
ejpam-6087	122	26	,	,	PUNCT
ejpam-6087	122	27	areas	area	NOUN
ejpam-6087	122	28	not	not	PART
ejpam-6087	122	29	covered	cover	VERB
ejpam-6087	122	30	in	in	ADP
ejpam-6087	122	31	the	the	DET
ejpam-6087	122	32	previous	previous	ADJ
ejpam-6087	122	33	work	work	NOUN
ejpam-6087	122	34	.	.	PUNCT
ejpam-6087	123	1	in	in	ADP
ejpam-6087	123	2	contrast	contrast	NOUN
ejpam-6087	123	3	to	to	ADP
ejpam-6087	123	4	these	these	DET
ejpam-6087	123	5	studies	study	NOUN
ejpam-6087	123	6	,	,	PUNCT
ejpam-6087	123	7	this	this	DET
ejpam-6087	123	8	research	research	NOUN
ejpam-6087	123	9	presents	present	VERB
ejpam-6087	123	10	several	several	ADJ
ejpam-6087	123	11	distinctive	distinctive	ADJ
ejpam-6087	123	12	strengths	strength	NOUN
ejpam-6087	123	13	.	.	PUNCT
ejpam-6087	124	1	firstly	firstly	ADV
ejpam-6087	124	2	,	,	PUNCT
ejpam-6087	124	3	this	this	DET
ejpam-6087	124	4	study	study	NOUN
ejpam-6087	124	5	employs	employ	VERB
ejpam-6087	124	6	a	a	DET
ejpam-6087	124	7	diverse	diverse	ADJ
ejpam-6087	124	8	set	set	NOUN
ejpam-6087	124	9	of	of	ADP
ejpam-6087	124	10	individual	individual	ADJ
ejpam-6087	124	11	classifiers	classifier	NOUN
ejpam-6087	124	12	,	,	PUNCT
ejpam-6087	124	13	including	include	VERB
ejpam-6087	124	14	dt	dt	PROPN
ejpam-6087	124	15	,	,	PUNCT
ejpam-6087	124	16	nb	nb	PROPN
ejpam-6087	124	17	,	,	PUNCT
ejpam-6087	124	18	svm	svm	PROPN
ejpam-6087	124	19	,	,	PUNCT
ejpam-6087	124	20	knn	knn	PROPN
ejpam-6087	124	21	,	,	PUNCT
ejpam-6087	124	22	rf	rf	PROPN
ejpam-6087	124	23	,	,	PUNCT
ejpam-6087	124	24	and	and	CCONJ
ejpam-6087	124	25	mlp	mlp	NOUN
ejpam-6087	124	26	,	,	PUNCT
ejpam-6087	124	27	along	along	ADP
ejpam-6087	124	28	with	with	ADP
ejpam-6087	124	29	ensemble	ensemble	ADJ
ejpam-6087	124	30	methods	method	NOUN
ejpam-6087	124	31	such	such	ADJ
ejpam-6087	124	32	as	as	ADP
ejpam-6087	124	33	bagging	bagging	NOUN
ejpam-6087	124	34	,	,	PUNCT
ejpam-6087	124	35	boosting	boost	VERB
ejpam-6087	124	36	,	,	PUNCT
ejpam-6087	124	37	and	and	CCONJ
ejpam-6087	124	38	stacking	stacking	NOUN
ejpam-6087	124	39	.	.	PUNCT
ejpam-6087	125	1	this	this	DET
ejpam-6087	125	2	comprehensive	comprehensive	ADJ
ejpam-6087	125	3	approach	approach	NOUN
ejpam-6087	125	4	ensures	ensure	VERB
ejpam-6087	125	5	a	a	DET
ejpam-6087	125	6	robust	robust	ADJ
ejpam-6087	125	7	evaluation	evaluation	NOUN
ejpam-6087	125	8	of	of	ADP
ejpam-6087	125	9	various	various	ADJ
ejpam-6087	125	10	algorithms	algorithm	NOUN
ejpam-6087	125	11	for	for	ADP
ejpam-6087	125	12	the	the	DET
ejpam-6087	125	13	detection	detection	NOUN
ejpam-6087	125	14	of	of	ADP
ejpam-6087	125	15	fake	fake	ADJ
ejpam-6087	125	16	news	news	NOUN
ejpam-6087	125	17	.	.	PUNCT
ejpam-6087	126	1	in	in	ADP
ejpam-6087	126	2	addition	addition	NOUN
ejpam-6087	126	3	,	,	PUNCT
ejpam-6087	126	4	the	the	DET
ejpam-6087	126	5	research	research	NOUN
ejpam-6087	126	6	systematically	systematically	ADV
ejpam-6087	126	7	investigates	investigate	VERB
ejpam-6087	126	8	the	the	DET
ejpam-6087	126	9	impact	impact	NOUN
ejpam-6087	126	10	of	of	ADP
ejpam-6087	126	11	various	various	ADJ
ejpam-6087	126	12	feature	feature	NOUN
ejpam-6087	126	13	selection	selection	NOUN
ejpam-6087	126	14	methods	method	NOUN
ejpam-6087	126	15	,	,	PUNCT
ejpam-6087	126	16	including	include	VERB
ejpam-6087	126	17	rfe	rfe	NOUN
ejpam-6087	126	18	,	,	PUNCT
ejpam-6087	126	19	selectkbest	selectkb	ADJ
ejpam-6087	126	20	,	,	PUNCT
ejpam-6087	126	21	pca	pca	PROPN
ejpam-6087	126	22	,	,	PUNCT
ejpam-6087	126	23	and	and	CCONJ
ejpam-6087	126	24	ga	ga	PROPN
ejpam-6087	126	25	,	,	PUNCT
ejpam-6087	126	26	on	on	ADP
ejpam-6087	126	27	model	model	NOUN
ejpam-6087	126	28	performance	performance	NOUN
ejpam-6087	126	29	.	.	PUNCT
ejpam-6087	127	1	this	this	DET
ejpam-6087	127	2	detailed	detailed	ADJ
ejpam-6087	127	3	examination	examination	NOUN
ejpam-6087	127	4	of	of	ADP
ejpam-6087	127	5	feature	feature	NOUN
ejpam-6087	127	6	selection	selection	NOUN
ejpam-6087	127	7	techniques	technique	NOUN
ejpam-6087	127	8	provides	provide	VERB
ejpam-6087	127	9	valuable	valuable	ADJ
ejpam-6087	127	10	insight	insight	NOUN
ejpam-6087	127	11	into	into	ADP
ejpam-6087	127	12	how	how	SCONJ
ejpam-6087	127	13	different	different	ADJ
ejpam-6087	127	14	methods	method	NOUN
ejpam-6087	127	15	influence	influence	VERB
ejpam-6087	127	16	the	the	DET
ejpam-6087	127	17	effectiveness	effectiveness	NOUN
ejpam-6087	127	18	of	of	ADP
ejpam-6087	127	19	fake	fake	ADJ
ejpam-6087	127	20	news	news	NOUN
ejpam-6087	127	21	detection	detection	NOUN
ejpam-6087	127	22	models	model	NOUN
ejpam-6087	127	23	.	.	PUNCT
ejpam-6087	128	1	furthermore	furthermore	ADV
ejpam-6087	128	2	,	,	PUNCT
ejpam-6087	128	3	the	the	DET
ejpam-6087	128	4	study	study	NOUN
ejpam-6087	128	5	uses	use	VERB
ejpam-6087	128	6	the	the	DET
ejpam-6087	128	7	truthseeker	truthseeker	NOUN
ejpam-6087	128	8	dataset	dataset	NOUN
ejpam-6087	128	9	,	,	PUNCT
ejpam-6087	128	10	which	which	PRON
ejpam-6087	128	11	offers	offer	VERB
ejpam-6087	128	12	a	a	DET
ejpam-6087	128	13	rich	rich	ADJ
ejpam-6087	128	14	and	and	CCONJ
ejpam-6087	128	15	varied	varied	ADJ
ejpam-6087	128	16	source	source	NOUN
ejpam-6087	128	17	of	of	ADP
ejpam-6087	128	18	data	datum	NOUN
ejpam-6087	128	19	comprising	comprise	VERB
ejpam-6087	128	20	textual	textual	ADJ
ejpam-6087	128	21	,	,	PUNCT
ejpam-6087	128	22	lexical	lexical	ADJ
ejpam-6087	128	23	,	,	PUNCT
ejpam-6087	128	24	and	and	CCONJ
ejpam-6087	128	25	metadata	metadata	NOUN
ejpam-6087	128	26	features	feature	NOUN
ejpam-6087	128	27	.	.	PUNCT
ejpam-6087	129	1	spanning	span	VERB
ejpam-6087	129	2	more	more	ADJ
ejpam-6087	129	3	than	than	ADP
ejpam-6087	129	4	a	a	PRON
ejpam-6087	129	5	decade	decade	NOUN
ejpam-6087	129	6	,	,	PUNCT
ejpam-6087	129	7	this	this	DET
ejpam-6087	129	8	extensive	extensive	ADJ
ejpam-6087	129	9	dataset	dataset	NOUN
ejpam-6087	129	10	ensures	ensure	VERB
ejpam-6087	129	11	that	that	SCONJ
ejpam-6087	129	12	the	the	DET
ejpam-6087	129	13	models	model	NOUN
ejpam-6087	129	14	are	be	AUX
ejpam-6087	129	15	tested	test	VERB
ejpam-6087	129	16	on	on	ADP
ejpam-6087	129	17	a	a	DET
ejpam-6087	129	18	wide	wide	ADJ
ejpam-6087	129	19	range	range	NOUN
ejpam-6087	129	20	of	of	ADP
ejpam-6087	129	21	real	real	ADJ
ejpam-6087	129	22	-	-	PUNCT
ejpam-6087	129	23	world	world	NOUN
ejpam-6087	129	24	scenarios	scenario	NOUN
ejpam-6087	129	25	.	.	PUNCT
ejpam-6087	130	1	the	the	DET
ejpam-6087	130	2	inclusion	inclusion	NOUN
ejpam-6087	130	3	of	of	ADP
ejpam-6087	130	4	metadata	metadata	NOUN
ejpam-6087	130	5	features	feature	NOUN
ejpam-6087	130	6	,	,	PUNCT
ejpam-6087	130	7	such	such	ADJ
ejpam-6087	130	8	as	as	ADP
ejpam-6087	130	9	user	user	NOUN
ejpam-6087	130	10	behavior	behavior	NOUN
ejpam-6087	130	11	and	and	CCONJ
ejpam-6087	130	12	engagement	engagement	NOUN
ejpam-6087	130	13	metrics	metric	NOUN
ejpam-6087	130	14	,	,	PUNCT
ejpam-6087	130	15	allows	allow	VERB
ejpam-6087	130	16	for	for	ADP
ejpam-6087	130	17	a	a	DET
ejpam-6087	130	18	more	more	ADV
ejpam-6087	130	19	nuanced	nuanced	ADJ
ejpam-6087	130	20	analysis	analysis	NOUN
ejpam-6087	130	21	of	of	ADP
ejpam-6087	130	22	fake	fake	ADJ
ejpam-6087	130	23	news	news	NOUN
ejpam-6087	130	24	,	,	PUNCT
ejpam-6087	130	25	capturing	capture	VERB
ejpam-6087	130	26	patterns	pattern	NOUN
ejpam-6087	130	27	that	that	PRON
ejpam-6087	130	28	may	may	AUX
ejpam-6087	130	29	not	not	PART
ejpam-6087	130	30	be	be	AUX
ejpam-6087	130	31	evident	evident	ADJ
ejpam-6087	130	32	from	from	ADP
ejpam-6087	130	33	textual	textual	ADJ
ejpam-6087	130	34	analysis	analysis	NOUN
ejpam-6087	130	35	alone	alone	ADV
ejpam-6087	130	36	.	.	PUNCT
ejpam-6087	131	1	this	this	DET
ejpam-6087	131	2	holistic	holistic	ADJ
ejpam-6087	131	3	approach	approach	NOUN
ejpam-6087	131	4	improves	improve	VERB
ejpam-6087	131	5	the	the	DET
ejpam-6087	131	6	accuracy	accuracy	NOUN
ejpam-6087	131	7	and	and	CCONJ
ejpam-6087	131	8	robustness	robustness	NOUN
ejpam-6087	131	9	of	of	ADP
ejpam-6087	131	10	fake	fake	ADJ
ejpam-6087	131	11	news	news	NOUN
ejpam-6087	131	12	detection	detection	NOUN
ejpam-6087	131	13	models	model	NOUN
ejpam-6087	131	14	,	,	PUNCT
ejpam-6087	131	15	which	which	PRON
ejpam-6087	131	16	separates	separate	VERB
ejpam-6087	131	17	this	this	DET
ejpam-6087	131	18	study	study	NOUN
ejpam-6087	131	19	from	from	ADP
ejpam-6087	131	20	others	other	NOUN
ejpam-6087	131	21	that	that	PRON
ejpam-6087	131	22	focus	focus	VERB
ejpam-6087	131	23	primarily	primarily	ADV
ejpam-6087	131	24	on	on	ADP
ejpam-6087	131	25	advanced	advanced	ADJ
ejpam-6087	131	26	model	model	NOUN
ejpam-6087	131	27	architectures	architecture	NOUN
ejpam-6087	131	28	or	or	CCONJ
ejpam-6087	131	29	specific	specific	ADJ
ejpam-6087	131	30	feature	feature	NOUN
ejpam-6087	131	31	sets	set	NOUN
ejpam-6087	131	32	.	.	PUNCT
ejpam-6087	132	1	in	in	ADP
ejpam-6087	132	2	summary	summary	NOUN
ejpam-6087	132	3	,	,	PUNCT
ejpam-6087	132	4	m.	m.	PROPN
ejpam-6087	132	5	al	al	PROPN
ejpam-6087	132	6	-	-	PUNCT
ejpam-6087	132	7	tarawneh	tarawneh	PROPN
ejpam-6087	132	8	et	et	PROPN
ejpam-6087	132	9	al	al	PROPN
ejpam-6087	132	10	.	.	PUNCT
ejpam-6087	132	11	/	/	SYM
ejpam-6087	132	12	eur	eur	PROPN
ejpam-6087	132	13	.	.	PUNCT
ejpam-6087	133	1	j.	j.	PROPN
ejpam-6087	133	2	pure	pure	PROPN
ejpam-6087	133	3	appl	appl	PROPN
ejpam-6087	133	4	.	.	PROPN
ejpam-6087	133	5	math	math	PROPN
ejpam-6087	133	6	,	,	PUNCT
ejpam-6087	133	7	18	18	NUM
ejpam-6087	133	8	(	(	PUNCT
ejpam-6087	133	9	2	2	NUM
ejpam-6087	133	10	)	)	PUNCT
ejpam-6087	133	11	(	(	PUNCT
ejpam-6087	133	12	2025	2025	NUM
ejpam-6087	133	13	)	)	PUNCT
ejpam-6087	133	14	,	,	PUNCT
ejpam-6087	133	15	6087	6087	NUM
ejpam-6087	133	16	7	7	NUM
ejpam-6087	133	17	of	of	ADP
ejpam-6087	133	18	39	39	NUM
ejpam-6087	133	19	while	while	SCONJ
ejpam-6087	133	20	recent	recent	ADJ
ejpam-6087	133	21	studies	study	NOUN
ejpam-6087	133	22	have	have	AUX
ejpam-6087	133	23	made	make	VERB
ejpam-6087	133	24	significant	significant	ADJ
ejpam-6087	133	25	contributions	contribution	NOUN
ejpam-6087	133	26	to	to	ADP
ejpam-6087	133	27	the	the	DET
ejpam-6087	133	28	field	field	NOUN
ejpam-6087	133	29	of	of	ADP
ejpam-6087	133	30	fake	fake	ADJ
ejpam-6087	133	31	news	news	NOUN
ejpam-6087	133	32	detection	detection	NOUN
ejpam-6087	133	33	through	through	ADP
ejpam-6087	133	34	various	various	ADJ
ejpam-6087	133	35	ml	ml	NOUN
ejpam-6087	133	36	techniques	technique	NOUN
ejpam-6087	133	37	and	and	CCONJ
ejpam-6087	133	38	datasets	dataset	NOUN
ejpam-6087	133	39	,	,	PUNCT
ejpam-6087	133	40	current	current	ADJ
ejpam-6087	133	41	research	research	NOUN
ejpam-6087	133	42	distinguishes	distinguish	VERB
ejpam-6087	133	43	itself	itself	PRON
ejpam-6087	133	44	through	through	ADP
ejpam-6087	133	45	a	a	DET
ejpam-6087	133	46	comprehensive	comprehensive	ADJ
ejpam-6087	133	47	evaluation	evaluation	NOUN
ejpam-6087	133	48	of	of	ADP
ejpam-6087	133	49	multiple	multiple	ADJ
ejpam-6087	133	50	classifiers	classifier	NOUN
ejpam-6087	133	51	and	and	CCONJ
ejpam-6087	133	52	feature	feature	NOUN
ejpam-6087	133	53	selection	selection	NOUN
ejpam-6087	133	54	methods	method	NOUN
ejpam-6087	133	55	using	use	VERB
ejpam-6087	133	56	the	the	DET
ejpam-6087	133	57	diverse	diverse	ADJ
ejpam-6087	133	58	and	and	CCONJ
ejpam-6087	133	59	extensive	extensive	ADJ
ejpam-6087	133	60	truthseeker	truthseeker	NOUN
ejpam-6087	133	61	dataset	dataset	NOUN
ejpam-6087	133	62	.	.	PUNCT
ejpam-6087	134	1	this	this	DET
ejpam-6087	134	2	multifaceted	multifaceted	ADJ
ejpam-6087	134	3	approach	approach	NOUN
ejpam-6087	134	4	provides	provide	VERB
ejpam-6087	134	5	a	a	DET
ejpam-6087	134	6	more	more	ADV
ejpam-6087	134	7	detailed	detailed	ADJ
ejpam-6087	134	8	understanding	understanding	NOUN
ejpam-6087	134	9	of	of	ADP
ejpam-6087	134	10	effective	effective	ADJ
ejpam-6087	134	11	strategies	strategy	NOUN
ejpam-6087	134	12	for	for	ADP
ejpam-6087	134	13	detecting	detect	VERB
ejpam-6087	134	14	fake	fake	ADJ
ejpam-6087	134	15	news	news	NOUN
ejpam-6087	134	16	and	and	CCONJ
ejpam-6087	134	17	advancing	advance	VERB
ejpam-6087	134	18	the	the	DET
ejpam-6087	134	19	state	state	NOUN
ejpam-6087	134	20	of	of	ADP
ejpam-6087	134	21	the	the	DET
ejpam-6087	134	22	art	art	NOUN
ejpam-6087	134	23	in	in	ADP
ejpam-6087	134	24	this	this	DET
ejpam-6087	134	25	critical	critical	ADJ
ejpam-6087	134	26	area	area	NOUN
ejpam-6087	134	27	of	of	ADP
ejpam-6087	134	28	research	research	NOUN
ejpam-6087	134	29	.	.	PUNCT
ejpam-6087	135	1	3	3	X
ejpam-6087	135	2	.	.	X
ejpam-6087	135	3	methodology	methodology	NOUN
ejpam-6087	135	4	this	this	DET
ejpam-6087	135	5	section	section	NOUN
ejpam-6087	135	6	highlights	highlight	VERB
ejpam-6087	135	7	the	the	DET
ejpam-6087	135	8	stages	stage	NOUN
ejpam-6087	135	9	followed	follow	VERB
ejpam-6087	135	10	in	in	ADP
ejpam-6087	135	11	this	this	DET
ejpam-6087	135	12	investigation	investigation	NOUN
ejpam-6087	135	13	,	,	PUNCT
ejpam-6087	135	14	including	include	VERB
ejpam-6087	135	15	data	data	NOUN
ejpam-6087	135	16	collection	collection	NOUN
ejpam-6087	135	17	,	,	PUNCT
ejpam-6087	135	18	implementation	implementation	NOUN
ejpam-6087	135	19	of	of	ADP
ejpam-6087	135	20	machine	machine	NOUN
ejpam-6087	135	21	learning	learn	VERB
ejpam-6087	135	22	for	for	ADP
ejpam-6087	135	23	the	the	DET
ejpam-6087	135	24	detection	detection	NOUN
ejpam-6087	135	25	of	of	ADP
ejpam-6087	135	26	fake	fake	ADJ
ejpam-6087	135	27	news	news	NOUN
ejpam-6087	135	28	,	,	PUNCT
ejpam-6087	135	29	selection	selection	NOUN
ejpam-6087	135	30	of	of	ADP
ejpam-6087	135	31	features	feature	NOUN
ejpam-6087	135	32	and	and	CCONJ
ejpam-6087	135	33	performance	performance	NOUN
ejpam-6087	135	34	evaluation	evaluation	NOUN
ejpam-6087	135	35	.	.	PUNCT
ejpam-6087	136	1	3.1	3.1	NUM
ejpam-6087	136	2	.	.	PUNCT
ejpam-6087	136	3	data	datum	NOUN
ejpam-6087	136	4	collection	collection	NOUN
ejpam-6087	136	5	this	this	DET
ejpam-6087	136	6	study	study	NOUN
ejpam-6087	136	7	employs	employ	VERB
ejpam-6087	136	8	the	the	DET
ejpam-6087	136	9	truthseeker	truthseeker	NOUN
ejpam-6087	136	10	dataset	dataset	NOUN
ejpam-6087	136	11	,	,	PUNCT
ejpam-6087	136	12	one	one	NUM
ejpam-6087	136	13	of	of	ADP
ejpam-6087	136	14	the	the	DET
ejpam-6087	136	15	largest	large	ADJ
ejpam-6087	136	16	ground	ground	NOUN
ejpam-6087	136	17	-	-	PUNCT
ejpam-6087	136	18	truth	truth	NOUN
ejpam-6087	136	19	collections	collection	NOUN
ejpam-6087	136	20	for	for	ADP
ejpam-6087	136	21	fake	fake	ADJ
ejpam-6087	136	22	news	news	NOUN
ejpam-6087	136	23	detection	detection	NOUN
ejpam-6087	136	24	on	on	ADP
ejpam-6087	136	25	twitter	twitter	NOUN
ejpam-6087	136	26	,	,	PUNCT
ejpam-6087	136	27	which	which	PRON
ejpam-6087	136	28	spans	span	VERB
ejpam-6087	136	29	from	from	ADP
ejpam-6087	136	30	2009	2009	NUM
ejpam-6087	136	31	to	to	ADP
ejpam-6087	136	32	2022	2022	NUM
ejpam-6087	136	33	and	and	CCONJ
ejpam-6087	136	34	includes	include	VERB
ejpam-6087	136	35	more	more	ADJ
ejpam-6087	136	36	than	than	ADP
ejpam-6087	136	37	180,000	180,000	NUM
ejpam-6087	136	38	labeled	label	VERB
ejpam-6087	136	39	tweets	tweet	NOUN
ejpam-6087	136	40	.	.	PUNCT
ejpam-6087	137	1	the	the	DET
ejpam-6087	137	2	dataset	dataset	NOUN
ejpam-6087	137	3	creation	creation	NOUN
ejpam-6087	137	4	followed	follow	VERB
ejpam-6087	137	5	a	a	DET
ejpam-6087	137	6	structured	structured	ADJ
ejpam-6087	137	7	multi	multi	ADJ
ejpam-6087	137	8	-	-	ADJ
ejpam-6087	137	9	phase	phase	NOUN
ejpam-6087	137	10	annotation	annotation	NOUN
ejpam-6087	137	11	process	process	NOUN
ejpam-6087	137	12	designed	design	VERB
ejpam-6087	137	13	to	to	PART
ejpam-6087	137	14	ensure	ensure	VERB
ejpam-6087	137	15	high	high	ADJ
ejpam-6087	137	16	reliability	reliability	NOUN
ejpam-6087	137	17	.	.	PUNCT
ejpam-6087	138	1	initial	initial	ADJ
ejpam-6087	138	2	real	real	ADJ
ejpam-6087	138	3	and	and	CCONJ
ejpam-6087	138	4	fake	fake	ADJ
ejpam-6087	138	5	statements	statement	NOUN
ejpam-6087	138	6	were	be	AUX
ejpam-6087	138	7	collected	collect	VERB
ejpam-6087	138	8	from	from	ADP
ejpam-6087	138	9	politifact	politifact	NOUN
ejpam-6087	138	10	and	and	CCONJ
ejpam-6087	138	11	keyword	keyword	NOUN
ejpam-6087	138	12	-	-	PUNCT
ejpam-6087	138	13	based	base	VERB
ejpam-6087	138	14	crawling	crawling	NOUN
ejpam-6087	138	15	was	be	AUX
ejpam-6087	138	16	used	use	VERB
ejpam-6087	138	17	to	to	PART
ejpam-6087	138	18	retrieve	retrieve	VERB
ejpam-6087	138	19	related	related	ADJ
ejpam-6087	138	20	tweets	tweet	NOUN
ejpam-6087	138	21	.	.	PUNCT
ejpam-6087	139	1	the	the	DET
ejpam-6087	139	2	annotation	annotation	NOUN
ejpam-6087	139	3	process	process	NOUN
ejpam-6087	139	4	involved	involve	VERB
ejpam-6087	139	5	amazon	amazon	PROPN
ejpam-6087	139	6	mechanical	mechanical	ADJ
ejpam-6087	139	7	turk	turk	PROPN
ejpam-6087	139	8	(	(	PUNCT
ejpam-6087	139	9	mturk	mturk	PROPN
ejpam-6087	139	10	)	)	PUNCT
ejpam-6087	139	11	,	,	PUNCT
ejpam-6087	139	12	where	where	SCONJ
ejpam-6087	139	13	each	each	DET
ejpam-6087	139	14	tweet	tweet	NOUN
ejpam-6087	139	15	-	-	PUNCT
ejpam-6087	139	16	statement	statement	NOUN
ejpam-6087	139	17	pair	pair	NOUN
ejpam-6087	139	18	was	be	AUX
ejpam-6087	139	19	evaluated	evaluate	VERB
ejpam-6087	139	20	by	by	ADP
ejpam-6087	139	21	three	three	NUM
ejpam-6087	139	22	independent	independent	ADJ
ejpam-6087	139	23	master	master	NOUN
ejpam-6087	139	24	turkers	turker	NOUN
ejpam-6087	139	25	.	.	PUNCT
ejpam-6087	140	1	these	these	DET
ejpam-6087	140	2	highly	highly	ADV
ejpam-6087	140	3	qualified	qualified	ADJ
ejpam-6087	140	4	annotators	annotator	NOUN
ejpam-6087	140	5	completed	complete	VERB
ejpam-6087	140	6	semantic	semantic	ADJ
ejpam-6087	140	7	similarity	similarity	NOUN
ejpam-6087	140	8	tasks	task	NOUN
ejpam-6087	140	9	,	,	PUNCT
ejpam-6087	140	10	judging	judge	VERB
ejpam-6087	140	11	how	how	SCONJ
ejpam-6087	140	12	much	much	ADJ
ejpam-6087	140	13	a	a	DET
ejpam-6087	140	14	tweet	tweet	NOUN
ejpam-6087	140	15	agreed	agree	VERB
ejpam-6087	140	16	or	or	CCONJ
ejpam-6087	140	17	disagreed	disagree	VERB
ejpam-6087	140	18	with	with	ADP
ejpam-6087	140	19	the	the	DET
ejpam-6087	140	20	source	source	NOUN
ejpam-6087	140	21	statement	statement	NOUN
ejpam-6087	140	22	using	use	VERB
ejpam-6087	140	23	predefined	predefine	VERB
ejpam-6087	140	24	five	five	NUM
ejpam-6087	140	25	-	-	PUNCT
ejpam-6087	140	26	class	class	NOUN
ejpam-6087	140	27	and	and	CCONJ
ejpam-6087	140	28	three	three	NUM
ejpam-6087	140	29	-	-	PUNCT
ejpam-6087	140	30	class	class	NOUN
ejpam-6087	140	31	labeling	labeling	NOUN
ejpam-6087	140	32	schemes	scheme	NOUN
ejpam-6087	140	33	.	.	PUNCT
ejpam-6087	141	1	only	only	ADV
ejpam-6087	141	2	responses	response	NOUN
ejpam-6087	141	3	with	with	ADP
ejpam-6087	141	4	a	a	DET
ejpam-6087	141	5	two	two	NUM
ejpam-6087	141	6	-	-	PUNCT
ejpam-6087	141	7	thirds	third	NOUN
ejpam-6087	141	8	majority	majority	NOUN
ejpam-6087	141	9	agreement	agreement	NOUN
ejpam-6087	141	10	were	be	AUX
ejpam-6087	141	11	retained	retain	VERB
ejpam-6087	141	12	to	to	PART
ejpam-6087	141	13	ensure	ensure	VERB
ejpam-6087	141	14	label	label	NOUN
ejpam-6087	141	15	consistency.this	consistency.this	DET
ejpam-6087	141	16	robust	robust	ADJ
ejpam-6087	141	17	process	process	NOUN
ejpam-6087	141	18	,	,	PUNCT
ejpam-6087	141	19	grounded	ground	VERB
ejpam-6087	141	20	in	in	ADP
ejpam-6087	141	21	active	active	ADJ
ejpam-6087	141	22	learning	learning	NOUN
ejpam-6087	141	23	and	and	CCONJ
ejpam-6087	141	24	human	human	ADJ
ejpam-6087	141	25	verification	verification	NOUN
ejpam-6087	141	26	,	,	PUNCT
ejpam-6087	141	27	makes	make	VERB
ejpam-6087	141	28	truthseeker	truthseeker	NOUN
ejpam-6087	141	29	a	a	DET
ejpam-6087	141	30	reliable	reliable	ADJ
ejpam-6087	141	31	resource	resource	NOUN
ejpam-6087	141	32	for	for	ADP
ejpam-6087	141	33	both	both	CCONJ
ejpam-6087	141	34	binary	binary	ADJ
ejpam-6087	141	35	and	and	CCONJ
ejpam-6087	141	36	multiclass	multiclass	ADJ
ejpam-6087	141	37	fake	fake	ADJ
ejpam-6087	141	38	news	news	NOUN
ejpam-6087	141	39	detection	detection	NOUN
ejpam-6087	141	40	.	.	PUNCT
ejpam-6087	142	1	in	in	ADP
ejpam-6087	142	2	addition	addition	NOUN
ejpam-6087	142	3	to	to	ADP
ejpam-6087	142	4	labels	label	NOUN
ejpam-6087	142	5	,	,	PUNCT
ejpam-6087	142	6	the	the	DET
ejpam-6087	142	7	truthseeker	truthseeker	NOUN
ejpam-6087	142	8	dataset	dataset	VERB
ejpam-6087	142	9	offers	offer	VERB
ejpam-6087	142	10	enriched	enriched	ADJ
ejpam-6087	142	11	information	information	NOUN
ejpam-6087	142	12	to	to	PART
ejpam-6087	142	13	enhance	enhance	VERB
ejpam-6087	142	14	its	its	PRON
ejpam-6087	142	15	analytical	analytical	ADJ
ejpam-6087	142	16	capabilities	capability	NOUN
ejpam-6087	142	17	.	.	PUNCT
ejpam-6087	143	1	each	each	DET
ejpam-6087	143	2	tweet	tweet	NOUN
ejpam-6087	143	3	is	be	AUX
ejpam-6087	143	4	supplemented	supplement	VERB
ejpam-6087	143	5	with	with	ADP
ejpam-6087	143	6	scores	score	NOUN
ejpam-6087	143	7	that	that	PRON
ejpam-6087	143	8	assess	assess	VERB
ejpam-6087	143	9	bot	bot	NOUN
ejpam-6087	143	10	likelihood	likelihood	NOUN
ejpam-6087	143	11	,	,	PUNCT
ejpam-6087	143	12	user	user	NOUN
ejpam-6087	143	13	credibility	credibility	NOUN
ejpam-6087	143	14	,	,	PUNCT
ejpam-6087	143	15	and	and	CCONJ
ejpam-6087	143	16	influence	influence	NOUN
ejpam-6087	143	17	,	,	PUNCT
ejpam-6087	143	18	providing	provide	VERB
ejpam-6087	143	19	a	a	DET
ejpam-6087	143	20	comprehensive	comprehensive	ADJ
ejpam-6087	143	21	view	view	NOUN
ejpam-6087	143	22	of	of	ADP
ejpam-6087	143	23	the	the	DET
ejpam-6087	143	24	source	source	NOUN
ejpam-6087	143	25	and	and	CCONJ
ejpam-6087	143	26	its	its	PRON
ejpam-6087	143	27	potential	potential	ADJ
ejpam-6087	143	28	impact	impact	NOUN
ejpam-6087	143	29	.	.	PUNCT
ejpam-6087	144	1	these	these	DET
ejpam-6087	144	2	scores	score	NOUN
ejpam-6087	144	3	are	be	AUX
ejpam-6087	144	4	critical	critical	ADJ
ejpam-6087	144	5	to	to	ADP
ejpam-6087	144	6	understanding	understand	VERB
ejpam-6087	144	7	the	the	DET
ejpam-6087	144	8	mechanisms	mechanism	NOUN
ejpam-6087	144	9	of	of	ADP
ejpam-6087	144	10	automated	automate	VERB
ejpam-6087	144	11	misinformation	misinformation	NOUN
ejpam-6087	144	12	,	,	PUNCT
ejpam-6087	144	13	assessing	assess	VERB
ejpam-6087	144	14	content	content	ADJ
ejpam-6087	144	15	credibility	credibility	NOUN
ejpam-6087	144	16	,	,	PUNCT
ejpam-6087	144	17	and	and	CCONJ
ejpam-6087	144	18	studying	study	VERB
ejpam-6087	144	19	the	the	DET
ejpam-6087	144	20	role	role	NOUN
ejpam-6087	144	21	of	of	ADP
ejpam-6087	144	22	influential	influential	ADJ
ejpam-6087	144	23	users	user	NOUN
ejpam-6087	144	24	in	in	ADP
ejpam-6087	144	25	spreading	spread	VERB
ejpam-6087	144	26	false	false	ADJ
ejpam-6087	144	27	information	information	NOUN
ejpam-6087	144	28	.	.	PUNCT
ejpam-6087	145	1	furthermore	furthermore	ADV
ejpam-6087	145	2	,	,	PUNCT
ejpam-6087	145	3	the	the	DET
ejpam-6087	145	4	dataset	dataset	NOUN
ejpam-6087	145	5	includes	include	VERB
ejpam-6087	145	6	rich	rich	ADJ
ejpam-6087	145	7	metadata	metadata	NOUN
ejpam-6087	145	8	,	,	PUNCT
ejpam-6087	145	9	such	such	ADJ
ejpam-6087	145	10	as	as	ADP
ejpam-6087	145	11	user	user	NOUN
ejpam-6087	145	12	profiles	profile	NOUN
ejpam-6087	145	13	,	,	PUNCT
ejpam-6087	145	14	tweet	tweet	NOUN
ejpam-6087	145	15	timestamps	timestamp	NOUN
ejpam-6087	145	16	,	,	PUNCT
ejpam-6087	145	17	and	and	CCONJ
ejpam-6087	145	18	engagement	engagement	NOUN
ejpam-6087	145	19	metrics	metric	NOUN
ejpam-6087	145	20	(	(	PUNCT
ejpam-6087	145	21	likes	like	NOUN
ejpam-6087	145	22	,	,	PUNCT
ejpam-6087	145	23	retweets	retweet	NOUN
ejpam-6087	145	24	,	,	PUNCT
ejpam-6087	145	25	replies	reply	NOUN
ejpam-6087	145	26	)	)	PUNCT
ejpam-6087	145	27	,	,	PUNCT
ejpam-6087	145	28	offering	offer	VERB
ejpam-6087	145	29	invaluable	invaluable	ADJ
ejpam-6087	145	30	insights	insight	NOUN
ejpam-6087	145	31	into	into	ADP
ejpam-6087	145	32	the	the	DET
ejpam-6087	145	33	spread	spread	NOUN
ejpam-6087	145	34	and	and	CCONJ
ejpam-6087	145	35	impact	impact	NOUN
ejpam-6087	145	36	of	of	ADP
ejpam-6087	145	37	fake	fake	ADJ
ejpam-6087	145	38	news	news	NOUN
ejpam-6087	145	39	across	across	ADP
ejpam-6087	145	40	networks	network	NOUN
ejpam-6087	145	41	.	.	PUNCT
ejpam-6087	146	1	the	the	DET
ejpam-6087	146	2	truthseeker	truthseeker	NOUN
ejpam-6087	146	3	dataset	dataset	NOUN
ejpam-6087	146	4	is	be	AUX
ejpam-6087	146	5	made	make	VERB
ejpam-6087	146	6	available	available	ADJ
ejpam-6087	146	7	through	through	ADP
ejpam-6087	146	8	the	the	DET
ejpam-6087	146	9	canadian	canadian	PROPN
ejpam-6087	146	10	institute	institute	PROPN
ejpam-6087	146	11	for	for	ADP
ejpam-6087	146	12	cybersecurity	cybersecurity	NOUN
ejpam-6087	146	13	(	(	PUNCT
ejpam-6087	146	14	cic	cic	PROPN
ejpam-6087	146	15	)	)	PUNCT
ejpam-6087	146	16	as	as	ADP
ejpam-6087	146	17	an	an	DET
ejpam-6087	146	18	open	open	ADJ
ejpam-6087	146	19	resource	resource	NOUN
ejpam-6087	146	20	for	for	ADP
ejpam-6087	146	21	researchers	researcher	NOUN
ejpam-6087	146	22	worldwide	worldwide	ADV
ejpam-6087	146	23	.	.	PUNCT
ejpam-6087	147	1	its	its	PRON
ejpam-6087	147	2	high	high	ADJ
ejpam-6087	147	3	validation	validation	NOUN
ejpam-6087	147	4	standards	standard	NOUN
ejpam-6087	147	5	and	and	CCONJ
ejpam-6087	147	6	extensive	extensive	ADJ
ejpam-6087	147	7	metadata	metadata	NOUN
ejpam-6087	147	8	make	make	VERB
ejpam-6087	147	9	it	it	PRON
ejpam-6087	147	10	indispensable	indispensable	ADJ
ejpam-6087	147	11	for	for	ADP
ejpam-6087	147	12	advancing	advance	VERB
ejpam-6087	147	13	fake	fake	ADJ
ejpam-6087	147	14	news	news	NOUN
ejpam-6087	147	15	detection	detection	NOUN
ejpam-6087	147	16	methodologies	methodology	NOUN
ejpam-6087	147	17	.	.	PUNCT
ejpam-6087	148	1	researchers	researcher	NOUN
ejpam-6087	148	2	can	can	AUX
ejpam-6087	148	3	use	use	VERB
ejpam-6087	148	4	it	it	PRON
ejpam-6087	148	5	to	to	PART
ejpam-6087	148	6	develop	develop	VERB
ejpam-6087	148	7	and	and	CCONJ
ejpam-6087	148	8	test	test	VERB
ejpam-6087	148	9	new	new	ADJ
ejpam-6087	148	10	machine	machine	NOUN
ejpam-6087	148	11	learning	learning	NOUN
ejpam-6087	148	12	models	model	NOUN
ejpam-6087	148	13	,	,	PUNCT
ejpam-6087	148	14	explore	explore	VERB
ejpam-6087	148	15	patterns	pattern	NOUN
ejpam-6087	148	16	in	in	ADP
ejpam-6087	148	17	misinformation	misinformation	NOUN
ejpam-6087	148	18	dissemination	dissemination	NOUN
ejpam-6087	148	19	,	,	PUNCT
ejpam-6087	148	20	and	and	CCONJ
ejpam-6087	148	21	design	design	NOUN
ejpam-6087	148	22	interventions	intervention	NOUN
ejpam-6087	148	23	to	to	PART
ejpam-6087	148	24	mitigate	mitigate	VERB
ejpam-6087	148	25	its	its	PRON
ejpam-6087	148	26	spread	spread	NOUN
ejpam-6087	148	27	.	.	PUNCT
ejpam-6087	149	1	in	in	ADP
ejpam-6087	149	2	general	general	ADJ
ejpam-6087	149	3	,	,	PUNCT
ejpam-6087	149	4	the	the	DET
ejpam-6087	149	5	truthseeker	truthseeker	NOUN
ejpam-6087	149	6	dataset	dataset	VERB
ejpam-6087	149	7	is	be	AUX
ejpam-6087	149	8	a	a	DET
ejpam-6087	149	9	meticulously	meticulously	ADV
ejpam-6087	149	10	validated	validate	VERB
ejpam-6087	149	11	and	and	CCONJ
ejpam-6087	149	12	richly	richly	ADV
ejpam-6087	149	13	annotated	annotate	VERB
ejpam-6087	149	14	resource	resource	NOUN
ejpam-6087	149	15	that	that	PRON
ejpam-6087	149	16	provides	provide	VERB
ejpam-6087	149	17	a	a	DET
ejpam-6087	149	18	solid	solid	ADJ
ejpam-6087	149	19	foundation	foundation	NOUN
ejpam-6087	149	20	to	to	PART
ejpam-6087	149	21	tackle	tackle	VERB
ejpam-6087	149	22	the	the	DET
ejpam-6087	149	23	complex	complex	ADJ
ejpam-6087	149	24	issue	issue	NOUN
ejpam-6087	149	25	of	of	ADP
ejpam-6087	149	26	misinformation	misinformation	NOUN
ejpam-6087	149	27	on	on	ADP
ejpam-6087	149	28	social	social	ADJ
ejpam-6087	149	29	networks	network	NOUN
ejpam-6087	149	30	.	.	PUNCT
ejpam-6087	150	1	tables	table	NOUN
ejpam-6087	150	2	1	1	NUM
ejpam-6087	150	3	,	,	PUNCT
ejpam-6087	150	4	2	2	NUM
ejpam-6087	150	5	,	,	PUNCT
ejpam-6087	150	6	and	and	CCONJ
ejpam-6087	150	7	3	3	NUM
ejpam-6087	150	8	summarize	summarize	VERB
ejpam-6087	150	9	the	the	DET
ejpam-6087	150	10	text	text	NOUN
ejpam-6087	150	11	,	,	PUNCT
ejpam-6087	150	12	lexical	lexical	ADJ
ejpam-6087	150	13	,	,	PUNCT
ejpam-6087	150	14	and	and	CCONJ
ejpam-6087	150	15	metadata	metadata	VERB
ejpam-6087	150	16	features	feature	NOUN
ejpam-6087	150	17	of	of	ADP
ejpam-6087	150	18	m.	m.	PROPN
ejpam-6087	150	19	al	al	PROPN
ejpam-6087	150	20	-	-	PUNCT
ejpam-6087	150	21	tarawneh	tarawneh	PROPN
ejpam-6087	150	22	et	et	PROPN
ejpam-6087	150	23	al	al	PROPN
ejpam-6087	150	24	.	.	PUNCT
ejpam-6087	150	25	/	/	SYM
ejpam-6087	150	26	eur	eur	PROPN
ejpam-6087	150	27	.	.	PUNCT
ejpam-6087	151	1	j.	j.	PROPN
ejpam-6087	151	2	pure	pure	PROPN
ejpam-6087	151	3	appl	appl	PROPN
ejpam-6087	151	4	.	.	PROPN
ejpam-6087	151	5	math	math	PROPN
ejpam-6087	151	6	,	,	PUNCT
ejpam-6087	151	7	18	18	NUM
ejpam-6087	151	8	(	(	PUNCT
ejpam-6087	151	9	2	2	NUM
ejpam-6087	151	10	)	)	PUNCT
ejpam-6087	151	11	(	(	PUNCT
ejpam-6087	151	12	2025	2025	NUM
ejpam-6087	151	13	)	)	PUNCT
ejpam-6087	151	14	,	,	PUNCT
ejpam-6087	151	15	6087	6087	NUM
ejpam-6087	151	16	8	8	NUM
ejpam-6087	151	17	of	of	ADP
ejpam-6087	151	18	39	39	NUM
ejpam-6087	151	19	table	table	NOUN
ejpam-6087	151	20	1	1	NUM
ejpam-6087	151	21	:	:	PUNCT
ejpam-6087	151	22	text	text	NOUN
ejpam-6087	151	23	features	feature	VERB
ejpam-6087	151	24	for	for	ADP
ejpam-6087	151	25	fake	fake	ADJ
ejpam-6087	151	26	news	news	NOUN
ejpam-6087	151	27	detection	detection	NOUN
ejpam-6087	151	28	name	name	NOUN
ejpam-6087	151	29	description	description	NOUN
ejpam-6087	151	30	unique	unique	ADJ
ejpam-6087	151	31	count	count	NOUN
ejpam-6087	151	32	total	total	ADJ
ejpam-6087	151	33	count	count	NOUN
ejpam-6087	151	34	of	of	ADP
ejpam-6087	151	35	unique	unique	ADJ
ejpam-6087	151	36	,	,	PUNCT
ejpam-6087	151	37	complex	complex	ADJ
ejpam-6087	151	38	terms	term	NOUN
ejpam-6087	151	39	in	in	ADP
ejpam-6087	151	40	text	text	NOUN
ejpam-6087	151	41	total	total	ADJ
ejpam-6087	151	42	count	count	NOUN
ejpam-6087	151	43	overall	overall	ADJ
ejpam-6087	151	44	word	word	NOUN
ejpam-6087	151	45	count	count	NOUN
ejpam-6087	151	46	in	in	ADP
ejpam-6087	151	47	the	the	DET
ejpam-6087	151	48	text	text	NOUN
ejpam-6087	151	49	org	org	ADJ
ejpam-6087	151	50	percent	percent	NOUN
ejpam-6087	151	51	percentage	percentage	NOUN
ejpam-6087	151	52	of	of	ADP
ejpam-6087	151	53	text	text	NOUN
ejpam-6087	151	54	referencing	referencing	NOUN
ejpam-6087	151	55	organizations	organization	NOUN
ejpam-6087	151	56	norp	norp	VERB
ejpam-6087	151	57	percent	percent	NOUN
ejpam-6087	151	58	percentage	percentage	NOUN
ejpam-6087	151	59	of	of	ADP
ejpam-6087	151	60	text	text	NOUN
ejpam-6087	151	61	mentioning	mentioning	NOUN
ejpam-6087	151	62	groups	group	NOUN
ejpam-6087	151	63	or	or	CCONJ
ejpam-6087	151	64	affiliations	affiliation	NOUN
ejpam-6087	151	65	gpe	gpe	NOUN
ejpam-6087	151	66	percent	percent	NOUN
ejpam-6087	151	67	percentage	percentage	NOUN
ejpam-6087	151	68	of	of	ADP
ejpam-6087	151	69	text	text	NOUN
ejpam-6087	151	70	related	relate	VERB
ejpam-6087	151	71	to	to	ADP
ejpam-6087	151	72	geopolitical	geopolitical	ADJ
ejpam-6087	151	73	entities	entity	NOUN
ejpam-6087	151	74	person	person	NOUN
ejpam-6087	151	75	percent	percent	NOUN
ejpam-6087	151	76	percentage	percentage	NOUN
ejpam-6087	151	77	of	of	ADP
ejpam-6087	151	78	text	text	NOUN
ejpam-6087	151	79	including	include	VERB
ejpam-6087	151	80	references	reference	NOUN
ejpam-6087	151	81	to	to	ADP
ejpam-6087	151	82	individuals	individual	NOUN
ejpam-6087	151	83	money	money	NOUN
ejpam-6087	151	84	percent	percent	NOUN
ejpam-6087	151	85	percentage	percentage	NOUN
ejpam-6087	151	86	of	of	ADP
ejpam-6087	151	87	text	text	NOUN
ejpam-6087	151	88	discussing	discuss	VERB
ejpam-6087	151	89	monetary	monetary	ADJ
ejpam-6087	151	90	amounts	amount	NOUN
ejpam-6087	151	91	data	data	NOUN
ejpam-6087	151	92	percent	percent	NOUN
ejpam-6087	151	93	percentage	percentage	NOUN
ejpam-6087	151	94	of	of	ADP
ejpam-6087	151	95	text	text	NOUN
ejpam-6087	151	96	containing	contain	VERB
ejpam-6087	151	97	references	reference	NOUN
ejpam-6087	151	98	to	to	ADP
ejpam-6087	151	99	dates	date	VERB
ejpam-6087	151	100	cardinal	cardinal	ADJ
ejpam-6087	151	101	percent	percent	NOUN
ejpam-6087	151	102	percentage	percentage	NOUN
ejpam-6087	151	103	of	of	ADP
ejpam-6087	151	104	text	text	NOUN
ejpam-6087	151	105	using	use	VERB
ejpam-6087	151	106	cardinal	cardinal	ADJ
ejpam-6087	151	107	numbers	number	NOUN
ejpam-6087	151	108	percent	percent	NOUN
ejpam-6087	151	109	percent	percent	NOUN
ejpam-6087	151	110	percentage	percentage	NOUN
ejpam-6087	151	111	of	of	ADP
ejpam-6087	151	112	text	text	NOUN
ejpam-6087	151	113	involving	involve	VERB
ejpam-6087	151	114	percentages	percentage	NOUN
ejpam-6087	151	115	ordinal	ordinal	ADJ
ejpam-6087	151	116	percent	percent	NOUN
ejpam-6087	151	117	percentage	percentage	NOUN
ejpam-6087	151	118	of	of	ADP
ejpam-6087	151	119	text	text	NOUN
ejpam-6087	151	120	using	use	VERB
ejpam-6087	151	121	ordinal	ordinal	ADJ
ejpam-6087	151	122	numbers	number	NOUN
ejpam-6087	151	123	law	law	NOUN
ejpam-6087	151	124	percent	percent	NOUN
ejpam-6087	151	125	percentage	percentage	NOUN
ejpam-6087	151	126	of	of	ADP
ejpam-6087	151	127	text	text	NOUN
ejpam-6087	151	128	mentioning	mention	VERB
ejpam-6087	151	129	legal	legal	ADJ
ejpam-6087	151	130	documents	document	NOUN
ejpam-6087	151	131	product	product	NOUN
ejpam-6087	151	132	percent	percent	NOUN
ejpam-6087	151	133	percentage	percentage	NOUN
ejpam-6087	151	134	of	of	ADP
ejpam-6087	151	135	text	text	NOUN
ejpam-6087	151	136	referring	refer	VERB
ejpam-6087	151	137	to	to	ADP
ejpam-6087	151	138	products	product	NOUN
ejpam-6087	151	139	event	event	NOUN
ejpam-6087	151	140	percent	percent	NOUN
ejpam-6087	151	141	percentage	percentage	NOUN
ejpam-6087	151	142	of	of	ADP
ejpam-6087	151	143	text	text	NOUN
ejpam-6087	151	144	mentioning	mention	VERB
ejpam-6087	151	145	specific	specific	ADJ
ejpam-6087	151	146	events	event	NOUN
ejpam-6087	151	147	time	time	NOUN
ejpam-6087	151	148	percent	percent	NOUN
ejpam-6087	151	149	percentage	percentage	NOUN
ejpam-6087	151	150	of	of	ADP
ejpam-6087	151	151	text	text	NOUN
ejpam-6087	151	152	referring	refer	VERB
ejpam-6087	151	153	to	to	ADP
ejpam-6087	151	154	time	time	NOUN
ejpam-6087	151	155	-	-	PUNCT
ejpam-6087	151	156	related	relate	VERB
ejpam-6087	151	157	details	detail	NOUN
ejpam-6087	151	158	loc	loc	NOUN
ejpam-6087	151	159	percent	percent	NOUN
ejpam-6087	151	160	percentage	percentage	NOUN
ejpam-6087	151	161	of	of	ADP
ejpam-6087	151	162	text	text	NOUN
ejpam-6087	151	163	with	with	ADP
ejpam-6087	151	164	references	reference	NOUN
ejpam-6087	151	165	to	to	ADP
ejpam-6087	151	166	locations	location	NOUN
ejpam-6087	151	167	work	work	NOUN
ejpam-6087	151	168	of	of	ADP
ejpam-6087	151	169	art	art	NOUN
ejpam-6087	151	170	percent	percent	NOUN
ejpam-6087	151	171	percentage	percentage	NOUN
ejpam-6087	151	172	of	of	ADP
ejpam-6087	151	173	text	text	NOUN
ejpam-6087	151	174	mentioning	mention	VERB
ejpam-6087	151	175	artistic	artistic	ADJ
ejpam-6087	151	176	works	work	NOUN
ejpam-6087	151	177	quantity	quantity	NOUN
ejpam-6087	151	178	percent	percent	NOUN
ejpam-6087	151	179	percentage	percentage	NOUN
ejpam-6087	151	180	of	of	ADP
ejpam-6087	151	181	text	text	NOUN
ejpam-6087	151	182	involving	involve	VERB
ejpam-6087	151	183	quantities	quantity	NOUN
ejpam-6087	151	184	language	language	NOUN
ejpam-6087	151	185	percent	percent	NOUN
ejpam-6087	151	186	percentage	percentage	NOUN
ejpam-6087	151	187	of	of	ADP
ejpam-6087	151	188	text	text	NOUN
ejpam-6087	151	189	referring	refer	VERB
ejpam-6087	151	190	to	to	ADP
ejpam-6087	151	191	languages	language	NOUN
ejpam-6087	151	192	max	max	PROPN
ejpam-6087	151	193	word	word	NOUN
ejpam-6087	151	194	length	length	NOUN
ejpam-6087	151	195	of	of	ADP
ejpam-6087	151	196	the	the	DET
ejpam-6087	151	197	longest	long	ADJ
ejpam-6087	151	198	term	term	NOUN
ejpam-6087	151	199	in	in	ADP
ejpam-6087	151	200	the	the	DET
ejpam-6087	151	201	text	text	NOUN
ejpam-6087	151	202	min	min	NOUN
ejpam-6087	151	203	word	word	NOUN
ejpam-6087	151	204	length	length	NOUN
ejpam-6087	151	205	of	of	ADP
ejpam-6087	151	206	the	the	DET
ejpam-6087	151	207	shortest	short	ADJ
ejpam-6087	151	208	term	term	NOUN
ejpam-6087	151	209	in	in	ADP
ejpam-6087	151	210	the	the	DET
ejpam-6087	151	211	text	text	NOUN
ejpam-6087	151	212	avg	avg	NOUN
ejpam-6087	151	213	word	word	NOUN
ejpam-6087	151	214	length	length	NOUN
ejpam-6087	151	215	average	average	ADJ
ejpam-6087	151	216	length	length	NOUN
ejpam-6087	151	217	of	of	ADP
ejpam-6087	151	218	terms	term	NOUN
ejpam-6087	151	219	within	within	ADP
ejpam-6087	151	220	the	the	DET
ejpam-6087	151	221	text	text	NOUN
ejpam-6087	151	222	the	the	DET
ejpam-6087	151	223	dataset	dataset	NOUN
ejpam-6087	151	224	,	,	PUNCT
ejpam-6087	151	225	respectively	respectively	ADV
ejpam-6087	151	226	.	.	PUNCT
ejpam-6087	152	1	3.2	3.2	NUM
ejpam-6087	152	2	.	.	PUNCT
ejpam-6087	153	1	feature	feature	NOUN
ejpam-6087	153	2	preprocessing	preprocessing	NOUN
ejpam-6087	153	3	:	:	PUNCT
ejpam-6087	153	4	standardization	standardization	VERB
ejpam-6087	153	5	an	an	DET
ejpam-6087	153	6	essential	essential	ADJ
ejpam-6087	153	7	phase	phase	NOUN
ejpam-6087	153	8	in	in	ADP
ejpam-6087	153	9	the	the	DET
ejpam-6087	153	10	machine	machine	NOUN
ejpam-6087	153	11	learning	learn	VERB
ejpam-6087	153	12	workflow	workflow	NOUN
ejpam-6087	153	13	is	be	AUX
ejpam-6087	153	14	feature	feature	NOUN
ejpam-6087	153	15	preprocessing	preprocessing	NOUN
ejpam-6087	153	16	,	,	PUNCT
ejpam-6087	153	17	with	with	ADP
ejpam-6087	153	18	standardization	standardization	NOUN
ejpam-6087	153	19	being	be	AUX
ejpam-6087	153	20	one	one	NUM
ejpam-6087	153	21	of	of	ADP
ejpam-6087	153	22	the	the	DET
ejpam-6087	153	23	most	most	ADV
ejpam-6087	153	24	significant	significant	ADJ
ejpam-6087	153	25	techniques	technique	NOUN
ejpam-6087	153	26	applied	apply	VERB
ejpam-6087	153	27	[	[	PUNCT
ejpam-6087	153	28	26	26	NUM
ejpam-6087	153	29	]	]	PUNCT
ejpam-6087	153	30	.	.	PUNCT
ejpam-6087	154	1	standardization	standardization	NOUN
ejpam-6087	154	2	ensures	ensure	VERB
ejpam-6087	154	3	that	that	SCONJ
ejpam-6087	154	4	all	all	DET
ejpam-6087	154	5	features	feature	NOUN
ejpam-6087	154	6	are	be	AUX
ejpam-6087	154	7	rescaled	rescale	VERB
ejpam-6087	154	8	to	to	PART
ejpam-6087	154	9	have	have	VERB
ejpam-6087	154	10	a	a	DET
ejpam-6087	154	11	mean	mean	NOUN
ejpam-6087	154	12	of	of	ADP
ejpam-6087	154	13	0	0	NUM
ejpam-6087	154	14	and	and	CCONJ
ejpam-6087	154	15	a	a	DET
ejpam-6087	154	16	standard	standard	ADJ
ejpam-6087	154	17	deviation	deviation	NOUN
ejpam-6087	154	18	of	of	ADP
ejpam-6087	154	19	1	1	NUM
ejpam-6087	154	20	,	,	PUNCT
ejpam-6087	154	21	which	which	PRON
ejpam-6087	154	22	is	be	AUX
ejpam-6087	154	23	vital	vital	ADJ
ejpam-6087	154	24	for	for	ADP
ejpam-6087	154	25	the	the	DET
ejpam-6087	154	26	effective	effective	ADJ
ejpam-6087	154	27	functioning	functioning	NOUN
ejpam-6087	154	28	of	of	ADP
ejpam-6087	154	29	many	many	ADJ
ejpam-6087	154	30	machine	machine	NOUN
ejpam-6087	154	31	learning	learn	VERB
ejpam-6087	154	32	algorithms	algorithm	NOUN
ejpam-6087	154	33	.	.	PUNCT
ejpam-6087	155	1	this	this	DET
ejpam-6087	155	2	transformation	transformation	NOUN
ejpam-6087	155	3	is	be	AUX
ejpam-6087	155	4	performed	perform	VERB
ejpam-6087	155	5	independently	independently	ADV
ejpam-6087	155	6	for	for	ADP
ejpam-6087	155	7	each	each	DET
ejpam-6087	155	8	feature	feature	NOUN
ejpam-6087	155	9	using	use	VERB
ejpam-6087	155	10	the	the	DET
ejpam-6087	155	11	following	follow	VERB
ejpam-6087	155	12	formula	formula	NOUN
ejpam-6087	155	13	:	:	PUNCT
ejpam-6087	155	14	x′	x′	PROPN
ejpam-6087	155	15	=	=	SYM
ejpam-6087	155	16	x−	x−	PROPN
ejpam-6087	155	17	µ	µ	PROPN
ejpam-6087	155	18	σ	σ	X
ejpam-6087	155	19	(	(	PUNCT
ejpam-6087	155	20	1	1	NUM
ejpam-6087	155	21	)	)	PUNCT
ejpam-6087	155	22	here	here	ADV
ejpam-6087	155	23	,	,	PUNCT
ejpam-6087	155	24	x	x	PRON
ejpam-6087	155	25	represents	represent	VERB
ejpam-6087	155	26	the	the	DET
ejpam-6087	155	27	original	original	ADJ
ejpam-6087	155	28	feature	feature	NOUN
ejpam-6087	155	29	value	value	NOUN
ejpam-6087	155	30	,	,	PUNCT
ejpam-6087	155	31	µ	µ	X
ejpam-6087	155	32	is	be	AUX
ejpam-6087	155	33	the	the	DET
ejpam-6087	155	34	mean	mean	NOUN
ejpam-6087	155	35	of	of	ADP
ejpam-6087	155	36	the	the	DET
ejpam-6087	155	37	feature	feature	NOUN
ejpam-6087	155	38	,	,	PUNCT
ejpam-6087	155	39	and	and	CCONJ
ejpam-6087	155	40	σ	σ	PROPN
ejpam-6087	155	41	denotes	denote	VERB
ejpam-6087	155	42	the	the	DET
ejpam-6087	155	43	standard	standard	ADJ
ejpam-6087	155	44	deviation	deviation	NOUN
ejpam-6087	155	45	of	of	ADP
ejpam-6087	155	46	the	the	DET
ejpam-6087	155	47	feature	feature	NOUN
ejpam-6087	155	48	.	.	PUNCT
ejpam-6087	156	1	after	after	ADP
ejpam-6087	156	2	applying	apply	VERB
ejpam-6087	156	3	this	this	DET
ejpam-6087	156	4	formula	formula	NOUN
ejpam-6087	156	5	,	,	PUNCT
ejpam-6087	156	6	each	each	DET
ejpam-6087	156	7	feature	feature	NOUN
ejpam-6087	156	8	is	be	AUX
ejpam-6087	156	9	standardized	standardize	VERB
ejpam-6087	156	10	to	to	PART
ejpam-6087	156	11	have	have	VERB
ejpam-6087	156	12	a	a	DET
ejpam-6087	156	13	mean	mean	NOUN
ejpam-6087	156	14	of	of	ADP
ejpam-6087	156	15	0	0	NUM
ejpam-6087	156	16	and	and	CCONJ
ejpam-6087	156	17	a	a	DET
ejpam-6087	156	18	standard	standard	ADJ
ejpam-6087	156	19	deviation	deviation	NOUN
ejpam-6087	156	20	of	of	ADP
ejpam-6087	156	21	1	1	NUM
ejpam-6087	156	22	.	.	PUNCT
ejpam-6087	157	1	standardization	standardization	NOUN
ejpam-6087	157	2	is	be	AUX
ejpam-6087	157	3	particularly	particularly	ADV
ejpam-6087	157	4	critical	critical	ADJ
ejpam-6087	157	5	for	for	ADP
ejpam-6087	157	6	algorithms	algorithm	NOUN
ejpam-6087	157	7	that	that	PRON
ejpam-6087	157	8	are	be	AUX
ejpam-6087	157	9	sensitive	sensitive	ADJ
ejpam-6087	157	10	to	to	ADP
ejpam-6087	157	11	the	the	DET
ejpam-6087	157	12	scale	scale	NOUN
ejpam-6087	157	13	of	of	ADP
ejpam-6087	157	14	input	input	NOUN
ejpam-6087	157	15	data	datum	NOUN
ejpam-6087	157	16	.	.	PUNCT
ejpam-6087	158	1	models	model	NOUN
ejpam-6087	158	2	such	such	ADJ
ejpam-6087	158	3	as	as	ADP
ejpam-6087	158	4	logistic	logistic	ADJ
ejpam-6087	158	5	regression	regression	NOUN
ejpam-6087	158	6	(	(	PUNCT
ejpam-6087	158	7	lr	lr	NOUN
ejpam-6087	158	8	)	)	PUNCT
ejpam-6087	158	9	,	,	PUNCT
ejpam-6087	158	10	support	support	VERB
ejpam-6087	158	11	vector	vector	NOUN
ejpam-6087	158	12	machines	machine	NOUN
ejpam-6087	158	13	(	(	PUNCT
ejpam-6087	158	14	svm	svm	PROPN
ejpam-6087	158	15	)	)	PUNCT
ejpam-6087	158	16	,	,	PUNCT
ejpam-6087	158	17	and	and	CCONJ
ejpam-6087	158	18	neural	neural	ADJ
ejpam-6087	158	19	networks	network	NOUN
ejpam-6087	158	20	(	(	PUNCT
ejpam-6087	158	21	nn	nn	NOUN
ejpam-6087	158	22	)	)	PUNCT
ejpam-6087	158	23	rely	rely	VERB
ejpam-6087	158	24	on	on	ADP
ejpam-6087	158	25	consistent	consistent	ADJ
ejpam-6087	158	26	feature	feature	NOUN
ejpam-6087	158	27	scaling	scale	VERB
ejpam-6087	158	28	to	to	PART
ejpam-6087	158	29	ensure	ensure	VERB
ejpam-6087	158	30	effective	effective	ADJ
ejpam-6087	158	31	learning	learning	NOUN
ejpam-6087	158	32	.	.	PUNCT
ejpam-6087	159	1	without	without	ADP
ejpam-6087	159	2	standardization	standardization	NOUN
ejpam-6087	159	3	,	,	PUNCT
ejpam-6087	159	4	algorithms	algorithm	NOUN
ejpam-6087	159	5	may	may	AUX
ejpam-6087	159	6	disproportionately	disproportionately	ADV
ejpam-6087	159	7	emphasize	emphasize	VERB
ejpam-6087	159	8	features	feature	NOUN
ejpam-6087	159	9	with	with	ADP
ejpam-6087	159	10	larger	large	ADJ
ejpam-6087	159	11	ranges	range	NOUN
ejpam-6087	159	12	,	,	PUNCT
ejpam-6087	159	13	resulting	result	VERB
ejpam-6087	159	14	in	in	ADP
ejpam-6087	159	15	suboptimal	suboptimal	ADJ
ejpam-6087	159	16	model	model	NOUN
ejpam-6087	159	17	performance	performance	NOUN
ejpam-6087	159	18	.	.	PUNCT
ejpam-6087	160	1	by	by	ADP
ejpam-6087	160	2	normalizing	normalize	VERB
ejpam-6087	160	3	the	the	DET
ejpam-6087	160	4	feature	feature	NOUN
ejpam-6087	160	5	scales	scale	NOUN
ejpam-6087	160	6	,	,	PUNCT
ejpam-6087	160	7	these	these	DET
ejpam-6087	160	8	algorithms	algorithm	NOUN
ejpam-6087	160	9	can	can	AUX
ejpam-6087	160	10	focus	focus	VERB
ejpam-6087	160	11	on	on	ADP
ejpam-6087	160	12	identifying	identify	VERB
ejpam-6087	160	13	the	the	DET
ejpam-6087	160	14	true	true	ADJ
ejpam-6087	160	15	relationships	relationship	NOUN
ejpam-6087	160	16	between	between	ADP
ejpam-6087	160	17	the	the	DET
ejpam-6087	160	18	input	input	NOUN
ejpam-6087	160	19	features	feature	NOUN
ejpam-6087	160	20	and	and	CCONJ
ejpam-6087	160	21	the	the	DET
ejpam-6087	160	22	target	target	NOUN
ejpam-6087	160	23	variable	variable	NOUN
ejpam-6087	160	24	,	,	PUNCT
ejpam-6087	160	25	rather	rather	ADV
ejpam-6087	160	26	than	than	ADP
ejpam-6087	160	27	being	be	AUX
ejpam-6087	160	28	misled	mislead	VERB
ejpam-6087	160	29	by	by	ADP
ejpam-6087	160	30	differences	difference	NOUN
ejpam-6087	160	31	in	in	ADP
ejpam-6087	160	32	feature	feature	NOUN
ejpam-6087	160	33	magnitudes	magnitude	NOUN
ejpam-6087	160	34	.	.	PUNCT
ejpam-6087	161	1	in	in	ADP
ejpam-6087	161	2	addition	addition	NOUN
ejpam-6087	161	3	,	,	PUNCT
ejpam-6087	161	4	standardization	standardization	NOUN
ejpam-6087	161	5	improves	improve	VERB
ejpam-6087	161	6	the	the	DET
ejpam-6087	161	7	numerical	numerical	ADJ
ejpam-6087	161	8	stability	stability	NOUN
ejpam-6087	161	9	and	and	CCONJ
ejpam-6087	161	10	efficiency	efficiency	NOUN
ejpam-6087	161	11	of	of	ADP
ejpam-6087	161	12	optimization	optimization	NOUN
ejpam-6087	161	13	algorithms	algorithm	NOUN
ejpam-6087	161	14	.	.	PUNCT
ejpam-6087	162	1	standardized	standardized	ADJ
ejpam-6087	162	2	features	feature	NOUN
ejpam-6087	162	3	minimize	minimize	VERB
ejpam-6087	162	4	the	the	DET
ejpam-6087	162	5	m.	m.	NOUN
ejpam-6087	162	6	al	al	PROPN
ejpam-6087	162	7	-	-	PUNCT
ejpam-6087	162	8	tarawneh	tarawneh	PROPN
ejpam-6087	162	9	et	et	PROPN
ejpam-6087	162	10	al	al	PROPN
ejpam-6087	162	11	.	.	PUNCT
ejpam-6087	162	12	/	/	SYM
ejpam-6087	162	13	eur	eur	PROPN
ejpam-6087	162	14	.	.	PUNCT
ejpam-6087	163	1	j.	j.	PROPN
ejpam-6087	163	2	pure	pure	PROPN
ejpam-6087	163	3	appl	appl	PROPN
ejpam-6087	163	4	.	.	PROPN
ejpam-6087	163	5	math	math	PROPN
ejpam-6087	163	6	,	,	PUNCT
ejpam-6087	163	7	18	18	NUM
ejpam-6087	163	8	(	(	PUNCT
ejpam-6087	163	9	2	2	NUM
ejpam-6087	163	10	)	)	PUNCT
ejpam-6087	163	11	(	(	PUNCT
ejpam-6087	163	12	2025	2025	NUM
ejpam-6087	163	13	)	)	PUNCT
ejpam-6087	163	14	,	,	PUNCT
ejpam-6087	163	15	6087	6087	NUM
ejpam-6087	163	16	9	9	NUM
ejpam-6087	163	17	of	of	ADP
ejpam-6087	163	18	39	39	NUM
ejpam-6087	163	19	table	table	NOUN
ejpam-6087	163	20	2	2	NUM
ejpam-6087	163	21	:	:	PUNCT
ejpam-6087	163	22	lexical	lexical	ADJ
ejpam-6087	163	23	features	feature	NOUN
ejpam-6087	163	24	for	for	ADP
ejpam-6087	163	25	fake	fake	ADJ
ejpam-6087	163	26	news	news	NOUN
ejpam-6087	163	27	detection	detection	NOUN
ejpam-6087	163	28	name	name	NOUN
ejpam-6087	163	29	description	description	NOUN
ejpam-6087	163	30	present	present	ADJ
ejpam-6087	163	31	verb	verb	NOUN
ejpam-6087	163	32	count	count	NOUN
ejpam-6087	163	33	of	of	ADP
ejpam-6087	163	34	present	present	ADJ
ejpam-6087	163	35	tense	tense	ADJ
ejpam-6087	163	36	verbs	verbs	PROPN
ejpam-6087	163	37	past	past	ADP
ejpam-6087	163	38	verb	verb	NOUN
ejpam-6087	163	39	count	count	NOUN
ejpam-6087	163	40	of	of	ADP
ejpam-6087	163	41	past	past	ADJ
ejpam-6087	163	42	tense	tense	PROPN
ejpam-6087	163	43	verbs	verbs	PROPN
ejpam-6087	163	44	adjectives	adjective	NOUN
ejpam-6087	163	45	total	total	ADJ
ejpam-6087	163	46	number	number	NOUN
ejpam-6087	163	47	of	of	ADP
ejpam-6087	163	48	adjectives	adjective	NOUN
ejpam-6087	163	49	in	in	ADP
ejpam-6087	163	50	the	the	DET
ejpam-6087	163	51	text	text	NOUN
ejpam-6087	163	52	pronouns	pronoun	NOUN
ejpam-6087	163	53	total	total	ADJ
ejpam-6087	163	54	number	number	NOUN
ejpam-6087	163	55	of	of	ADP
ejpam-6087	163	56	pronouns	pronoun	NOUN
ejpam-6087	163	57	in	in	ADP
ejpam-6087	163	58	the	the	DET
ejpam-6087	163	59	text	text	NOUN
ejpam-6087	163	60	to	to	AUX
ejpam-6087	163	61	’s	’s	NOUN
ejpam-6087	163	62	frequency	frequency	NOUN
ejpam-6087	163	63	of	of	ADP
ejpam-6087	163	64	the	the	DET
ejpam-6087	163	65	word	word	NOUN
ejpam-6087	163	66	”	"	PUNCT
ejpam-6087	163	67	to	to	ADP
ejpam-6087	163	68	”	"	PUNCT
ejpam-6087	163	69	determiners	determiner	NOUN
ejpam-6087	163	70	count	count	VERB
ejpam-6087	163	71	of	of	ADP
ejpam-6087	163	72	determiner	determiner	NOUN
ejpam-6087	163	73	words	word	NOUN
ejpam-6087	163	74	conjunctions	conjunction	NOUN
ejpam-6087	163	75	total	total	ADJ
ejpam-6087	163	76	count	count	NOUN
ejpam-6087	163	77	of	of	ADP
ejpam-6087	163	78	conjunctions	conjunction	NOUN
ejpam-6087	163	79	dots	dot	NOUN
ejpam-6087	163	80	instances	instance	NOUN
ejpam-6087	163	81	of	of	ADP
ejpam-6087	163	82	period	period	NOUN
ejpam-6087	163	83	(	(	PUNCT
ejpam-6087	163	84	.	.	PUNCT
ejpam-6087	163	85	)	)	PUNCT
ejpam-6087	163	86	usage	usage	NOUN
ejpam-6087	163	87	exclamations	exclamation	NOUN
ejpam-6087	163	88	instances	instance	NOUN
ejpam-6087	163	89	of	of	ADP
ejpam-6087	163	90	exclamation	exclamation	NOUN
ejpam-6087	163	91	marks	mark	NOUN
ejpam-6087	163	92	(	(	PUNCT
ejpam-6087	163	93	!	!	PUNCT
ejpam-6087	163	94	)	)	PUNCT
ejpam-6087	163	95	usage	usage	NOUN
ejpam-6087	163	96	question	question	NOUN
ejpam-6087	163	97	instances	instance	NOUN
ejpam-6087	163	98	of	of	ADP
ejpam-6087	163	99	question	question	NOUN
ejpam-6087	163	100	marks	mark	NOUN
ejpam-6087	163	101	(	(	PUNCT
ejpam-6087	163	102	?	?	PUNCT
ejpam-6087	163	103	)	)	PUNCT
ejpam-6087	163	104	usage	usage	NOUN
ejpam-6087	163	105	ampersand	ampersand	ADJ
ejpam-6087	163	106	frequency	frequency	NOUN
ejpam-6087	163	107	of	of	ADP
ejpam-6087	163	108	ampersand	ampersand	ADJ
ejpam-6087	163	109	(	(	PUNCT
ejpam-6087	163	110	&	&	CCONJ
ejpam-6087	163	111	)	)	PUNCT
ejpam-6087	163	112	usage	usage	NOUN
ejpam-6087	163	113	capitals	capital	NOUN
ejpam-6087	163	114	count	count	NOUN
ejpam-6087	163	115	of	of	ADP
ejpam-6087	163	116	capitalized	capitalize	VERB
ejpam-6087	163	117	letters	letter	NOUN
ejpam-6087	163	118	quotes	quote	VERB
ejpam-6087	163	119	total	total	ADJ
ejpam-6087	163	120	instances	instance	NOUN
ejpam-6087	163	121	of	of	ADP
ejpam-6087	163	122	quotation	quotation	NOUN
ejpam-6087	163	123	marks	mark	NOUN
ejpam-6087	163	124	digits	digit	NOUN
ejpam-6087	163	125	frequency	frequency	NOUN
ejpam-6087	163	126	of	of	ADP
ejpam-6087	163	127	numerical	numerical	ADJ
ejpam-6087	163	128	digits	digit	NOUN
ejpam-6087	163	129	(	(	PUNCT
ejpam-6087	163	130	0	0	NUM
ejpam-6087	163	131	-	-	SYM
ejpam-6087	163	132	9	9	NUM
ejpam-6087	163	133	)	)	PUNCT
ejpam-6087	163	134	long	long	ADJ
ejpam-6087	163	135	word	word	NOUN
ejpam-6087	163	136	freq	freq	ADJ
ejpam-6087	163	137	frequency	frequency	NOUN
ejpam-6087	163	138	of	of	ADP
ejpam-6087	163	139	long	long	ADJ
ejpam-6087	163	140	words	word	NOUN
ejpam-6087	163	141	short	short	ADJ
ejpam-6087	163	142	word	word	NOUN
ejpam-6087	163	143	freq	freq	ADJ
ejpam-6087	163	144	frequency	frequency	NOUN
ejpam-6087	163	145	of	of	ADP
ejpam-6087	163	146	short	short	ADJ
ejpam-6087	163	147	words	word	NOUN
ejpam-6087	163	148	table	table	NOUN
ejpam-6087	163	149	3	3	NUM
ejpam-6087	163	150	:	:	PUNCT
ejpam-6087	163	151	meta	meta	ADJ
ejpam-6087	163	152	-	-	PUNCT
ejpam-6087	163	153	data	data	NOUN
ejpam-6087	163	154	features	feature	NOUN
ejpam-6087	163	155	for	for	ADP
ejpam-6087	163	156	twitter	twitter	NOUN
ejpam-6087	163	157	users	user	NOUN
ejpam-6087	163	158	name	name	NOUN
ejpam-6087	163	159	description	description	NOUN
ejpam-6087	163	160	followers	follower	NOUN
ejpam-6087	163	161	count	count	VERB
ejpam-6087	163	162	total	total	ADJ
ejpam-6087	163	163	number	number	NOUN
ejpam-6087	163	164	of	of	ADP
ejpam-6087	163	165	followers	follower	NOUN
ejpam-6087	163	166	of	of	ADP
ejpam-6087	163	167	the	the	DET
ejpam-6087	163	168	user	user	NOUN
ejpam-6087	163	169	friends	friend	NOUN
ejpam-6087	163	170	count	count	VERB
ejpam-6087	163	171	total	total	ADJ
ejpam-6087	163	172	number	number	NOUN
ejpam-6087	163	173	of	of	ADP
ejpam-6087	163	174	friends	friend	NOUN
ejpam-6087	163	175	connected	connect	VERB
ejpam-6087	163	176	to	to	ADP
ejpam-6087	163	177	the	the	DET
ejpam-6087	163	178	user	user	NOUN
ejpam-6087	163	179	favourites	favourite	NOUN
ejpam-6087	163	180	count	count	VERB
ejpam-6087	163	181	total	total	ADJ
ejpam-6087	163	182	likes	like	NOUN
ejpam-6087	163	183	given	give	VERB
ejpam-6087	163	184	across	across	ADP
ejpam-6087	163	185	all	all	DET
ejpam-6087	163	186	user	user	NOUN
ejpam-6087	163	187	tweets	tweet	NOUN
ejpam-6087	163	188	statuses	status	NOUN
ejpam-6087	163	189	count	count	VERB
ejpam-6087	163	190	overall	overall	ADJ
ejpam-6087	163	191	count	count	NOUN
ejpam-6087	163	192	of	of	ADP
ejpam-6087	163	193	tweets	tweet	NOUN
ejpam-6087	163	194	by	by	ADP
ejpam-6087	163	195	the	the	DET
ejpam-6087	163	196	user	user	NOUN
ejpam-6087	163	197	listed	list	VERB
ejpam-6087	163	198	count	count	NOUN
ejpam-6087	163	199	number	number	NOUN
ejpam-6087	163	200	of	of	ADP
ejpam-6087	163	201	times	time	NOUN
ejpam-6087	163	202	the	the	DET
ejpam-6087	163	203	user	user	NOUN
ejpam-6087	163	204	’s	’s	PART
ejpam-6087	163	205	tweets	tweet	NOUN
ejpam-6087	163	206	are	be	AUX
ejpam-6087	163	207	listed	list	VERB
ejpam-6087	163	208	mentions	mention	NOUN
ejpam-6087	163	209	total	total	ADJ
ejpam-6087	163	210	mentions	mention	NOUN
ejpam-6087	163	211	of	of	ADP
ejpam-6087	163	212	the	the	DET
ejpam-6087	163	213	user	user	NOUN
ejpam-6087	163	214	in	in	ADP
ejpam-6087	163	215	tweets	tweet	NOUN
ejpam-6087	163	216	quotes	quote	NOUN
ejpam-6087	163	217	count	count	NOUN
ejpam-6087	163	218	of	of	ADP
ejpam-6087	163	219	times	time	NOUN
ejpam-6087	163	220	the	the	DET
ejpam-6087	163	221	user	user	NOUN
ejpam-6087	163	222	’s	’s	PART
ejpam-6087	163	223	tweets	tweet	NOUN
ejpam-6087	163	224	were	be	AUX
ejpam-6087	163	225	quoted	quote	VERB
ejpam-6087	163	226	replies	reply	NOUN
ejpam-6087	163	227	count	count	NOUN
ejpam-6087	163	228	of	of	ADP
ejpam-6087	163	229	replies	reply	NOUN
ejpam-6087	163	230	received	receive	VERB
ejpam-6087	163	231	by	by	ADP
ejpam-6087	163	232	the	the	DET
ejpam-6087	163	233	user	user	NOUN
ejpam-6087	163	234	retweets	retweet	VERB
ejpam-6087	163	235	total	total	ADJ
ejpam-6087	163	236	retweets	retweet	NOUN
ejpam-6087	163	237	of	of	ADP
ejpam-6087	163	238	the	the	DET
ejpam-6087	163	239	user	user	NOUN
ejpam-6087	163	240	’s	’s	PART
ejpam-6087	163	241	content	content	NOUN
ejpam-6087	163	242	favourites	favourite	NOUN
ejpam-6087	163	243	count	count	NOUN
ejpam-6087	163	244	of	of	ADP
ejpam-6087	163	245	times	time	NOUN
ejpam-6087	163	246	the	the	DET
ejpam-6087	163	247	user	user	NOUN
ejpam-6087	163	248	’s	’s	PART
ejpam-6087	163	249	tweets	tweet	NOUN
ejpam-6087	163	250	were	be	AUX
ejpam-6087	163	251	favorited	favorite	VERB
ejpam-6087	163	252	hashtags	hashtag	NOUN
ejpam-6087	163	253	number	number	NOUN
ejpam-6087	163	254	of	of	ADP
ejpam-6087	163	255	hashtags	hashtag	NOUN
ejpam-6087	163	256	used	use	VERB
ejpam-6087	163	257	in	in	ADP
ejpam-6087	163	258	the	the	DET
ejpam-6087	163	259	user	user	NOUN
ejpam-6087	163	260	’s	’s	PART
ejpam-6087	163	261	tweets	tweet	NOUN
ejpam-6087	163	262	urls	url	NOUN
ejpam-6087	163	263	whether	whether	SCONJ
ejpam-6087	163	264	the	the	DET
ejpam-6087	163	265	user	user	NOUN
ejpam-6087	163	266	’s	’s	PART
ejpam-6087	163	267	profile	profile	NOUN
ejpam-6087	163	268	contains	contain	VERB
ejpam-6087	163	269	a	a	DET
ejpam-6087	163	270	url	url	NOUN
ejpam-6087	163	271	botscorebinary	botscorebinary	NOUN
ejpam-6087	163	272	binary	binary	ADJ
ejpam-6087	163	273	indicator	indicator	NOUN
ejpam-6087	163	274	of	of	ADP
ejpam-6087	163	275	the	the	DET
ejpam-6087	163	276	user	user	NOUN
ejpam-6087	163	277	being	be	AUX
ejpam-6087	163	278	a	a	DET
ejpam-6087	163	279	bot	bot	NOUN
ejpam-6087	163	280	cred	cre	VERB
ejpam-6087	163	281	score	score	NOUN
ejpam-6087	163	282	measuring	measure	VERB
ejpam-6087	163	283	the	the	DET
ejpam-6087	163	284	user	user	NOUN
ejpam-6087	163	285	’s	’s	PART
ejpam-6087	163	286	credibility	credibility	NOUN
ejpam-6087	163	287	normalized	normalize	VERB
ejpam-6087	163	288	influence	influence	NOUN
ejpam-6087	163	289	normalized	normalize	VERB
ejpam-6087	163	290	score	score	NOUN
ejpam-6087	163	291	of	of	ADP
ejpam-6087	163	292	the	the	DET
ejpam-6087	163	293	user	user	NOUN
ejpam-6087	163	294	’s	’s	PART
ejpam-6087	163	295	influence	influence	NOUN
ejpam-6087	163	296	risk	risk	NOUN
ejpam-6087	163	297	of	of	ADP
ejpam-6087	163	298	numerical	numerical	ADJ
ejpam-6087	163	299	errors	error	NOUN
ejpam-6087	163	300	,	,	PUNCT
ejpam-6087	163	301	such	such	ADJ
ejpam-6087	163	302	as	as	ADP
ejpam-6087	163	303	overflow	overflow	NOUN
ejpam-6087	163	304	or	or	CCONJ
ejpam-6087	163	305	underflow	underflow	NOUN
ejpam-6087	163	306	,	,	PUNCT
ejpam-6087	163	307	that	that	PRON
ejpam-6087	163	308	can	can	AUX
ejpam-6087	163	309	arise	arise	VERB
ejpam-6087	163	310	during	during	ADP
ejpam-6087	163	311	complex	complex	ADJ
ejpam-6087	163	312	calculations	calculation	NOUN
ejpam-6087	163	313	.	.	PUNCT
ejpam-6087	164	1	another	another	DET
ejpam-6087	164	2	significant	significant	ADJ
ejpam-6087	164	3	benefit	benefit	NOUN
ejpam-6087	164	4	of	of	ADP
ejpam-6087	164	5	standardization	standardization	NOUN
ejpam-6087	164	6	is	be	AUX
ejpam-6087	164	7	its	its	PRON
ejpam-6087	164	8	role	role	NOUN
ejpam-6087	164	9	in	in	ADP
ejpam-6087	164	10	accelerating	accelerate	VERB
ejpam-6087	164	11	optimization	optimization	NOUN
ejpam-6087	164	12	algorithms	algorithm	NOUN
ejpam-6087	164	13	,	,	PUNCT
ejpam-6087	164	14	such	such	ADJ
ejpam-6087	164	15	as	as	ADP
ejpam-6087	164	16	gradient	gradient	ADJ
ejpam-6087	164	17	descent	descent	NOUN
ejpam-6087	164	18	.	.	PUNCT
ejpam-6087	165	1	when	when	SCONJ
ejpam-6087	165	2	features	feature	NOUN
ejpam-6087	165	3	are	be	AUX
ejpam-6087	165	4	on	on	ADP
ejpam-6087	165	5	a	a	DET
ejpam-6087	165	6	uniform	uniform	ADJ
ejpam-6087	165	7	scale	scale	NOUN
ejpam-6087	165	8	,	,	PUNCT
ejpam-6087	165	9	the	the	DET
ejpam-6087	165	10	optimization	optimization	NOUN
ejpam-6087	165	11	process	process	NOUN
ejpam-6087	165	12	becomes	become	VERB
ejpam-6087	165	13	more	more	ADV
ejpam-6087	165	14	stable	stable	ADJ
ejpam-6087	165	15	and	and	CCONJ
ejpam-6087	165	16	achieves	achieve	VERB
ejpam-6087	165	17	convergence	convergence	NOUN
ejpam-6087	165	18	faster	fast	ADV
ejpam-6087	165	19	,	,	PUNCT
ejpam-6087	165	20	helping	help	VERB
ejpam-6087	165	21	the	the	DET
ejpam-6087	165	22	model	model	NOUN
ejpam-6087	165	23	to	to	PART
ejpam-6087	165	24	find	find	VERB
ejpam-6087	165	25	the	the	DET
ejpam-6087	165	26	optimal	optimal	ADJ
ejpam-6087	165	27	solution	solution	NOUN
ejpam-6087	165	28	efficiently	efficiently	ADV
ejpam-6087	165	29	.	.	PUNCT
ejpam-6087	166	1	3.3	3.3	NUM
ejpam-6087	166	2	.	.	PUNCT
ejpam-6087	167	1	classification	classification	NOUN
ejpam-6087	167	2	techniques	technique	NOUN
ejpam-6087	167	3	this	this	DET
ejpam-6087	167	4	section	section	NOUN
ejpam-6087	167	5	outlines	outline	VERB
ejpam-6087	167	6	the	the	DET
ejpam-6087	167	7	classification	classification	NOUN
ejpam-6087	167	8	techniques	technique	NOUN
ejpam-6087	167	9	employed	employ	VERB
ejpam-6087	167	10	in	in	ADP
ejpam-6087	167	11	the	the	DET
ejpam-6087	167	12	study	study	NOUN
ejpam-6087	167	13	,	,	PUNCT
ejpam-6087	167	14	focusing	focus	VERB
ejpam-6087	167	15	on	on	ADP
ejpam-6087	167	16	both	both	PRON
ejpam-6087	167	17	individual	individual	ADJ
ejpam-6087	167	18	classifiers	classifier	NOUN
ejpam-6087	167	19	and	and	CCONJ
ejpam-6087	167	20	ensemble	ensemble	ADJ
ejpam-6087	167	21	methods	method	NOUN
ejpam-6087	167	22	.	.	PUNCT
ejpam-6087	168	1	the	the	DET
ejpam-6087	168	2	evaluation	evaluation	NOUN
ejpam-6087	168	3	process	process	NOUN
ejpam-6087	168	4	aims	aim	VERB
ejpam-6087	168	5	to	to	PART
ejpam-6087	168	6	assess	assess	VERB
ejpam-6087	168	7	their	their	PRON
ejpam-6087	168	8	effectiveness	effectiveness	NOUN
ejpam-6087	168	9	and	and	CCONJ
ejpam-6087	168	10	identify	identify	VERB
ejpam-6087	168	11	the	the	DET
ejpam-6087	168	12	most	most	ADV
ejpam-6087	168	13	suitable	suitable	ADJ
ejpam-6087	168	14	models	model	NOUN
ejpam-6087	168	15	for	for	ADP
ejpam-6087	168	16	integration	integration	NOUN
ejpam-6087	168	17	into	into	ADP
ejpam-6087	168	18	ensemble	ensemble	ADJ
ejpam-6087	168	19	frameworks	framework	NOUN
ejpam-6087	168	20	.	.	PUNCT
ejpam-6087	169	1	the	the	DET
ejpam-6087	169	2	classification	classification	NOUN
ejpam-6087	169	3	approaches	approach	NOUN
ejpam-6087	169	4	are	be	AUX
ejpam-6087	169	5	categorized	categorize	VERB
ejpam-6087	169	6	into	into	ADP
ejpam-6087	169	7	two	two	NUM
ejpam-6087	169	8	groups	group	NOUN
ejpam-6087	169	9	:	:	PUNCT
ejpam-6087	169	10	base	base	NOUN
ejpam-6087	169	11	classifiers	classifier	NOUN
ejpam-6087	169	12	and	and	CCONJ
ejpam-6087	169	13	ensemble	ensemble	ADJ
ejpam-6087	169	14	methods	method	NOUN
ejpam-6087	169	15	,	,	PUNCT
ejpam-6087	169	16	with	with	ADP
ejpam-6087	169	17	the	the	DET
ejpam-6087	169	18	latter	latter	ADJ
ejpam-6087	169	19	further	far	ADV
ejpam-6087	169	20	divided	divide	VERB
ejpam-6087	169	21	into	into	ADP
ejpam-6087	169	22	homogeneous	homogeneous	ADJ
ejpam-6087	169	23	and	and	CCONJ
ejpam-6087	169	24	heterogem	heterogem	PROPN
ejpam-6087	169	25	.	.	PUNCT
ejpam-6087	170	1	al	al	PROPN
ejpam-6087	170	2	-	-	PUNCT
ejpam-6087	170	3	tarawneh	tarawneh	PROPN
ejpam-6087	170	4	et	et	PROPN
ejpam-6087	170	5	al	al	PROPN
ejpam-6087	170	6	.	.	PUNCT
ejpam-6087	170	7	/	/	SYM
ejpam-6087	170	8	eur	eur	PROPN
ejpam-6087	170	9	.	.	PUNCT
ejpam-6087	171	1	j.	j.	PROPN
ejpam-6087	171	2	pure	pure	PROPN
ejpam-6087	171	3	appl	appl	PROPN
ejpam-6087	171	4	.	.	PROPN
ejpam-6087	171	5	math	math	PROPN
ejpam-6087	171	6	,	,	PUNCT
ejpam-6087	171	7	18	18	NUM
ejpam-6087	171	8	(	(	PUNCT
ejpam-6087	171	9	2	2	NUM
ejpam-6087	171	10	)	)	PUNCT
ejpam-6087	171	11	(	(	PUNCT
ejpam-6087	171	12	2025	2025	NUM
ejpam-6087	171	13	)	)	PUNCT
ejpam-6087	171	14	,	,	PUNCT
ejpam-6087	171	15	6087	6087	NUM
ejpam-6087	171	16	10	10	NUM
ejpam-6087	171	17	of	of	ADP
ejpam-6087	171	18	39	39	NUM
ejpam-6087	171	19	neous	neous	ADJ
ejpam-6087	171	20	strategies	strategy	NOUN
ejpam-6087	171	21	.	.	PUNCT
ejpam-6087	172	1	3.3.1	3.3.1	X
ejpam-6087	172	2	.	.	PUNCT
ejpam-6087	172	3	base	base	NOUN
ejpam-6087	172	4	classifiers	classifier	NOUN
ejpam-6087	172	5	the	the	DET
ejpam-6087	172	6	study	study	NOUN
ejpam-6087	172	7	begins	begin	VERB
ejpam-6087	172	8	by	by	ADP
ejpam-6087	172	9	evaluating	evaluate	VERB
ejpam-6087	172	10	various	various	ADJ
ejpam-6087	172	11	individual	individual	ADJ
ejpam-6087	172	12	classifiers	classifier	NOUN
ejpam-6087	172	13	to	to	PART
ejpam-6087	172	14	determine	determine	VERB
ejpam-6087	172	15	their	their	PRON
ejpam-6087	172	16	effectiveness	effectiveness	NOUN
ejpam-6087	172	17	for	for	ADP
ejpam-6087	172	18	inclusion	inclusion	NOUN
ejpam-6087	172	19	in	in	ADP
ejpam-6087	172	20	the	the	DET
ejpam-6087	172	21	ensemble	ensemble	NOUN
ejpam-6087	172	22	.	.	PUNCT
ejpam-6087	173	1	identifying	identify	VERB
ejpam-6087	173	2	high	high	ADJ
ejpam-6087	173	3	-	-	PUNCT
ejpam-6087	173	4	performing	perform	VERB
ejpam-6087	173	5	models	model	NOUN
ejpam-6087	173	6	is	be	AUX
ejpam-6087	173	7	essential	essential	ADJ
ejpam-6087	173	8	for	for	ADP
ejpam-6087	173	9	constructing	construct	VERB
ejpam-6087	173	10	robust	robust	ADJ
ejpam-6087	173	11	ensemble	ensemble	ADJ
ejpam-6087	173	12	methods	method	NOUN
ejpam-6087	173	13	[	[	X
ejpam-6087	173	14	27–29	27–29	NUM
ejpam-6087	173	15	]	]	PUNCT
ejpam-6087	173	16	.	.	PUNCT
ejpam-6087	174	1	•	•	NUM
ejpam-6087	174	2	decision	decision	NOUN
ejpam-6087	174	3	trees	tree	NOUN
ejpam-6087	174	4	(	(	PUNCT
ejpam-6087	174	5	dt	dt	NOUN
ejpam-6087	174	6	)	)	PUNCT
ejpam-6087	174	7	use	use	VERB
ejpam-6087	174	8	a	a	DET
ejpam-6087	174	9	tree	tree	NOUN
ejpam-6087	174	10	-	-	PUNCT
ejpam-6087	174	11	like	like	ADJ
ejpam-6087	174	12	structure	structure	NOUN
ejpam-6087	174	13	to	to	PART
ejpam-6087	174	14	model	model	VERB
ejpam-6087	174	15	decision	decision	NOUN
ejpam-6087	174	16	-	-	PUNCT
ejpam-6087	174	17	making	make	VERB
ejpam-6087	174	18	processes	process	NOUN
ejpam-6087	174	19	,	,	PUNCT
ejpam-6087	174	20	where	where	SCONJ
ejpam-6087	174	21	nodes	node	NOUN
ejpam-6087	174	22	represent	represent	VERB
ejpam-6087	174	23	feature	feature	NOUN
ejpam-6087	174	24	tests	test	NOUN
ejpam-6087	174	25	,	,	PUNCT
ejpam-6087	174	26	branches	branch	NOUN
ejpam-6087	174	27	denote	denote	VERB
ejpam-6087	174	28	outcomes	outcome	NOUN
ejpam-6087	174	29	,	,	PUNCT
ejpam-6087	174	30	and	and	CCONJ
ejpam-6087	174	31	leaves	leave	NOUN
ejpam-6087	174	32	indicate	indicate	VERB
ejpam-6087	174	33	class	class	NOUN
ejpam-6087	174	34	labels	label	NOUN
ejpam-6087	174	35	.	.	PUNCT
ejpam-6087	175	1	the	the	DET
ejpam-6087	175	2	tree	tree	NOUN
ejpam-6087	175	3	is	be	AUX
ejpam-6087	175	4	built	build	VERB
ejpam-6087	175	5	by	by	ADP
ejpam-6087	175	6	partitioning	partition	VERB
ejpam-6087	175	7	the	the	DET
ejpam-6087	175	8	data	datum	NOUN
ejpam-6087	175	9	based	base	VERB
ejpam-6087	175	10	on	on	ADP
ejpam-6087	175	11	attributes	attribute	NOUN
ejpam-6087	175	12	that	that	PRON
ejpam-6087	175	13	maximize	maximize	VERB
ejpam-6087	175	14	information	information	NOUN
ejpam-6087	175	15	gain	gain	NOUN
ejpam-6087	175	16	.	.	PUNCT
ejpam-6087	176	1	dts	dt	NOUN
ejpam-6087	176	2	are	be	AUX
ejpam-6087	176	3	easy	easy	ADJ
ejpam-6087	176	4	to	to	PART
ejpam-6087	176	5	interpret	interpret	VERB
ejpam-6087	176	6	and	and	CCONJ
ejpam-6087	176	7	visualize	visualize	VERB
ejpam-6087	176	8	,	,	PUNCT
ejpam-6087	176	9	accommodating	accommodate	VERB
ejpam-6087	176	10	both	both	CCONJ
ejpam-6087	176	11	numerical	numerical	ADJ
ejpam-6087	176	12	and	and	CCONJ
ejpam-6087	176	13	categorical	categorical	ADJ
ejpam-6087	176	14	features	feature	NOUN
ejpam-6087	176	15	.	.	PUNCT
ejpam-6087	177	1	however	however	ADV
ejpam-6087	177	2	,	,	PUNCT
ejpam-6087	177	3	they	they	PRON
ejpam-6087	177	4	are	be	AUX
ejpam-6087	177	5	susceptible	susceptible	ADJ
ejpam-6087	177	6	to	to	ADP
ejpam-6087	177	7	overfitting	overfitte	VERB
ejpam-6087	177	8	,	,	PUNCT
ejpam-6087	177	9	especially	especially	ADV
ejpam-6087	177	10	with	with	ADP
ejpam-6087	177	11	deep	deep	ADJ
ejpam-6087	177	12	trees	tree	NOUN
ejpam-6087	177	13	or	or	CCONJ
ejpam-6087	177	14	small	small	ADJ
ejpam-6087	177	15	datasets	dataset	NOUN
ejpam-6087	177	16	[	[	X
ejpam-6087	177	17	30	30	NUM
ejpam-6087	177	18	,	,	PUNCT
ejpam-6087	177	19	31	31	NUM
ejpam-6087	177	20	]	]	PUNCT
ejpam-6087	177	21	.	.	PUNCT
ejpam-6087	178	1	•	•	NUM
ejpam-6087	178	2	naive	naive	ADJ
ejpam-6087	178	3	bayes	bayes	NOUN
ejpam-6087	178	4	(	(	PUNCT
ejpam-6087	178	5	nb	nb	INTJ
ejpam-6087	178	6	)	)	PUNCT
ejpam-6087	178	7	is	be	AUX
ejpam-6087	178	8	a	a	DET
ejpam-6087	178	9	probabilistic	probabilistic	ADJ
ejpam-6087	178	10	classifier	classifier	NOUN
ejpam-6087	178	11	based	base	VERB
ejpam-6087	178	12	on	on	ADP
ejpam-6087	178	13	bayes	bayes	PROPN
ejpam-6087	178	14	’	'	PUNCT
ejpam-6087	178	15	theorem	theorem	VERB
ejpam-6087	178	16	,	,	PUNCT
ejpam-6087	178	17	often	often	ADV
ejpam-6087	178	18	applied	apply	VERB
ejpam-6087	178	19	to	to	ADP
ejpam-6087	178	20	text	text	NOUN
ejpam-6087	178	21	classification	classification	NOUN
ejpam-6087	178	22	tasks	task	NOUN
ejpam-6087	178	23	.	.	PUNCT
ejpam-6087	179	1	it	it	PRON
ejpam-6087	179	2	assumes	assume	VERB
ejpam-6087	179	3	feature	feature	NOUN
ejpam-6087	179	4	independence	independence	NOUN
ejpam-6087	179	5	given	give	VERB
ejpam-6087	179	6	the	the	DET
ejpam-6087	179	7	class	class	NOUN
ejpam-6087	179	8	,	,	PUNCT
ejpam-6087	179	9	simplifies	simplifie	NOUN
ejpam-6087	179	10	computations	computation	NOUN
ejpam-6087	179	11	,	,	PUNCT
ejpam-6087	179	12	and	and	CCONJ
ejpam-6087	179	13	enables	enable	VERB
ejpam-6087	179	14	efficient	efficient	ADJ
ejpam-6087	179	15	training	training	NOUN
ejpam-6087	179	16	.	.	PUNCT
ejpam-6087	180	1	despite	despite	SCONJ
ejpam-6087	180	2	its	its	PRON
ejpam-6087	180	3	simplicity	simplicity	NOUN
ejpam-6087	180	4	,	,	PUNCT
ejpam-6087	180	5	nb	nb	NOUN
ejpam-6087	180	6	performs	perform	VERB
ejpam-6087	180	7	well	well	ADV
ejpam-6087	180	8	in	in	ADP
ejpam-6087	180	9	areas	area	NOUN
ejpam-6087	180	10	such	such	ADJ
ejpam-6087	180	11	as	as	ADP
ejpam-6087	180	12	spam	spam	NOUN
ejpam-6087	180	13	detection	detection	NOUN
ejpam-6087	180	14	and	and	CCONJ
ejpam-6087	180	15	sentiment	sentiment	NOUN
ejpam-6087	180	16	analysis	analysis	NOUN
ejpam-6087	180	17	,	,	PUNCT
ejpam-6087	180	18	but	but	CCONJ
ejpam-6087	180	19	struggles	struggle	VERB
ejpam-6087	180	20	with	with	ADP
ejpam-6087	180	21	highly	highly	ADV
ejpam-6087	180	22	correlated	correlate	VERB
ejpam-6087	180	23	features	feature	NOUN
ejpam-6087	180	24	or	or	CCONJ
ejpam-6087	180	25	non	non	ADJ
ejpam-6087	180	26	-	-	ADJ
ejpam-6087	180	27	gaussian	gaussian	ADJ
ejpam-6087	180	28	data	datum	NOUN
ejpam-6087	180	29	distributions	distribution	NOUN
ejpam-6087	180	30	[	[	X
ejpam-6087	180	31	32	32	NUM
ejpam-6087	180	32	]	]	PUNCT
ejpam-6087	180	33	.	.	PUNCT
ejpam-6087	181	1	•	•	NOUN
ejpam-6087	181	2	support	support	NOUN
ejpam-6087	181	3	vector	vector	NOUN
ejpam-6087	181	4	machines	machine	NOUN
ejpam-6087	181	5	(	(	PUNCT
ejpam-6087	181	6	svms	svms	NOUN
ejpam-6087	181	7	)	)	PUNCT
ejpam-6087	181	8	are	be	AUX
ejpam-6087	181	9	versatile	versatile	ADJ
ejpam-6087	181	10	algorithms	algorithm	NOUN
ejpam-6087	181	11	that	that	PRON
ejpam-6087	181	12	separate	separate	ADJ
ejpam-6087	181	13	classes	class	NOUN
ejpam-6087	181	14	by	by	ADP
ejpam-6087	181	15	finding	find	VERB
ejpam-6087	181	16	the	the	DET
ejpam-6087	181	17	hyperplane	hyperplane	NOUN
ejpam-6087	181	18	that	that	PRON
ejpam-6087	181	19	maximizes	maximize	VERB
ejpam-6087	181	20	the	the	DET
ejpam-6087	181	21	margin	margin	NOUN
ejpam-6087	181	22	between	between	ADP
ejpam-6087	181	23	them	they	PRON
ejpam-6087	181	24	.	.	PUNCT
ejpam-6087	182	1	svms	svms	NOUN
ejpam-6087	182	2	can	can	AUX
ejpam-6087	182	3	handle	handle	VERB
ejpam-6087	182	4	linear	linear	ADJ
ejpam-6087	182	5	and	and	CCONJ
ejpam-6087	182	6	non	non	ADJ
ejpam-6087	182	7	-	-	ADJ
ejpam-6087	182	8	linear	linear	ADJ
ejpam-6087	182	9	problems	problem	NOUN
ejpam-6087	182	10	using	use	VERB
ejpam-6087	182	11	kernel	kernel	NOUN
ejpam-6087	182	12	functions	function	NOUN
ejpam-6087	182	13	such	such	ADJ
ejpam-6087	182	14	as	as	ADP
ejpam-6087	182	15	the	the	DET
ejpam-6087	182	16	linear	linear	ADJ
ejpam-6087	182	17	,	,	PUNCT
ejpam-6087	182	18	polynomial	polynomial	ADJ
ejpam-6087	182	19	,	,	PUNCT
ejpam-6087	182	20	and	and	CCONJ
ejpam-6087	182	21	radial	radial	ADJ
ejpam-6087	182	22	basis	basis	NOUN
ejpam-6087	182	23	function	function	NOUN
ejpam-6087	182	24	(	(	PUNCT
ejpam-6087	182	25	rbf	rbf	PROPN
ejpam-6087	182	26	)	)	PUNCT
ejpam-6087	182	27	.	.	PUNCT
ejpam-6087	183	1	the	the	DET
ejpam-6087	183	2	regularization	regularization	NOUN
ejpam-6087	183	3	parameter	parameter	NOUN
ejpam-6087	183	4	c	c	PROPN
ejpam-6087	183	5	controls	control	VERB
ejpam-6087	183	6	the	the	DET
ejpam-6087	183	7	tradeoff	tradeoff	NOUN
ejpam-6087	183	8	between	between	ADP
ejpam-6087	183	9	the	the	DET
ejpam-6087	183	10	margin	margin	NOUN
ejpam-6087	183	11	size	size	NOUN
ejpam-6087	183	12	and	and	CCONJ
ejpam-6087	183	13	the	the	DET
ejpam-6087	183	14	classification	classification	NOUN
ejpam-6087	183	15	errors	error	NOUN
ejpam-6087	183	16	[	[	X
ejpam-6087	183	17	33	33	NUM
ejpam-6087	183	18	,	,	PUNCT
ejpam-6087	183	19	34	34	NUM
ejpam-6087	183	20	]	]	PUNCT
ejpam-6087	183	21	.	.	PUNCT
ejpam-6087	184	1	•	•	NUM
ejpam-6087	184	2	k	k	X
ejpam-6087	184	3	-	-	PUNCT
ejpam-6087	184	4	nearest	near	ADJ
ejpam-6087	184	5	neighbors	neighbor	NOUN
ejpam-6087	184	6	(	(	PUNCT
ejpam-6087	184	7	knn	knn	PROPN
ejpam-6087	184	8	)	)	PUNCT
ejpam-6087	184	9	is	be	AUX
ejpam-6087	184	10	a	a	DET
ejpam-6087	184	11	straightforward	straightforward	ADJ
ejpam-6087	184	12	algorithm	algorithm	NOUN
ejpam-6087	184	13	that	that	PRON
ejpam-6087	184	14	assigns	assign	VERB
ejpam-6087	184	15	class	class	NOUN
ejpam-6087	184	16	labels	label	NOUN
ejpam-6087	184	17	based	base	VERB
ejpam-6087	184	18	on	on	ADP
ejpam-6087	184	19	the	the	DET
ejpam-6087	184	20	majority	majority	NOUN
ejpam-6087	184	21	class	class	NOUN
ejpam-6087	184	22	among	among	ADP
ejpam-6087	184	23	the	the	DET
ejpam-6087	184	24	k	k	PROPN
ejpam-6087	184	25	nearest	near	ADJ
ejpam-6087	184	26	neighbors	neighbor	NOUN
ejpam-6087	184	27	.	.	PUNCT
ejpam-6087	185	1	distance	distance	NOUN
ejpam-6087	185	2	metrics	metric	NOUN
ejpam-6087	185	3	such	such	ADJ
ejpam-6087	185	4	as	as	ADP
ejpam-6087	185	5	euclidean	euclidean	ADJ
ejpam-6087	185	6	or	or	CCONJ
ejpam-6087	185	7	manhattan	manhattan	PROPN
ejpam-6087	185	8	are	be	AUX
ejpam-6087	185	9	used	use	VERB
ejpam-6087	185	10	to	to	PART
ejpam-6087	185	11	find	find	VERB
ejpam-6087	185	12	neighbors	neighbor	NOUN
ejpam-6087	185	13	.	.	PUNCT
ejpam-6087	186	1	knn	knn	PROPN
ejpam-6087	186	2	is	be	AUX
ejpam-6087	186	3	effective	effective	ADJ
ejpam-6087	186	4	for	for	ADP
ejpam-6087	186	5	multiclass	multiclass	ADJ
ejpam-6087	186	6	classification	classification	NOUN
ejpam-6087	186	7	,	,	PUNCT
ejpam-6087	186	8	but	but	CCONJ
ejpam-6087	186	9	can	can	AUX
ejpam-6087	186	10	be	be	AUX
ejpam-6087	186	11	computationally	computationally	ADV
ejpam-6087	186	12	expensive	expensive	ADJ
ejpam-6087	186	13	for	for	ADP
ejpam-6087	186	14	large	large	ADJ
ejpam-6087	186	15	datasets	dataset	NOUN
ejpam-6087	186	16	and	and	CCONJ
ejpam-6087	186	17	is	be	AUX
ejpam-6087	186	18	sensitive	sensitive	ADJ
ejpam-6087	186	19	to	to	ADP
ejpam-6087	186	20	values	value	NOUN
ejpam-6087	186	21	of	of	ADP
ejpam-6087	186	22	k	k	PROPN
ejpam-6087	186	23	and	and	CCONJ
ejpam-6087	186	24	the	the	DET
ejpam-6087	186	25	dimensionality	dimensionality	NOUN
ejpam-6087	186	26	of	of	ADP
ejpam-6087	186	27	features	feature	NOUN
ejpam-6087	186	28	[	[	X
ejpam-6087	186	29	31	31	NUM
ejpam-6087	186	30	,	,	PUNCT
ejpam-6087	186	31	35	35	NUM
ejpam-6087	186	32	]	]	PUNCT
ejpam-6087	186	33	.	.	PUNCT
ejpam-6087	187	1	•	•	NUM
ejpam-6087	187	2	random	random	ADJ
ejpam-6087	187	3	forest	forest	NOUN
ejpam-6087	187	4	(	(	PUNCT
ejpam-6087	187	5	rf	rf	NOUN
ejpam-6087	187	6	)	)	PUNCT
ejpam-6087	187	7	builds	build	VERB
ejpam-6087	187	8	multiple	multiple	ADJ
ejpam-6087	187	9	decision	decision	NOUN
ejpam-6087	187	10	trees	tree	NOUN
ejpam-6087	187	11	on	on	ADP
ejpam-6087	187	12	subsets	subset	NOUN
ejpam-6087	187	13	of	of	ADP
ejpam-6087	187	14	random	random	ADJ
ejpam-6087	187	15	data	datum	NOUN
ejpam-6087	187	16	and	and	CCONJ
ejpam-6087	187	17	aggregates	aggregate	VERB
ejpam-6087	187	18	their	their	PRON
ejpam-6087	187	19	predictions	prediction	NOUN
ejpam-6087	187	20	.	.	PUNCT
ejpam-6087	188	1	this	this	DET
ejpam-6087	188	2	ensemble	ensemble	ADJ
ejpam-6087	188	3	approach	approach	NOUN
ejpam-6087	188	4	reduces	reduce	VERB
ejpam-6087	188	5	overfitting	overfitting	NOUN
ejpam-6087	188	6	and	and	CCONJ
ejpam-6087	188	7	variance	variance	NOUN
ejpam-6087	188	8	,	,	PUNCT
ejpam-6087	188	9	making	make	VERB
ejpam-6087	188	10	rf	rf	ADJ
ejpam-6087	188	11	suitable	suitable	ADJ
ejpam-6087	188	12	for	for	ADP
ejpam-6087	188	13	noisy	noisy	ADJ
ejpam-6087	188	14	and	and	CCONJ
ejpam-6087	188	15	high	high	ADJ
ejpam-6087	188	16	-	-	PUNCT
ejpam-6087	188	17	dimensional	dimensional	ADJ
ejpam-6087	188	18	data	datum	NOUN
ejpam-6087	188	19	.	.	PUNCT
ejpam-6087	189	1	it	it	PRON
ejpam-6087	189	2	also	also	ADV
ejpam-6087	189	3	provides	provide	VERB
ejpam-6087	189	4	feature	feature	NOUN
ejpam-6087	189	5	importance	importance	NOUN
ejpam-6087	189	6	scores	score	NOUN
ejpam-6087	189	7	,	,	PUNCT
ejpam-6087	189	8	but	but	CCONJ
ejpam-6087	189	9	can	can	AUX
ejpam-6087	189	10	be	be	AUX
ejpam-6087	189	11	computationally	computationally	ADV
ejpam-6087	189	12	intensive	intensive	ADJ
ejpam-6087	189	13	for	for	ADP
ejpam-6087	189	14	large	large	ADJ
ejpam-6087	189	15	datasets	dataset	NOUN
ejpam-6087	189	16	[	[	X
ejpam-6087	189	17	36	36	NUM
ejpam-6087	189	18	,	,	PUNCT
ejpam-6087	189	19	37	37	NUM
ejpam-6087	189	20	]	]	PUNCT
ejpam-6087	189	21	.	.	PUNCT
ejpam-6087	190	1	•	•	NUM
ejpam-6087	190	2	multilayer	multilayer	PROPN
ejpam-6087	190	3	perceptron	perceptron	PROPN
ejpam-6087	190	4	(	(	PUNCT
ejpam-6087	190	5	mlp	mlp	PROPN
ejpam-6087	190	6	)	)	PUNCT
ejpam-6087	190	7	is	be	AUX
ejpam-6087	190	8	a	a	DET
ejpam-6087	190	9	type	type	NOUN
ejpam-6087	190	10	of	of	ADP
ejpam-6087	190	11	neural	neural	ADJ
ejpam-6087	190	12	network	network	NOUN
ejpam-6087	190	13	with	with	ADP
ejpam-6087	190	14	multiple	multiple	ADJ
ejpam-6087	190	15	layers	layer	NOUN
ejpam-6087	190	16	that	that	PRON
ejpam-6087	190	17	capture	capture	VERB
ejpam-6087	190	18	nonlinear	nonlinear	ADJ
ejpam-6087	190	19	relationships	relationship	NOUN
ejpam-6087	190	20	in	in	ADP
ejpam-6087	190	21	data	datum	NOUN
ejpam-6087	190	22	.	.	PUNCT
ejpam-6087	191	1	using	use	VERB
ejpam-6087	191	2	activation	activation	NOUN
ejpam-6087	191	3	functions	function	NOUN
ejpam-6087	191	4	such	such	ADJ
ejpam-6087	191	5	as	as	ADP
ejpam-6087	191	6	relu	relu	NOUN
ejpam-6087	191	7	or	or	CCONJ
ejpam-6087	191	8	sigmoid	sigmoid	NOUN
ejpam-6087	191	9	,	,	PUNCT
ejpam-6087	191	10	mlps	mlp	NOUN
ejpam-6087	191	11	can	can	AUX
ejpam-6087	191	12	model	model	VERB
ejpam-6087	191	13	complex	complex	ADJ
ejpam-6087	191	14	patterns	pattern	NOUN
ejpam-6087	191	15	.	.	PUNCT
ejpam-6087	192	1	however	however	ADV
ejpam-6087	192	2	,	,	PUNCT
ejpam-6087	192	3	they	they	PRON
ejpam-6087	192	4	require	require	VERB
ejpam-6087	192	5	careful	careful	ADJ
ejpam-6087	192	6	regularization	regularization	NOUN
ejpam-6087	192	7	to	to	PART
ejpam-6087	192	8	avoid	avoid	VERB
ejpam-6087	192	9	overfitting	overfitte	VERB
ejpam-6087	192	10	and	and	CCONJ
ejpam-6087	192	11	can	can	AUX
ejpam-6087	192	12	be	be	AUX
ejpam-6087	192	13	computationally	computationally	ADV
ejpam-6087	192	14	demanding	demand	VERB
ejpam-6087	192	15	[	[	X
ejpam-6087	192	16	38	38	NUM
ejpam-6087	192	17	]	]	PUNCT
ejpam-6087	192	18	.	.	PUNCT
ejpam-6087	193	1	m.	m.	PROPN
ejpam-6087	193	2	al	al	PROPN
ejpam-6087	193	3	-	-	PUNCT
ejpam-6087	193	4	tarawneh	tarawneh	PROPN
ejpam-6087	193	5	et	et	PROPN
ejpam-6087	193	6	al	al	PROPN
ejpam-6087	193	7	.	.	PUNCT
ejpam-6087	193	8	/	/	SYM
ejpam-6087	193	9	eur	eur	PROPN
ejpam-6087	193	10	.	.	PUNCT
ejpam-6087	194	1	j.	j.	PROPN
ejpam-6087	194	2	pure	pure	PROPN
ejpam-6087	194	3	appl	appl	PROPN
ejpam-6087	194	4	.	.	PROPN
ejpam-6087	194	5	math	math	PROPN
ejpam-6087	194	6	,	,	PUNCT
ejpam-6087	194	7	18	18	NUM
ejpam-6087	194	8	(	(	PUNCT
ejpam-6087	194	9	2	2	NUM
ejpam-6087	194	10	)	)	PUNCT
ejpam-6087	194	11	(	(	PUNCT
ejpam-6087	194	12	2025	2025	NUM
ejpam-6087	194	13	)	)	PUNCT
ejpam-6087	194	14	,	,	PUNCT
ejpam-6087	194	15	6087	6087	NUM
ejpam-6087	194	16	11	11	NUM
ejpam-6087	194	17	of	of	ADP
ejpam-6087	194	18	39	39	NUM
ejpam-6087	194	19	3.3.2	3.3.2	NUM
ejpam-6087	194	20	.	.	PUNCT
ejpam-6087	195	1	homogeneous	homogeneous	ADJ
ejpam-6087	195	2	ensemble	ensemble	ADJ
ejpam-6087	195	3	classifiers	classifier	NOUN
ejpam-6087	195	4	homogeneous	homogeneous	ADJ
ejpam-6087	195	5	ensembles	ensemble	NOUN
ejpam-6087	195	6	combine	combine	VERB
ejpam-6087	195	7	multiple	multiple	ADJ
ejpam-6087	195	8	models	model	NOUN
ejpam-6087	195	9	of	of	ADP
ejpam-6087	195	10	the	the	DET
ejpam-6087	195	11	same	same	ADJ
ejpam-6087	195	12	type	type	NOUN
ejpam-6087	195	13	to	to	PART
ejpam-6087	195	14	improve	improve	VERB
ejpam-6087	195	15	predictive	predictive	ADJ
ejpam-6087	195	16	accuracy	accuracy	NOUN
ejpam-6087	195	17	and	and	CCONJ
ejpam-6087	195	18	reduce	reduce	VERB
ejpam-6087	195	19	variance	variance	NOUN
ejpam-6087	195	20	.	.	PUNCT
ejpam-6087	196	1	these	these	DET
ejpam-6087	196	2	methods	method	NOUN
ejpam-6087	196	3	utilize	utilize	VERB
ejpam-6087	196	4	the	the	DET
ejpam-6087	196	5	strengths	strength	NOUN
ejpam-6087	196	6	of	of	ADP
ejpam-6087	196	7	individual	individual	ADJ
ejpam-6087	196	8	models	model	NOUN
ejpam-6087	196	9	to	to	PART
ejpam-6087	196	10	enhance	enhance	VERB
ejpam-6087	196	11	stability	stability	NOUN
ejpam-6087	196	12	and	and	CCONJ
ejpam-6087	196	13	reliability	reliability	NOUN
ejpam-6087	196	14	.	.	PUNCT
ejpam-6087	197	1	the	the	DET
ejpam-6087	197	2	following	follow	VERB
ejpam-6087	197	3	are	be	AUX
ejpam-6087	197	4	key	key	ADJ
ejpam-6087	197	5	homogeneous	homogeneous	ADJ
ejpam-6087	197	6	ensemble	ensemble	ADJ
ejpam-6087	197	7	techniques	technique	NOUN
ejpam-6087	197	8	used	use	VERB
ejpam-6087	197	9	in	in	ADP
ejpam-6087	197	10	this	this	DET
ejpam-6087	197	11	study	study	NOUN
ejpam-6087	197	12	[	[	X
ejpam-6087	197	13	28	28	NUM
ejpam-6087	197	14	,	,	PUNCT
ejpam-6087	197	15	35	35	NUM
ejpam-6087	197	16	]	]	PUNCT
ejpam-6087	197	17	.	.	PUNCT
ejpam-6087	198	1	•	•	NOUN
ejpam-6087	198	2	bagging	bagging	NOUN
ejpam-6087	198	3	(	(	PUNCT
ejpam-6087	198	4	bootstrap	bootstrap	NOUN
ejpam-6087	198	5	aggregating	aggregating	NOUN
ejpam-6087	198	6	)	)	PUNCT
ejpam-6087	198	7	creates	create	VERB
ejpam-6087	198	8	diverse	diverse	ADJ
ejpam-6087	198	9	training	training	NOUN
ejpam-6087	198	10	sets	set	NOUN
ejpam-6087	198	11	by	by	ADP
ejpam-6087	198	12	bootstrapping	bootstrappe	VERB
ejpam-6087	198	13	and	and	CCONJ
ejpam-6087	198	14	trains	train	VERB
ejpam-6087	198	15	multiple	multiple	ADJ
ejpam-6087	198	16	instances	instance	NOUN
ejpam-6087	198	17	of	of	ADP
ejpam-6087	198	18	a	a	DET
ejpam-6087	198	19	base	base	NOUN
ejpam-6087	198	20	model	model	NOUN
ejpam-6087	198	21	.	.	PUNCT
ejpam-6087	199	1	the	the	DET
ejpam-6087	199	2	final	final	ADJ
ejpam-6087	199	3	prediction	prediction	NOUN
ejpam-6087	199	4	is	be	AUX
ejpam-6087	199	5	determined	determine	VERB
ejpam-6087	199	6	by	by	ADP
ejpam-6087	199	7	aggregating	aggregate	VERB
ejpam-6087	199	8	the	the	DET
ejpam-6087	199	9	output	output	NOUN
ejpam-6087	199	10	,	,	PUNCT
ejpam-6087	199	11	typically	typically	ADV
ejpam-6087	199	12	using	use	VERB
ejpam-6087	199	13	majority	majority	NOUN
ejpam-6087	199	14	voting	voting	NOUN
ejpam-6087	199	15	.	.	PUNCT
ejpam-6087	200	1	bagging	bagging	NOUN
ejpam-6087	200	2	is	be	AUX
ejpam-6087	200	3	particularly	particularly	ADV
ejpam-6087	200	4	effective	effective	ADJ
ejpam-6087	200	5	in	in	ADP
ejpam-6087	200	6	reducing	reduce	VERB
ejpam-6087	200	7	overfitting	overfitte	VERB
ejpam-6087	200	8	in	in	ADP
ejpam-6087	200	9	models	model	NOUN
ejpam-6087	200	10	prone	prone	ADJ
ejpam-6087	200	11	to	to	ADP
ejpam-6087	200	12	high	high	ADJ
ejpam-6087	200	13	variance	variance	NOUN
ejpam-6087	200	14	,	,	PUNCT
ejpam-6087	200	15	such	such	ADJ
ejpam-6087	200	16	as	as	ADP
ejpam-6087	200	17	decision	decision	NOUN
ejpam-6087	200	18	trees	tree	NOUN
ejpam-6087	200	19	[	[	X
ejpam-6087	200	20	27	27	NUM
ejpam-6087	200	21	]	]	PUNCT
ejpam-6087	200	22	.	.	PUNCT
ejpam-6087	201	1	•	•	NOUN
ejpam-6087	201	2	adaboost	adaboost	ADV
ejpam-6087	201	3	(	(	PUNCT
ejpam-6087	201	4	adaptive	adaptive	ADJ
ejpam-6087	201	5	boosting	boosting	NOUN
ejpam-6087	201	6	)	)	PUNCT
ejpam-6087	201	7	builds	build	VERB
ejpam-6087	201	8	a	a	DET
ejpam-6087	201	9	sequence	sequence	NOUN
ejpam-6087	201	10	of	of	ADP
ejpam-6087	201	11	models	model	NOUN
ejpam-6087	201	12	where	where	SCONJ
ejpam-6087	201	13	each	each	DET
ejpam-6087	201	14	subsequent	subsequent	ADJ
ejpam-6087	201	15	model	model	NOUN
ejpam-6087	201	16	focuses	focus	VERB
ejpam-6087	201	17	on	on	ADP
ejpam-6087	201	18	correcting	correct	VERB
ejpam-6087	201	19	errors	error	NOUN
ejpam-6087	201	20	made	make	VERB
ejpam-6087	201	21	by	by	ADP
ejpam-6087	201	22	its	its	PRON
ejpam-6087	201	23	predecessors	predecessor	NOUN
ejpam-6087	201	24	.	.	PUNCT
ejpam-6087	202	1	the	the	DET
ejpam-6087	202	2	final	final	ADJ
ejpam-6087	202	3	predictions	prediction	NOUN
ejpam-6087	202	4	are	be	AUX
ejpam-6087	202	5	weighted	weight	VERB
ejpam-6087	202	6	on	on	ADP
ejpam-6087	202	7	the	the	DET
ejpam-6087	202	8	basis	basis	NOUN
ejpam-6087	202	9	of	of	ADP
ejpam-6087	202	10	the	the	DET
ejpam-6087	202	11	models	model	NOUN
ejpam-6087	202	12	’	'	PUNCT
ejpam-6087	202	13	performance	performance	NOUN
ejpam-6087	202	14	.	.	PUNCT
ejpam-6087	203	1	adaboost	adaboost	ADV
ejpam-6087	203	2	improves	improve	VERB
ejpam-6087	203	3	weak	weak	ADJ
ejpam-6087	203	4	learners	learner	NOUN
ejpam-6087	203	5	,	,	PUNCT
ejpam-6087	203	6	but	but	CCONJ
ejpam-6087	203	7	can	can	AUX
ejpam-6087	203	8	be	be	AUX
ejpam-6087	203	9	sensitive	sensitive	ADJ
ejpam-6087	203	10	to	to	ADP
ejpam-6087	203	11	noisy	noisy	ADJ
ejpam-6087	203	12	data	datum	NOUN
ejpam-6087	203	13	and	and	CCONJ
ejpam-6087	203	14	class	class	NOUN
ejpam-6087	203	15	imbalances	imbalance	NOUN
ejpam-6087	203	16	[	[	X
ejpam-6087	203	17	35	35	NUM
ejpam-6087	203	18	]	]	PUNCT
ejpam-6087	203	19	.	.	PUNCT
ejpam-6087	204	1	•	•	NUM
ejpam-6087	204	2	gradient	gradient	NOUN
ejpam-6087	204	3	boosting	boost	VERB
ejpam-6087	204	4	iteratively	iteratively	ADV
ejpam-6087	204	5	trains	train	NOUN
ejpam-6087	204	6	models	model	NOUN
ejpam-6087	204	7	to	to	PART
ejpam-6087	204	8	minimize	minimize	VERB
ejpam-6087	204	9	residual	residual	ADJ
ejpam-6087	204	10	errors	error	NOUN
ejpam-6087	204	11	from	from	ADP
ejpam-6087	204	12	previous	previous	ADJ
ejpam-6087	204	13	models	model	NOUN
ejpam-6087	204	14	.	.	PUNCT
ejpam-6087	205	1	the	the	DET
ejpam-6087	205	2	final	final	ADJ
ejpam-6087	205	3	output	output	NOUN
ejpam-6087	205	4	combines	combine	VERB
ejpam-6087	205	5	all	all	DET
ejpam-6087	205	6	the	the	DET
ejpam-6087	205	7	models	model	NOUN
ejpam-6087	205	8	’	'	PUNCT
ejpam-6087	205	9	predictions	prediction	NOUN
ejpam-6087	205	10	.	.	PUNCT
ejpam-6087	206	1	this	this	DET
ejpam-6087	206	2	technique	technique	NOUN
ejpam-6087	206	3	is	be	AUX
ejpam-6087	206	4	highly	highly	ADV
ejpam-6087	206	5	flexible	flexible	ADJ
ejpam-6087	206	6	,	,	PUNCT
ejpam-6087	206	7	supports	support	VERB
ejpam-6087	206	8	various	various	ADJ
ejpam-6087	206	9	data	datum	NOUN
ejpam-6087	206	10	types	type	NOUN
ejpam-6087	206	11	,	,	PUNCT
ejpam-6087	206	12	and	and	CCONJ
ejpam-6087	206	13	provides	provide	VERB
ejpam-6087	206	14	insight	insight	NOUN
ejpam-6087	206	15	into	into	ADP
ejpam-6087	206	16	the	the	DET
ejpam-6087	206	17	importance	importance	NOUN
ejpam-6087	206	18	of	of	ADP
ejpam-6087	206	19	features	feature	NOUN
ejpam-6087	206	20	[	[	X
ejpam-6087	206	21	27	27	NUM
ejpam-6087	206	22	,	,	PUNCT
ejpam-6087	206	23	28	28	NUM
ejpam-6087	206	24	]	]	PUNCT
ejpam-6087	206	25	.	.	PUNCT
ejpam-6087	207	1	•	•	NUM
ejpam-6087	207	2	xgboost	xgboost	NOUN
ejpam-6087	207	3	(	(	PUNCT
ejpam-6087	207	4	extreme	extreme	ADJ
ejpam-6087	207	5	gradient	gradient	NOUN
ejpam-6087	207	6	boosting	boosting	NOUN
ejpam-6087	207	7	)	)	PUNCT
ejpam-6087	207	8	improves	improve	VERB
ejpam-6087	207	9	traditional	traditional	ADJ
ejpam-6087	207	10	gradient	gradient	NOUN
ejpam-6087	207	11	boosting	boost	VERB
ejpam-6087	207	12	with	with	ADP
ejpam-6087	207	13	optimizations	optimization	NOUN
ejpam-6087	207	14	such	such	ADJ
ejpam-6087	207	15	as	as	ADP
ejpam-6087	207	16	parallel	parallel	ADJ
ejpam-6087	207	17	processing	processing	NOUN
ejpam-6087	207	18	and	and	CCONJ
ejpam-6087	207	19	regularization	regularization	NOUN
ejpam-6087	207	20	,	,	PUNCT
ejpam-6087	207	21	making	make	VERB
ejpam-6087	207	22	it	it	PRON
ejpam-6087	207	23	computationally	computationally	ADV
ejpam-6087	207	24	efficient	efficient	ADJ
ejpam-6087	207	25	and	and	CCONJ
ejpam-6087	207	26	highly	highly	ADV
ejpam-6087	207	27	effective	effective	ADJ
ejpam-6087	207	28	for	for	ADP
ejpam-6087	207	29	complex	complex	ADJ
ejpam-6087	207	30	datasets	dataset	NOUN
ejpam-6087	207	31	[	[	X
ejpam-6087	207	32	39	39	NUM
ejpam-6087	207	33	]	]	PUNCT
ejpam-6087	207	34	.	.	PUNCT
ejpam-6087	208	1	•	•	NUM
ejpam-6087	208	2	extra	extra	ADJ
ejpam-6087	208	3	trees	tree	NOUN
ejpam-6087	208	4	generates	generate	VERB
ejpam-6087	208	5	an	an	DET
ejpam-6087	208	6	ensemble	ensemble	NOUN
ejpam-6087	208	7	of	of	ADP
ejpam-6087	208	8	extremely	extremely	ADV
ejpam-6087	208	9	randomized	randomized	ADJ
ejpam-6087	208	10	decision	decision	NOUN
ejpam-6087	208	11	trees	tree	NOUN
ejpam-6087	208	12	.	.	PUNCT
ejpam-6087	209	1	introduces	introduce	NOUN
ejpam-6087	209	2	randomness	randomness	NOUN
ejpam-6087	209	3	in	in	ADP
ejpam-6087	209	4	feature	feature	NOUN
ejpam-6087	209	5	selection	selection	NOUN
ejpam-6087	209	6	and	and	CCONJ
ejpam-6087	209	7	split	split	NOUN
ejpam-6087	209	8	point	point	NOUN
ejpam-6087	209	9	determination	determination	NOUN
ejpam-6087	209	10	,	,	PUNCT
ejpam-6087	209	11	improving	improve	VERB
ejpam-6087	209	12	generalization	generalization	NOUN
ejpam-6087	209	13	and	and	CCONJ
ejpam-6087	209	14	efficiency	efficiency	NOUN
ejpam-6087	209	15	[	[	X
ejpam-6087	209	16	39	39	NUM
ejpam-6087	209	17	,	,	PUNCT
ejpam-6087	209	18	40	40	NUM
ejpam-6087	209	19	]	]	PUNCT
ejpam-6087	209	20	.	.	PUNCT
ejpam-6087	210	1	•	•	NUM
ejpam-6087	210	2	catboost	catboost	NOUN
ejpam-6087	210	3	specializes	specialize	VERB
ejpam-6087	210	4	in	in	ADP
ejpam-6087	210	5	handling	handle	VERB
ejpam-6087	210	6	categorical	categorical	ADJ
ejpam-6087	210	7	data	datum	NOUN
ejpam-6087	210	8	through	through	ADP
ejpam-6087	210	9	efficient	efficient	ADJ
ejpam-6087	210	10	encoding	encoding	NOUN
ejpam-6087	210	11	techniques	technique	NOUN
ejpam-6087	210	12	.	.	PUNCT
ejpam-6087	211	1	it	it	PRON
ejpam-6087	211	2	includes	include	VERB
ejpam-6087	211	3	mechanisms	mechanism	NOUN
ejpam-6087	211	4	to	to	PART
ejpam-6087	211	5	prevent	prevent	VERB
ejpam-6087	211	6	overfitting	overfitting	NOUN
ejpam-6087	211	7	and	and	CCONJ
ejpam-6087	211	8	supports	support	VERB
ejpam-6087	211	9	gpu	gpu	NOUN
ejpam-6087	211	10	acceleration	acceleration	NOUN
ejpam-6087	211	11	,	,	PUNCT
ejpam-6087	211	12	making	make	VERB
ejpam-6087	211	13	it	it	PRON
ejpam-6087	211	14	suitable	suitable	ADJ
ejpam-6087	211	15	for	for	ADP
ejpam-6087	211	16	diverse	diverse	ADJ
ejpam-6087	211	17	datasets	dataset	NOUN
ejpam-6087	211	18	.	.	PUNCT
ejpam-6087	212	1	•	•	NUM
ejpam-6087	212	2	hist	hist	PROPN
ejpam-6087	212	3	gradient	gradient	NOUN
ejpam-6087	212	4	boosting	boost	VERB
ejpam-6087	212	5	employs	employ	VERB
ejpam-6087	212	6	histogram	histogram	NOUN
ejpam-6087	212	7	-	-	PUNCT
ejpam-6087	212	8	based	base	VERB
ejpam-6087	212	9	decision	decision	NOUN
ejpam-6087	212	10	trees	tree	NOUN
ejpam-6087	212	11	to	to	PART
ejpam-6087	212	12	improve	improve	VERB
ejpam-6087	212	13	computational	computational	ADJ
ejpam-6087	212	14	efficiency	efficiency	NOUN
ejpam-6087	212	15	by	by	ADP
ejpam-6087	212	16	approximating	approximate	VERB
ejpam-6087	212	17	continuous	continuous	ADJ
ejpam-6087	212	18	values	value	NOUN
ejpam-6087	212	19	with	with	ADP
ejpam-6087	212	20	histograms	histogram	NOUN
ejpam-6087	212	21	.	.	PUNCT
ejpam-6087	213	1	it	it	PRON
ejpam-6087	213	2	is	be	AUX
ejpam-6087	213	3	particularly	particularly	ADV
ejpam-6087	213	4	useful	useful	ADJ
ejpam-6087	213	5	for	for	ADP
ejpam-6087	213	6	large	large	ADJ
ejpam-6087	213	7	-	-	PUNCT
ejpam-6087	213	8	scale	scale	NOUN
ejpam-6087	213	9	problems	problem	NOUN
ejpam-6087	213	10	[	[	X
ejpam-6087	213	11	27	27	NUM
ejpam-6087	213	12	,	,	PUNCT
ejpam-6087	213	13	35	35	NUM
ejpam-6087	213	14	]	]	PUNCT
ejpam-6087	213	15	.	.	PUNCT
ejpam-6087	214	1	•	•	NOUN
ejpam-6087	214	2	isolation	isolation	NOUN
ejpam-6087	214	3	forests	forest	NOUN
ejpam-6087	214	4	focus	focus	VERB
ejpam-6087	214	5	on	on	ADP
ejpam-6087	214	6	identifying	identify	VERB
ejpam-6087	214	7	anomalies	anomaly	NOUN
ejpam-6087	214	8	by	by	ADP
ejpam-6087	214	9	isolating	isolate	VERB
ejpam-6087	214	10	outliers	outlier	NOUN
ejpam-6087	214	11	using	use	VERB
ejpam-6087	214	12	random	random	ADJ
ejpam-6087	214	13	feature	feature	NOUN
ejpam-6087	214	14	splits	split	NOUN
ejpam-6087	214	15	.	.	PUNCT
ejpam-6087	215	1	shorter	short	ADJ
ejpam-6087	215	2	isolation	isolation	NOUN
ejpam-6087	215	3	paths	path	NOUN
ejpam-6087	215	4	indicate	indicate	VERB
ejpam-6087	215	5	potential	potential	ADJ
ejpam-6087	215	6	anomalies	anomaly	NOUN
ejpam-6087	215	7	,	,	PUNCT
ejpam-6087	215	8	making	make	VERB
ejpam-6087	215	9	this	this	DET
ejpam-6087	215	10	method	method	NOUN
ejpam-6087	215	11	valuable	valuable	ADJ
ejpam-6087	215	12	for	for	ADP
ejpam-6087	215	13	detecting	detect	VERB
ejpam-6087	215	14	rare	rare	ADJ
ejpam-6087	215	15	patterns	pattern	NOUN
ejpam-6087	215	16	in	in	ADP
ejpam-6087	215	17	high	high	ADJ
ejpam-6087	215	18	-	-	PUNCT
ejpam-6087	215	19	dimensional	dimensional	ADJ
ejpam-6087	215	20	data	datum	NOUN
ejpam-6087	215	21	[	[	X
ejpam-6087	215	22	41	41	NUM
ejpam-6087	215	23	,	,	PUNCT
ejpam-6087	215	24	42	42	NUM
ejpam-6087	215	25	]	]	PUNCT
ejpam-6087	215	26	.	.	PUNCT
ejpam-6087	216	1	3.3.3	3.3.3	NUM
ejpam-6087	216	2	.	.	PUNCT
ejpam-6087	216	3	heterogeneous	heterogeneous	ADJ
ejpam-6087	216	4	ensemble	ensemble	ADJ
ejpam-6087	216	5	classifiers	classifier	NOUN
ejpam-6087	216	6	heterogeneous	heterogeneous	ADJ
ejpam-6087	216	7	ensembles	ensemble	NOUN
ejpam-6087	216	8	combine	combine	VERB
ejpam-6087	216	9	diverse	diverse	ADJ
ejpam-6087	216	10	classifiers	classifier	NOUN
ejpam-6087	216	11	to	to	PART
ejpam-6087	216	12	leverage	leverage	VERB
ejpam-6087	216	13	their	their	PRON
ejpam-6087	216	14	complementary	complementary	ADJ
ejpam-6087	216	15	strengths	strength	NOUN
ejpam-6087	216	16	,	,	PUNCT
ejpam-6087	216	17	improving	improve	VERB
ejpam-6087	216	18	the	the	DET
ejpam-6087	216	19	accuracy	accuracy	NOUN
ejpam-6087	216	20	and	and	CCONJ
ejpam-6087	216	21	robustness	robustness	NOUN
ejpam-6087	216	22	of	of	ADP
ejpam-6087	216	23	predictions	prediction	NOUN
ejpam-6087	216	24	.	.	PUNCT
ejpam-6087	217	1	the	the	DET
ejpam-6087	217	2	key	key	ADJ
ejpam-6087	217	3	methods	method	NOUN
ejpam-6087	217	4	explored	explore	VERB
ejpam-6087	217	5	in	in	ADP
ejpam-6087	217	6	this	this	DET
ejpam-6087	217	7	study	study	NOUN
ejpam-6087	217	8	include	include	VERB
ejpam-6087	217	9	stacking	stack	VERB
ejpam-6087	217	10	and	and	CCONJ
ejpam-6087	217	11	voting	voting	NOUN
ejpam-6087	217	12	,	,	PUNCT
ejpam-6087	217	13	both	both	PRON
ejpam-6087	217	14	of	of	ADP
ejpam-6087	217	15	which	which	PRON
ejpam-6087	217	16	are	be	AUX
ejpam-6087	217	17	effective	effective	ADJ
ejpam-6087	217	18	in	in	ADP
ejpam-6087	217	19	detecting	detect	VERB
ejpam-6087	217	20	fake	fake	ADJ
ejpam-6087	217	21	news	news	NOUN
ejpam-6087	217	22	.	.	PUNCT
ejpam-6087	218	1	m.	m.	PROPN
ejpam-6087	218	2	al	al	PROPN
ejpam-6087	218	3	-	-	PUNCT
ejpam-6087	218	4	tarawneh	tarawneh	PROPN
ejpam-6087	218	5	et	et	PROPN
ejpam-6087	218	6	al	al	PROPN
ejpam-6087	218	7	.	.	PUNCT
ejpam-6087	218	8	/	/	SYM
ejpam-6087	218	9	eur	eur	PROPN
ejpam-6087	218	10	.	.	PUNCT
ejpam-6087	219	1	j.	j.	PROPN
ejpam-6087	219	2	pure	pure	PROPN
ejpam-6087	219	3	appl	appl	PROPN
ejpam-6087	219	4	.	.	PROPN
ejpam-6087	219	5	math	math	PROPN
ejpam-6087	219	6	,	,	PUNCT
ejpam-6087	219	7	18	18	NUM
ejpam-6087	219	8	(	(	PUNCT
ejpam-6087	219	9	2	2	NUM
ejpam-6087	219	10	)	)	PUNCT
ejpam-6087	219	11	(	(	PUNCT
ejpam-6087	219	12	2025	2025	NUM
ejpam-6087	219	13	)	)	PUNCT
ejpam-6087	219	14	,	,	PUNCT
ejpam-6087	219	15	6087	6087	NUM
ejpam-6087	219	16	12	12	NUM
ejpam-6087	219	17	of	of	ADP
ejpam-6087	219	18	39	39	NUM
ejpam-6087	219	19	•	•	NOUN
ejpam-6087	219	20	voting	voting	NOUN
ejpam-6087	219	21	classifiers	classifier	NOUN
ejpam-6087	219	22	aggregate	aggregate	VERB
ejpam-6087	219	23	predictions	prediction	NOUN
ejpam-6087	219	24	from	from	ADP
ejpam-6087	219	25	different	different	ADJ
ejpam-6087	219	26	base	base	NOUN
ejpam-6087	219	27	models	model	NOUN
ejpam-6087	219	28	to	to	PART
ejpam-6087	219	29	create	create	VERB
ejpam-6087	219	30	a	a	DET
ejpam-6087	219	31	unified	unified	ADJ
ejpam-6087	219	32	output	output	NOUN
ejpam-6087	219	33	.	.	PUNCT
ejpam-6087	220	1	hard	hard	ADJ
ejpam-6087	220	2	voting	voting	NOUN
ejpam-6087	220	3	relies	rely	VERB
ejpam-6087	220	4	on	on	ADP
ejpam-6087	220	5	majority	majority	NOUN
ejpam-6087	220	6	class	class	NOUN
ejpam-6087	220	7	votes	vote	NOUN
ejpam-6087	220	8	,	,	PUNCT
ejpam-6087	220	9	while	while	SCONJ
ejpam-6087	220	10	soft	soft	ADJ
ejpam-6087	220	11	voting	voting	NOUN
ejpam-6087	220	12	uses	use	VERB
ejpam-6087	220	13	weighted	weight	VERB
ejpam-6087	220	14	probabilities	probability	NOUN
ejpam-6087	220	15	based	base	VERB
ejpam-6087	220	16	on	on	ADP
ejpam-6087	220	17	model	model	NOUN
ejpam-6087	220	18	performance	performance	NOUN
ejpam-6087	220	19	.	.	PUNCT
ejpam-6087	221	1	voting	voting	NOUN
ejpam-6087	221	2	classifiers	classifier	NOUN
ejpam-6087	221	3	benefit	benefit	VERB
ejpam-6087	221	4	from	from	ADP
ejpam-6087	221	5	the	the	DET
ejpam-6087	221	6	diversity	diversity	NOUN
ejpam-6087	221	7	of	of	ADP
ejpam-6087	221	8	base	base	NOUN
ejpam-6087	221	9	models	model	NOUN
ejpam-6087	221	10	,	,	PUNCT
ejpam-6087	221	11	which	which	PRON
ejpam-6087	221	12	enhances	enhance	VERB
ejpam-6087	221	13	overall	overall	ADJ
ejpam-6087	221	14	predictive	predictive	ADJ
ejpam-6087	221	15	performance	performance	NOUN
ejpam-6087	221	16	[	[	X
ejpam-6087	221	17	34	34	NUM
ejpam-6087	221	18	]	]	PUNCT
ejpam-6087	221	19	.	.	PUNCT
ejpam-6087	222	1	•	•	NOUN
ejpam-6087	222	2	stacking	stack	VERB
ejpam-6087	222	3	classifiers	classifier	NOUN
ejpam-6087	222	4	employ	employ	VERB
ejpam-6087	222	5	a	a	DET
ejpam-6087	222	6	meta	meta	ADJ
ejpam-6087	222	7	-	-	PUNCT
ejpam-6087	222	8	learner	learner	NOUN
ejpam-6087	222	9	to	to	PART
ejpam-6087	222	10	combine	combine	VERB
ejpam-6087	222	11	predictions	prediction	NOUN
ejpam-6087	222	12	from	from	ADP
ejpam-6087	222	13	multiple	multiple	ADJ
ejpam-6087	222	14	base	base	NOUN
ejpam-6087	222	15	models	model	NOUN
ejpam-6087	222	16	.	.	PUNCT
ejpam-6087	223	1	unlike	unlike	ADP
ejpam-6087	223	2	voting	voting	NOUN
ejpam-6087	223	3	,	,	PUNCT
ejpam-6087	223	4	stacking	stack	VERB
ejpam-6087	223	5	uses	use	VERB
ejpam-6087	223	6	a	a	DET
ejpam-6087	223	7	data	data	NOUN
ejpam-6087	223	8	-	-	PUNCT
ejpam-6087	223	9	driven	drive	VERB
ejpam-6087	223	10	approach	approach	NOUN
ejpam-6087	223	11	to	to	PART
ejpam-6087	223	12	learn	learn	VERB
ejpam-6087	223	13	the	the	DET
ejpam-6087	223	14	optimal	optimal	ADJ
ejpam-6087	223	15	combination	combination	NOUN
ejpam-6087	223	16	of	of	ADP
ejpam-6087	223	17	models	model	NOUN
ejpam-6087	223	18	output	output	VERB
ejpam-6087	223	19	.	.	PUNCT
ejpam-6087	224	1	using	use	VERB
ejpam-6087	224	2	the	the	DET
ejpam-6087	224	3	strengths	strength	NOUN
ejpam-6087	224	4	of	of	ADP
ejpam-6087	224	5	various	various	ADJ
ejpam-6087	224	6	models	model	NOUN
ejpam-6087	224	7	,	,	PUNCT
ejpam-6087	224	8	stacking	stack	VERB
ejpam-6087	224	9	often	often	ADV
ejpam-6087	224	10	achieves	achieve	VERB
ejpam-6087	224	11	superior	superior	ADJ
ejpam-6087	224	12	performance	performance	NOUN
ejpam-6087	224	13	compared	compare	VERB
ejpam-6087	224	14	to	to	ADP
ejpam-6087	224	15	individual	individual	ADJ
ejpam-6087	224	16	models	model	NOUN
ejpam-6087	224	17	or	or	CCONJ
ejpam-6087	224	18	simple	simple	ADJ
ejpam-6087	224	19	voting	voting	NOUN
ejpam-6087	224	20	schemes	scheme	NOUN
ejpam-6087	224	21	[	[	X
ejpam-6087	224	22	34	34	NUM
ejpam-6087	224	23	,	,	PUNCT
ejpam-6087	224	24	35	35	NUM
ejpam-6087	224	25	]	]	PUNCT
ejpam-6087	224	26	.	.	PUNCT
ejpam-6087	225	1	3.4	3.4	NUM
ejpam-6087	225	2	.	.	PUNCT
ejpam-6087	225	3	feature	feature	NOUN
ejpam-6087	225	4	selection	selection	NOUN
ejpam-6087	225	5	methods	method	NOUN
ejpam-6087	225	6	feature	feature	NOUN
ejpam-6087	225	7	selection	selection	NOUN
ejpam-6087	225	8	is	be	AUX
ejpam-6087	225	9	a	a	DET
ejpam-6087	225	10	crucial	crucial	ADJ
ejpam-6087	225	11	step	step	NOUN
ejpam-6087	225	12	in	in	ADP
ejpam-6087	225	13	machine	machine	NOUN
ejpam-6087	225	14	learning	learning	NOUN
ejpam-6087	225	15	that	that	PRON
ejpam-6087	225	16	aims	aim	VERB
ejpam-6087	225	17	to	to	PART
ejpam-6087	225	18	identify	identify	VERB
ejpam-6087	225	19	the	the	DET
ejpam-6087	225	20	most	most	ADV
ejpam-6087	225	21	relevant	relevant	ADJ
ejpam-6087	225	22	features	feature	NOUN
ejpam-6087	225	23	for	for	ADP
ejpam-6087	225	24	model	model	NOUN
ejpam-6087	225	25	building	building	NOUN
ejpam-6087	225	26	.	.	PUNCT
ejpam-6087	226	1	effective	effective	ADJ
ejpam-6087	226	2	feature	feature	NOUN
ejpam-6087	226	3	selection	selection	NOUN
ejpam-6087	226	4	not	not	PART
ejpam-6087	226	5	only	only	ADV
ejpam-6087	226	6	improves	improve	VERB
ejpam-6087	226	7	model	model	NOUN
ejpam-6087	226	8	performance	performance	NOUN
ejpam-6087	226	9	but	but	CCONJ
ejpam-6087	226	10	also	also	ADV
ejpam-6087	226	11	reduces	reduce	VERB
ejpam-6087	226	12	computational	computational	ADJ
ejpam-6087	226	13	cost	cost	NOUN
ejpam-6087	226	14	and	and	CCONJ
ejpam-6087	226	15	helps	help	VERB
ejpam-6087	226	16	prevent	prevent	VERB
ejpam-6087	226	17	overfitting	overfitting	NOUN
ejpam-6087	226	18	.	.	PUNCT
ejpam-6087	227	1	in	in	ADP
ejpam-6087	227	2	this	this	DET
ejpam-6087	227	3	context	context	NOUN
ejpam-6087	227	4	,	,	PUNCT
ejpam-6087	227	5	several	several	ADJ
ejpam-6087	227	6	techniques	technique	NOUN
ejpam-6087	227	7	can	can	AUX
ejpam-6087	227	8	be	be	AUX
ejpam-6087	227	9	employed	employ	VERB
ejpam-6087	227	10	to	to	PART
ejpam-6087	227	11	select	select	VERB
ejpam-6087	227	12	the	the	DET
ejpam-6087	227	13	most	most	ADV
ejpam-6087	227	14	informative	informative	ADJ
ejpam-6087	227	15	features	feature	NOUN
ejpam-6087	227	16	.	.	PUNCT
ejpam-6087	228	1	three	three	NUM
ejpam-6087	228	2	prominent	prominent	ADJ
ejpam-6087	228	3	methods	method	NOUN
ejpam-6087	228	4	are	be	AUX
ejpam-6087	228	5	recursive	recursive	ADJ
ejpam-6087	228	6	feature	feature	NOUN
ejpam-6087	228	7	elimination	elimination	NOUN
ejpam-6087	228	8	(	(	PUNCT
ejpam-6087	228	9	rfe	rfe	NOUN
ejpam-6087	228	10	)	)	PUNCT
ejpam-6087	228	11	,	,	PUNCT
ejpam-6087	228	12	selectkbest	selectkb	ADJ
ejpam-6087	228	13	,	,	PUNCT
ejpam-6087	228	14	and	and	CCONJ
ejpam-6087	228	15	principal	principal	ADJ
ejpam-6087	228	16	component	component	NOUN
ejpam-6087	228	17	analysis	analysis	NOUN
ejpam-6087	228	18	(	(	PUNCT
ejpam-6087	228	19	pca	pca	NOUN
ejpam-6087	228	20	)	)	PUNCT
ejpam-6087	228	21	.	.	PUNCT
ejpam-6087	229	1	each	each	PRON
ejpam-6087	229	2	of	of	ADP
ejpam-6087	229	3	these	these	DET
ejpam-6087	229	4	methods	method	NOUN
ejpam-6087	229	5	utilizes	utilize	VERB
ejpam-6087	229	6	distinct	distinct	ADJ
ejpam-6087	229	7	principles	principle	NOUN
ejpam-6087	229	8	and	and	CCONJ
ejpam-6087	229	9	mathematical	mathematical	ADJ
ejpam-6087	229	10	foundations	foundation	NOUN
ejpam-6087	229	11	to	to	PART
ejpam-6087	229	12	achieve	achieve	VERB
ejpam-6087	229	13	feature	feature	NOUN
ejpam-6087	229	14	selection	selection	NOUN
ejpam-6087	229	15	.	.	PUNCT
ejpam-6087	230	1	3.4.1	3.4.1	X
ejpam-6087	230	2	.	.	PUNCT
ejpam-6087	230	3	recursive	recursive	ADJ
ejpam-6087	230	4	feature	feature	NOUN
ejpam-6087	230	5	elimination	elimination	NOUN
ejpam-6087	230	6	rfe	rfe	NOUN
ejpam-6087	230	7	is	be	AUX
ejpam-6087	230	8	an	an	DET
ejpam-6087	230	9	iterative	iterative	NOUN
ejpam-6087	230	10	feature	feature	NOUN
ejpam-6087	230	11	selection	selection	NOUN
ejpam-6087	230	12	technique	technique	NOUN
ejpam-6087	230	13	that	that	PRON
ejpam-6087	230	14	recursively	recursively	ADV
ejpam-6087	230	15	removes	remove	VERB
ejpam-6087	230	16	the	the	DET
ejpam-6087	230	17	least	least	ADJ
ejpam-6087	230	18	significant	significant	ADJ
ejpam-6087	230	19	features	feature	NOUN
ejpam-6087	230	20	based	base	VERB
ejpam-6087	230	21	on	on	ADP
ejpam-6087	230	22	the	the	DET
ejpam-6087	230	23	performance	performance	NOUN
ejpam-6087	230	24	of	of	ADP
ejpam-6087	230	25	a	a	DET
ejpam-6087	230	26	chosen	choose	VERB
ejpam-6087	230	27	model	model	NOUN
ejpam-6087	230	28	.	.	PUNCT
ejpam-6087	231	1	the	the	DET
ejpam-6087	231	2	algorithm	algorithm	NOUN
ejpam-6087	231	3	continues	continue	VERB
ejpam-6087	231	4	to	to	PART
ejpam-6087	231	5	remove	remove	VERB
ejpam-6087	231	6	features	feature	NOUN
ejpam-6087	231	7	until	until	SCONJ
ejpam-6087	231	8	only	only	ADV
ejpam-6087	231	9	the	the	DET
ejpam-6087	231	10	desired	desire	VERB
ejpam-6087	231	11	number	number	NOUN
ejpam-6087	231	12	of	of	ADP
ejpam-6087	231	13	features	feature	NOUN
ejpam-6087	231	14	remains	remain	VERB
ejpam-6087	231	15	.	.	PUNCT
ejpam-6087	232	1	rfe	rfe	NOUN
ejpam-6087	232	2	works	work	VERB
ejpam-6087	232	3	by	by	ADP
ejpam-6087	232	4	fitting	fit	VERB
ejpam-6087	232	5	a	a	DET
ejpam-6087	232	6	model	model	NOUN
ejpam-6087	232	7	to	to	ADP
ejpam-6087	232	8	the	the	DET
ejpam-6087	232	9	data	datum	NOUN
ejpam-6087	232	10	and	and	CCONJ
ejpam-6087	232	11	ranking	rank	VERB
ejpam-6087	232	12	the	the	DET
ejpam-6087	232	13	features	feature	NOUN
ejpam-6087	232	14	based	base	VERB
ejpam-6087	232	15	on	on	ADP
ejpam-6087	232	16	their	their	PRON
ejpam-6087	232	17	importance	importance	NOUN
ejpam-6087	232	18	.	.	PUNCT
ejpam-6087	233	1	the	the	DET
ejpam-6087	233	2	features	feature	NOUN
ejpam-6087	233	3	are	be	AUX
ejpam-6087	233	4	then	then	ADV
ejpam-6087	233	5	removed	remove	VERB
ejpam-6087	233	6	on	on	ADP
ejpam-6087	233	7	the	the	DET
ejpam-6087	233	8	basis	basis	NOUN
ejpam-6087	233	9	of	of	ADP
ejpam-6087	233	10	this	this	DET
ejpam-6087	233	11	ranking	ranking	NOUN
ejpam-6087	233	12	,	,	PUNCT
ejpam-6087	233	13	and	and	CCONJ
ejpam-6087	233	14	the	the	DET
ejpam-6087	233	15	model	model	NOUN
ejpam-6087	233	16	is	be	AUX
ejpam-6087	233	17	re	re	VERB
ejpam-6087	233	18	-	-	VERB
ejpam-6087	233	19	trained	train	VERB
ejpam-6087	233	20	on	on	ADP
ejpam-6087	233	21	the	the	DET
ejpam-6087	233	22	reduced	reduce	VERB
ejpam-6087	233	23	feature	feature	NOUN
ejpam-6087	233	24	set	set	NOUN
ejpam-6087	233	25	.	.	PUNCT
ejpam-6087	234	1	this	this	DET
ejpam-6087	234	2	process	process	NOUN
ejpam-6087	234	3	is	be	AUX
ejpam-6087	234	4	repeated	repeat	VERB
ejpam-6087	234	5	until	until	SCONJ
ejpam-6087	234	6	the	the	DET
ejpam-6087	234	7	number	number	NOUN
ejpam-6087	234	8	of	of	ADP
ejpam-6087	234	9	features	feature	NOUN
ejpam-6087	234	10	is	be	AUX
ejpam-6087	234	11	equal	equal	ADJ
ejpam-6087	234	12	to	to	ADP
ejpam-6087	234	13	the	the	DET
ejpam-6087	234	14	target	target	NOUN
ejpam-6087	234	15	number	number	NOUN
ejpam-6087	234	16	,	,	PUNCT
ejpam-6087	235	1	k.	k.	PROPN
ejpam-6087	235	2	rfe	rfe	PROPN
ejpam-6087	235	3	is	be	AUX
ejpam-6087	235	4	typically	typically	ADV
ejpam-6087	235	5	used	use	VERB
ejpam-6087	235	6	with	with	ADP
ejpam-6087	235	7	models	model	NOUN
ejpam-6087	235	8	that	that	PRON
ejpam-6087	235	9	provide	provide	VERB
ejpam-6087	235	10	feature	feature	NOUN
ejpam-6087	235	11	importance	importance	NOUN
ejpam-6087	235	12	scores	score	NOUN
ejpam-6087	235	13	,	,	PUNCT
ejpam-6087	235	14	such	such	ADJ
ejpam-6087	235	15	as	as	ADP
ejpam-6087	235	16	lr	lr	PROPN
ejpam-6087	235	17	,	,	PUNCT
ejpam-6087	235	18	svm	svm	ADJ
ejpam-6087	235	19	,	,	PUNCT
ejpam-6087	235	20	or	or	CCONJ
ejpam-6087	235	21	dts	dts	NOUN
ejpam-6087	235	22	.	.	NOUN
ejpam-6087	236	1	•	•	NUM
ejpam-6087	236	2	steps	step	NOUN
ejpam-6087	236	3	of	of	ADP
ejpam-6087	236	4	rfe	rfe	NOUN
ejpam-6087	236	5	:	:	PUNCT
ejpam-6087	236	6	(	(	PUNCT
ejpam-6087	236	7	i	i	NOUN
ejpam-6087	236	8	)	)	PUNCT
ejpam-6087	236	9	model	model	NOUN
ejpam-6087	236	10	training	training	NOUN
ejpam-6087	236	11	:	:	PUNCT
ejpam-6087	236	12	train	train	VERB
ejpam-6087	236	13	a	a	DET
ejpam-6087	236	14	model	model	NOUN
ejpam-6087	236	15	m	m	VERB
ejpam-6087	236	16	on	on	ADP
ejpam-6087	236	17	the	the	DET
ejpam-6087	236	18	initial	initial	ADJ
ejpam-6087	236	19	set	set	NOUN
ejpam-6087	236	20	of	of	ADP
ejpam-6087	236	21	features	feature	NOUN
ejpam-6087	236	22	x	x	VERB
ejpam-6087	236	23	to	to	PART
ejpam-6087	236	24	predict	predict	VERB
ejpam-6087	236	25	the	the	DET
ejpam-6087	236	26	target	target	NOUN
ejpam-6087	236	27	y.	y.	PROPN
ejpam-6087	236	28	(	(	PUNCT
ejpam-6087	236	29	ii	ii	NOUN
ejpam-6087	236	30	)	)	PUNCT
ejpam-6087	236	31	feature	feature	NOUN
ejpam-6087	236	32	ranking	ranking	NOUN
ejpam-6087	236	33	:	:	PUNCT
ejpam-6087	236	34	evaluate	evaluate	VERB
ejpam-6087	236	35	the	the	DET
ejpam-6087	236	36	importance	importance	NOUN
ejpam-6087	236	37	of	of	ADP
ejpam-6087	236	38	each	each	DET
ejpam-6087	236	39	feature	feature	NOUN
ejpam-6087	236	40	based	base	VERB
ejpam-6087	236	41	on	on	ADP
ejpam-6087	236	42	the	the	DET
ejpam-6087	236	43	coefficients	coefficient	NOUN
ejpam-6087	236	44	of	of	ADP
ejpam-6087	236	45	the	the	DET
ejpam-6087	236	46	model	model	NOUN
ejpam-6087	236	47	or	or	CCONJ
ejpam-6087	236	48	feature	feature	NOUN
ejpam-6087	236	49	weights	weight	NOUN
ejpam-6087	236	50	.	.	PUNCT
ejpam-6087	237	1	(	(	PUNCT
ejpam-6087	237	2	iii	iii	X
ejpam-6087	237	3	)	)	PUNCT
ejpam-6087	237	4	feature	feature	NOUN
ejpam-6087	237	5	elimination	elimination	NOUN
ejpam-6087	237	6	:	:	PUNCT
ejpam-6087	237	7	remove	remove	VERB
ejpam-6087	237	8	the	the	DET
ejpam-6087	237	9	least	least	ADV
ejpam-6087	237	10	important	important	ADJ
ejpam-6087	237	11	feature(s	feature(s	NOUN
ejpam-6087	237	12	)	)	PUNCT
ejpam-6087	237	13	from	from	ADP
ejpam-6087	237	14	the	the	DET
ejpam-6087	237	15	feature	feature	NOUN
ejpam-6087	237	16	set	set	NOUN
ejpam-6087	237	17	.	.	PUNCT
ejpam-6087	238	1	(	(	PUNCT
ejpam-6087	238	2	iv	iv	X
ejpam-6087	238	3	)	)	PUNCT
ejpam-6087	238	4	model	model	NOUN
ejpam-6087	238	5	retraining	retraining	NOUN
ejpam-6087	238	6	:	:	PUNCT
ejpam-6087	238	7	re	re	VERB
ejpam-6087	238	8	-	-	VERB
ejpam-6087	238	9	train	train	VERB
ejpam-6087	238	10	the	the	DET
ejpam-6087	238	11	model	model	NOUN
ejpam-6087	238	12	with	with	ADP
ejpam-6087	238	13	the	the	DET
ejpam-6087	238	14	reduced	reduce	VERB
ejpam-6087	238	15	set	set	NOUN
ejpam-6087	238	16	of	of	ADP
ejpam-6087	238	17	features	feature	NOUN
ejpam-6087	238	18	.	.	PUNCT
ejpam-6087	239	1	(	(	PUNCT
ejpam-6087	239	2	v	v	NOUN
ejpam-6087	239	3	)	)	PUNCT
ejpam-6087	239	4	iteration	iteration	NOUN
ejpam-6087	239	5	:	:	PUNCT
ejpam-6087	239	6	repeat	repeat	VERB
ejpam-6087	239	7	the	the	DET
ejpam-6087	239	8	process	process	NOUN
ejpam-6087	239	9	until	until	SCONJ
ejpam-6087	239	10	the	the	DET
ejpam-6087	239	11	desired	desire	VERB
ejpam-6087	239	12	number	number	NOUN
ejpam-6087	239	13	of	of	ADP
ejpam-6087	239	14	features	feature	NOUN
ejpam-6087	239	15	k	k	PROPN
ejpam-6087	239	16	is	be	AUX
ejpam-6087	239	17	reached	reach	VERB
ejpam-6087	239	18	.	.	PUNCT
ejpam-6087	240	1	m.	m.	NOUN
ejpam-6087	240	2	al	al	PROPN
ejpam-6087	240	3	-	-	PUNCT
ejpam-6087	240	4	tarawneh	tarawneh	PROPN
ejpam-6087	240	5	et	et	PROPN
ejpam-6087	240	6	al	al	PROPN
ejpam-6087	240	7	.	.	PUNCT
ejpam-6087	240	8	/	/	SYM
ejpam-6087	240	9	eur	eur	PROPN
ejpam-6087	240	10	.	.	PUNCT
ejpam-6087	241	1	j.	j.	PROPN
ejpam-6087	241	2	pure	pure	PROPN
ejpam-6087	241	3	appl	appl	PROPN
ejpam-6087	241	4	.	.	PROPN
ejpam-6087	241	5	math	math	PROPN
ejpam-6087	241	6	,	,	PUNCT
ejpam-6087	241	7	18	18	NUM
ejpam-6087	241	8	(	(	PUNCT
ejpam-6087	241	9	2	2	NUM
ejpam-6087	241	10	)	)	PUNCT
ejpam-6087	241	11	(	(	PUNCT
ejpam-6087	241	12	2025	2025	NUM
ejpam-6087	241	13	)	)	PUNCT
ejpam-6087	241	14	,	,	PUNCT
ejpam-6087	241	15	6087	6087	NUM
ejpam-6087	241	16	13	13	NUM
ejpam-6087	241	17	of	of	ADP
ejpam-6087	241	18	39	39	NUM
ejpam-6087	241	19	•	•	NOUN
ejpam-6087	241	20	mathematical	mathematical	ADJ
ejpam-6087	241	21	formulation	formulation	NOUN
ejpam-6087	241	22	:	:	PUNCT
ejpam-6087	241	23	let	let	VERB
ejpam-6087	241	24	x	x	PRON
ejpam-6087	241	25	be	be	AUX
ejpam-6087	241	26	the	the	DET
ejpam-6087	241	27	matrix	matrix	NOUN
ejpam-6087	241	28	of	of	ADP
ejpam-6087	241	29	features	feature	NOUN
ejpam-6087	241	30	with	with	ADP
ejpam-6087	241	31	samples	sample	NOUN
ejpam-6087	241	32	n	n	PRON
ejpam-6087	241	33	and	and	CCONJ
ejpam-6087	241	34	features	feature	VERB
ejpam-6087	241	35	m	m	PRON
ejpam-6087	241	36	,	,	PUNCT
ejpam-6087	241	37	and	and	CCONJ
ejpam-6087	241	38	y	y	PROPN
ejpam-6087	241	39	be	be	VERB
ejpam-6087	241	40	the	the	DET
ejpam-6087	241	41	target	target	NOUN
ejpam-6087	241	42	vector	vector	NOUN
ejpam-6087	241	43	.	.	PUNCT
ejpam-6087	242	1	(	(	PUNCT
ejpam-6087	242	2	i	i	NOUN
ejpam-6087	242	3	)	)	PUNCT
ejpam-6087	242	4	model	model	NOUN
ejpam-6087	242	5	construction	construction	NOUN
ejpam-6087	242	6	:	:	PUNCT
ejpam-6087	242	7	train	train	VERB
ejpam-6087	242	8	a	a	DET
ejpam-6087	242	9	model	model	NOUN
ejpam-6087	242	10	m	m	VERB
ejpam-6087	242	11	with	with	ADP
ejpam-6087	242	12	features	feature	NOUN
ejpam-6087	242	13	x	x	NOUN
ejpam-6087	242	14	:	:	PUNCT
ejpam-6087	242	15	f(x	f(x	PROPN
ejpam-6087	242	16	)	)	PUNCT
ejpam-6087	242	17	=	=	PUNCT
ejpam-6087	243	1	ŷ	ŷ	NUM
ejpam-6087	243	2	(	(	PUNCT
ejpam-6087	243	3	2	2	X
ejpam-6087	243	4	)	)	PUNCT
ejpam-6087	243	5	where	where	SCONJ
ejpam-6087	243	6	ŷ	ŷ	NUM
ejpam-6087	243	7	is	be	AUX
ejpam-6087	243	8	the	the	DET
ejpam-6087	243	9	predicted	predict	VERB
ejpam-6087	243	10	target	target	NOUN
ejpam-6087	243	11	vector	vector	NOUN
ejpam-6087	243	12	.	.	PUNCT
ejpam-6087	244	1	(	(	PUNCT
ejpam-6087	244	2	ii	ii	NOUN
ejpam-6087	244	3	)	)	PUNCT
ejpam-6087	244	4	feature	feature	NOUN
ejpam-6087	244	5	importance	importance	NOUN
ejpam-6087	244	6	:	:	PUNCT
ejpam-6087	244	7	compute	compute	VERB
ejpam-6087	244	8	the	the	DET
ejpam-6087	244	9	importance	importance	NOUN
ejpam-6087	244	10	of	of	ADP
ejpam-6087	244	11	each	each	DET
ejpam-6087	244	12	feature	feature	NOUN
ejpam-6087	244	13	i	i	PRON
ejpam-6087	244	14	:	:	PUNCT
ejpam-6087	244	15	feature	feature	NOUN
ejpam-6087	244	16	importance	importance	NOUN
ejpam-6087	244	17	=	=	PUNCT
ejpam-6087	244	18	|wi|	|wi|	X
ejpam-6087	244	19	(	(	PUNCT
ejpam-6087	244	20	3	3	NUM
ejpam-6087	244	21	)	)	PUNCT
ejpam-6087	244	22	where	where	SCONJ
ejpam-6087	244	23	wi	wi	PROPN
ejpam-6087	244	24	is	be	AUX
ejpam-6087	244	25	the	the	DET
ejpam-6087	244	26	coefficient	coefficient	NOUN
ejpam-6087	244	27	for	for	ADP
ejpam-6087	244	28	feature	feature	NOUN
ejpam-6087	244	29	i.	i.	NOUN
ejpam-6087	244	30	(	(	PUNCT
ejpam-6087	244	31	iii	iii	NOUN
ejpam-6087	244	32	)	)	PUNCT
ejpam-6087	244	33	feature	feature	NOUN
ejpam-6087	244	34	elimination	elimination	NOUN
ejpam-6087	244	35	:	:	PUNCT
ejpam-6087	244	36	identify	identify	VERB
ejpam-6087	244	37	and	and	CCONJ
ejpam-6087	244	38	remove	remove	VERB
ejpam-6087	244	39	the	the	DET
ejpam-6087	244	40	least	least	ADV
ejpam-6087	244	41	important	important	ADJ
ejpam-6087	244	42	feature(s	feature(s	NOUN
ejpam-6087	244	43	)	)	PUNCT
ejpam-6087	244	44	based	base	VERB
ejpam-6087	244	45	on	on	ADP
ejpam-6087	244	46	|wi|	|wi|	PROPN
ejpam-6087	244	47	.	.	PUNCT
ejpam-6087	245	1	(	(	PUNCT
ejpam-6087	245	2	iv	iv	X
ejpam-6087	245	3	)	)	PUNCT
ejpam-6087	245	4	update	update	NOUN
ejpam-6087	245	5	feature	feature	NOUN
ejpam-6087	245	6	set	set	NOUN
ejpam-6087	245	7	:	:	PUNCT
ejpam-6087	245	8	update	update	VERB
ejpam-6087	245	9	the	the	DET
ejpam-6087	245	10	feature	feature	NOUN
ejpam-6087	245	11	set	set	VERB
ejpam-6087	245	12	x	x	PUNCT
ejpam-6087	245	13	and	and	CCONJ
ejpam-6087	245	14	repeat	repeat	VERB
ejpam-6087	245	15	the	the	DET
ejpam-6087	245	16	training	training	NOUN
ejpam-6087	245	17	and	and	CCONJ
ejpam-6087	245	18	elimination	elimination	NOUN
ejpam-6087	245	19	steps	step	NOUN
ejpam-6087	245	20	.	.	PUNCT
ejpam-6087	246	1	3.4.2	3.4.2	NUM
ejpam-6087	246	2	.	.	PUNCT
ejpam-6087	246	3	selectkbest	selectkb	ADJ
ejpam-6087	246	4	selectkbest	selectkb	ADJ
ejpam-6087	246	5	is	be	AUX
ejpam-6087	246	6	a	a	DET
ejpam-6087	246	7	univariate	univariate	ADJ
ejpam-6087	246	8	feature	feature	NOUN
ejpam-6087	246	9	selection	selection	NOUN
ejpam-6087	246	10	method	method	NOUN
ejpam-6087	246	11	that	that	PRON
ejpam-6087	246	12	selects	select	VERB
ejpam-6087	246	13	the	the	DET
ejpam-6087	246	14	top	top	ADJ
ejpam-6087	246	15	k	k	PROPN
ejpam-6087	246	16	features	feature	NOUN
ejpam-6087	246	17	based	base	VERB
ejpam-6087	246	18	on	on	ADP
ejpam-6087	246	19	univariate	univariate	ADJ
ejpam-6087	246	20	statistical	statistical	ADJ
ejpam-6087	246	21	tests	test	NOUN
ejpam-6087	246	22	.	.	PUNCT
ejpam-6087	247	1	it	it	PRON
ejpam-6087	247	2	evaluates	evaluate	VERB
ejpam-6087	247	3	each	each	DET
ejpam-6087	247	4	feature	feature	NOUN
ejpam-6087	247	5	independently	independently	ADV
ejpam-6087	247	6	to	to	PART
ejpam-6087	247	7	determine	determine	VERB
ejpam-6087	247	8	its	its	PRON
ejpam-6087	247	9	relevance	relevance	NOUN
ejpam-6087	247	10	to	to	ADP
ejpam-6087	247	11	the	the	DET
ejpam-6087	247	12	target	target	NOUN
ejpam-6087	247	13	variable	variable	NOUN
ejpam-6087	247	14	.	.	PUNCT
ejpam-6087	248	1	selectkbest	selectkb	ADJ
ejpam-6087	248	2	evaluates	evaluate	VERB
ejpam-6087	248	3	each	each	DET
ejpam-6087	248	4	feature	feature	NOUN
ejpam-6087	248	5	using	use	VERB
ejpam-6087	248	6	statistical	statistical	ADJ
ejpam-6087	248	7	tests	test	NOUN
ejpam-6087	248	8	and	and	CCONJ
ejpam-6087	248	9	selects	select	VERB
ejpam-6087	248	10	the	the	DET
ejpam-6087	248	11	top	top	ADJ
ejpam-6087	248	12	k	k	PROPN
ejpam-6087	248	13	features	feature	NOUN
ejpam-6087	248	14	based	base	VERB
ejpam-6087	248	15	on	on	ADP
ejpam-6087	248	16	test	test	NOUN
ejpam-6087	248	17	scores	score	NOUN
ejpam-6087	248	18	.	.	PUNCT
ejpam-6087	249	1	common	common	ADJ
ejpam-6087	249	2	statistical	statistical	ADJ
ejpam-6087	249	3	tests	test	NOUN
ejpam-6087	249	4	used	use	VERB
ejpam-6087	249	5	include	include	VERB
ejpam-6087	249	6	the	the	DET
ejpam-6087	249	7	chi	chi	ADJ
ejpam-6087	249	8	-	-	PUNCT
ejpam-6087	249	9	square	square	ADJ
ejpam-6087	249	10	test	test	NOUN
ejpam-6087	249	11	,	,	PUNCT
ejpam-6087	249	12	the	the	DET
ejpam-6087	249	13	anova	anova	PROPN
ejpam-6087	249	14	f	f	X
ejpam-6087	249	15	-	-	PUNCT
ejpam-6087	249	16	value	value	NOUN
ejpam-6087	249	17	,	,	PUNCT
ejpam-6087	249	18	and	and	CCONJ
ejpam-6087	249	19	mutual	mutual	ADJ
ejpam-6087	249	20	information	information	NOUN
ejpam-6087	249	21	.	.	PUNCT
ejpam-6087	250	1	•	•	NUM
ejpam-6087	250	2	steps	step	NOUN
ejpam-6087	250	3	of	of	ADP
ejpam-6087	250	4	selectkbest	selectkb	ADJ
ejpam-6087	250	5	:	:	PUNCT
ejpam-6087	250	6	(	(	PUNCT
ejpam-6087	250	7	i	i	NOUN
ejpam-6087	250	8	)	)	PUNCT
ejpam-6087	250	9	compute	compute	PROPN
ejpam-6087	250	10	test	test	NOUN
ejpam-6087	250	11	statistics	statistic	NOUN
ejpam-6087	250	12	:	:	PUNCT
ejpam-6087	250	13	calculate	calculate	VERB
ejpam-6087	250	14	the	the	DET
ejpam-6087	250	15	test	test	NOUN
ejpam-6087	250	16	statistic	statistic	NOUN
ejpam-6087	250	17	for	for	ADP
ejpam-6087	250	18	each	each	DET
ejpam-6087	250	19	feature	feature	NOUN
ejpam-6087	250	20	based	base	VERB
ejpam-6087	250	21	on	on	ADP
ejpam-6087	250	22	a	a	DET
ejpam-6087	250	23	chosen	choose	VERB
ejpam-6087	250	24	statistical	statistical	ADJ
ejpam-6087	250	25	test	test	NOUN
ejpam-6087	250	26	.	.	PUNCT
ejpam-6087	251	1	(	(	PUNCT
ejpam-6087	251	2	ii	ii	X
ejpam-6087	251	3	)	)	PUNCT
ejpam-6087	251	4	rank	rank	NOUN
ejpam-6087	251	5	features	feature	NOUN
ejpam-6087	251	6	:	:	PUNCT
ejpam-6087	251	7	rank	rank	NOUN
ejpam-6087	251	8	features	feature	NOUN
ejpam-6087	251	9	according	accord	VERB
ejpam-6087	251	10	to	to	ADP
ejpam-6087	251	11	their	their	PRON
ejpam-6087	251	12	test	test	NOUN
ejpam-6087	251	13	statistics	statistic	NOUN
ejpam-6087	251	14	.	.	PUNCT
ejpam-6087	252	1	(	(	PUNCT
ejpam-6087	252	2	iii	iii	X
ejpam-6087	252	3	)	)	PUNCT
ejpam-6087	252	4	select	select	ADJ
ejpam-6087	252	5	top	top	ADJ
ejpam-6087	252	6	features	feature	NOUN
ejpam-6087	252	7	:	:	PUNCT
ejpam-6087	252	8	choose	choose	VERB
ejpam-6087	252	9	the	the	DET
ejpam-6087	252	10	best	good	ADJ
ejpam-6087	252	11	k	k	X
ejpam-6087	252	12	features	feature	NOUN
ejpam-6087	252	13	with	with	ADP
ejpam-6087	252	14	the	the	DET
ejpam-6087	252	15	highest	high	ADJ
ejpam-6087	252	16	test	test	NOUN
ejpam-6087	252	17	statistics	statistic	NOUN
ejpam-6087	252	18	.	.	PUNCT
ejpam-6087	253	1	•	•	NUM
ejpam-6087	253	2	mathematical	mathematical	ADJ
ejpam-6087	253	3	formulation	formulation	NOUN
ejpam-6087	253	4	:	:	PUNCT
ejpam-6087	253	5	let	let	VERB
ejpam-6087	253	6	x	x	PRON
ejpam-6087	253	7	be	be	AUX
ejpam-6087	253	8	the	the	DET
ejpam-6087	253	9	matrix	matrix	NOUN
ejpam-6087	253	10	of	of	ADP
ejpam-6087	253	11	features	feature	NOUN
ejpam-6087	253	12	with	with	ADP
ejpam-6087	253	13	samples	sample	NOUN
ejpam-6087	253	14	n	n	PRON
ejpam-6087	253	15	and	and	CCONJ
ejpam-6087	253	16	features	feature	VERB
ejpam-6087	253	17	m	m	PRON
ejpam-6087	253	18	,	,	PUNCT
ejpam-6087	253	19	and	and	CCONJ
ejpam-6087	253	20	y	y	PROPN
ejpam-6087	253	21	be	be	VERB
ejpam-6087	253	22	the	the	DET
ejpam-6087	253	23	target	target	NOUN
ejpam-6087	253	24	vector	vector	NOUN
ejpam-6087	253	25	.	.	PUNCT
ejpam-6087	254	1	(	(	PUNCT
ejpam-6087	254	2	i	i	NOUN
ejpam-6087	254	3	)	)	PUNCT
ejpam-6087	254	4	compute	compute	PROPN
ejpam-6087	254	5	test	test	NOUN
ejpam-6087	254	6	statistics	statistic	NOUN
ejpam-6087	254	7	:	:	PUNCT
ejpam-6087	254	8	for	for	ADP
ejpam-6087	254	9	each	each	DET
ejpam-6087	254	10	feature	feature	NOUN
ejpam-6087	254	11	xi	xi	PROPN
ejpam-6087	254	12	,	,	PUNCT
ejpam-6087	254	13	compute	compute	VERB
ejpam-6087	254	14	the	the	DET
ejpam-6087	254	15	test	test	NOUN
ejpam-6087	254	16	statistic	statistic	NOUN
ejpam-6087	254	17	ti	ti	NOUN
ejpam-6087	254	18	:	:	PUNCT
ejpam-6087	254	19	ti	ti	X
ejpam-6087	254	20	=	=	NOUN
ejpam-6087	254	21	test	test	NOUN
ejpam-6087	254	22	statistic(xi	statistic(xi	NOUN
ejpam-6087	254	23	,	,	PUNCT
ejpam-6087	254	24	y	y	PROPN
ejpam-6087	254	25	)	)	PUNCT
ejpam-6087	254	26	(	(	PUNCT
ejpam-6087	254	27	4	4	NUM
ejpam-6087	254	28	)	)	PUNCT
ejpam-6087	254	29	(	(	PUNCT
ejpam-6087	254	30	ii	ii	NOUN
ejpam-6087	254	31	)	)	PUNCT
ejpam-6087	254	32	ranking	rank	VERB
ejpam-6087	254	33	features	feature	NOUN
ejpam-6087	254	34	:	:	PUNCT
ejpam-6087	254	35	rank	rank	NOUN
ejpam-6087	254	36	features	feature	NOUN
ejpam-6087	254	37	based	base	VERB
ejpam-6087	254	38	on	on	ADP
ejpam-6087	254	39	the	the	DET
ejpam-6087	254	40	computed	computed	ADJ
ejpam-6087	254	41	test	test	NOUN
ejpam-6087	254	42	statistics	statistic	NOUN
ejpam-6087	254	43	:	:	PUNCT
ejpam-6087	254	44	rank(xi	rank(xi	ADJ
ejpam-6087	254	45	)	)	PUNCT
ejpam-6087	254	46	=	=	NUM
ejpam-6087	254	47	sort(t1	sort(t1	NOUN
ejpam-6087	254	48	,	,	PUNCT
ejpam-6087	254	49	t2	t2	NOUN
ejpam-6087	254	50	,	,	PUNCT
ejpam-6087	254	51	.	.	PUNCT
ejpam-6087	254	52	.	.	PUNCT
ejpam-6087	255	1	.	.	PUNCT
ejpam-6087	256	1	,	,	PUNCT
ejpam-6087	256	2	tm	tm	PROPN
ejpam-6087	256	3	)	)	PUNCT
ejpam-6087	256	4	(	(	PUNCT
ejpam-6087	256	5	5	5	NUM
ejpam-6087	256	6	)	)	PUNCT
ejpam-6087	256	7	(	(	PUNCT
ejpam-6087	256	8	iii	iii	NOUN
ejpam-6087	256	9	)	)	PUNCT
ejpam-6087	256	10	selecting	select	VERB
ejpam-6087	256	11	top	top	ADJ
ejpam-6087	256	12	k	k	NOUN
ejpam-6087	256	13	:	:	PUNCT
ejpam-6087	256	14	select	select	VERB
ejpam-6087	256	15	the	the	DET
ejpam-6087	256	16	k	k	PROPN
ejpam-6087	256	17	features	feature	VERB
ejpam-6087	256	18	with	with	ADP
ejpam-6087	256	19	the	the	DET
ejpam-6087	256	20	highest	high	ADJ
ejpam-6087	256	21	ti	ti	NOUN
ejpam-6087	256	22	:	:	PUNCT
ejpam-6087	256	23	xselected	xselected	ADJ
ejpam-6087	256	24	=	=	PUNCT
ejpam-6087	256	25	top	top	NOUN
ejpam-6087	256	26	k	k	PROPN
ejpam-6087	256	27	features	feature	NOUN
ejpam-6087	256	28	based	base	VERB
ejpam-6087	256	29	on	on	ADP
ejpam-6087	256	30	ti	ti	PROPN
ejpam-6087	256	31	(	(	PUNCT
ejpam-6087	256	32	6	6	NUM
ejpam-6087	256	33	)	)	PUNCT
ejpam-6087	256	34	m.	m.	NOUN
ejpam-6087	256	35	al	al	PROPN
ejpam-6087	256	36	-	-	PUNCT
ejpam-6087	256	37	tarawneh	tarawneh	PROPN
ejpam-6087	256	38	et	et	PROPN
ejpam-6087	256	39	al	al	PROPN
ejpam-6087	256	40	.	.	PUNCT
ejpam-6087	256	41	/	/	SYM
ejpam-6087	256	42	eur	eur	PROPN
ejpam-6087	256	43	.	.	PUNCT
ejpam-6087	257	1	j.	j.	PROPN
ejpam-6087	257	2	pure	pure	PROPN
ejpam-6087	257	3	appl	appl	PROPN
ejpam-6087	257	4	.	.	PROPN
ejpam-6087	257	5	math	math	PROPN
ejpam-6087	257	6	,	,	PUNCT
ejpam-6087	257	7	18	18	NUM
ejpam-6087	257	8	(	(	PUNCT
ejpam-6087	257	9	2	2	NUM
ejpam-6087	257	10	)	)	PUNCT
ejpam-6087	257	11	(	(	PUNCT
ejpam-6087	257	12	2025	2025	NUM
ejpam-6087	257	13	)	)	PUNCT
ejpam-6087	257	14	,	,	PUNCT
ejpam-6087	257	15	6087	6087	NUM
ejpam-6087	257	16	14	14	NUM
ejpam-6087	257	17	of	of	ADP
ejpam-6087	257	18	39	39	NUM
ejpam-6087	257	19	3.4.3	3.4.3	NUM
ejpam-6087	257	20	.	.	PUNCT
ejpam-6087	258	1	principal	principal	ADJ
ejpam-6087	258	2	component	component	NOUN
ejpam-6087	258	3	analysis	analysis	NOUN
ejpam-6087	258	4	pca	pca	NOUN
ejpam-6087	258	5	is	be	AUX
ejpam-6087	258	6	a	a	DET
ejpam-6087	258	7	dimensionality	dimensionality	NOUN
ejpam-6087	258	8	reduction	reduction	NOUN
ejpam-6087	258	9	technique	technique	NOUN
ejpam-6087	258	10	that	that	PRON
ejpam-6087	258	11	transforms	transform	VERB
ejpam-6087	258	12	the	the	DET
ejpam-6087	258	13	original	original	ADJ
ejpam-6087	258	14	feature	feature	NOUN
ejpam-6087	258	15	space	space	NOUN
ejpam-6087	258	16	into	into	ADP
ejpam-6087	258	17	a	a	DET
ejpam-6087	258	18	new	new	ADJ
ejpam-6087	258	19	set	set	NOUN
ejpam-6087	258	20	of	of	ADP
ejpam-6087	258	21	uncorrelated	uncorrelated	ADJ
ejpam-6087	258	22	variables	variable	NOUN
ejpam-6087	258	23	called	call	VERB
ejpam-6087	258	24	principal	principal	ADJ
ejpam-6087	258	25	components	component	NOUN
ejpam-6087	258	26	.	.	PUNCT
ejpam-6087	259	1	pca	pca	PROPN
ejpam-6087	259	2	identifies	identify	VERB
ejpam-6087	259	3	the	the	DET
ejpam-6087	259	4	directions	direction	NOUN
ejpam-6087	259	5	(	(	PUNCT
ejpam-6087	259	6	principal	principal	ADJ
ejpam-6087	259	7	components	component	NOUN
ejpam-6087	259	8	)	)	PUNCT
ejpam-6087	259	9	along	along	ADP
ejpam-6087	259	10	which	which	PRON
ejpam-6087	259	11	the	the	DET
ejpam-6087	259	12	variance	variance	NOUN
ejpam-6087	259	13	of	of	ADP
ejpam-6087	259	14	the	the	DET
ejpam-6087	259	15	data	datum	NOUN
ejpam-6087	259	16	is	be	AUX
ejpam-6087	259	17	maximized	maximize	VERB
ejpam-6087	259	18	.	.	PUNCT
ejpam-6087	260	1	pca	pca	NOUN
ejpam-6087	260	2	transforms	transform	VERB
ejpam-6087	260	3	data	datum	NOUN
ejpam-6087	260	4	into	into	ADP
ejpam-6087	260	5	a	a	DET
ejpam-6087	260	6	new	new	ADJ
ejpam-6087	260	7	coordinate	coordinate	NOUN
ejpam-6087	260	8	system	system	NOUN
ejpam-6087	260	9	where	where	SCONJ
ejpam-6087	260	10	the	the	DET
ejpam-6087	260	11	first	first	ADJ
ejpam-6087	260	12	coordinate	coordinate	NOUN
ejpam-6087	260	13	accounts	account	VERB
ejpam-6087	260	14	for	for	ADP
ejpam-6087	260	15	the	the	DET
ejpam-6087	260	16	largest	large	ADJ
ejpam-6087	260	17	variance	variance	NOUN
ejpam-6087	260	18	in	in	ADP
ejpam-6087	260	19	the	the	DET
ejpam-6087	260	20	data	datum	NOUN
ejpam-6087	260	21	,	,	PUNCT
ejpam-6087	260	22	the	the	DET
ejpam-6087	260	23	second	second	ADJ
ejpam-6087	260	24	coordinate	coordinate	NOUN
ejpam-6087	260	25	accounts	account	VERB
ejpam-6087	260	26	for	for	ADP
ejpam-6087	260	27	the	the	DET
ejpam-6087	260	28	next	next	ADJ
ejpam-6087	260	29	largest	large	ADJ
ejpam-6087	260	30	variance	variance	NOUN
ejpam-6087	260	31	,	,	PUNCT
ejpam-6087	260	32	and	and	CCONJ
ejpam-6087	260	33	so	so	ADV
ejpam-6087	260	34	on	on	ADV
ejpam-6087	260	35	.	.	PUNCT
ejpam-6087	261	1	the	the	DET
ejpam-6087	261	2	goal	goal	NOUN
ejpam-6087	261	3	is	be	AUX
ejpam-6087	261	4	to	to	PART
ejpam-6087	261	5	reduce	reduce	VERB
ejpam-6087	261	6	the	the	DET
ejpam-6087	261	7	number	number	NOUN
ejpam-6087	261	8	of	of	ADP
ejpam-6087	261	9	dimensions	dimension	NOUN
ejpam-6087	261	10	while	while	SCONJ
ejpam-6087	261	11	retaining	retain	VERB
ejpam-6087	261	12	as	as	ADV
ejpam-6087	261	13	much	much	ADJ
ejpam-6087	261	14	variance	variance	NOUN
ejpam-6087	261	15	as	as	ADP
ejpam-6087	261	16	possible	possible	ADJ
ejpam-6087	261	17	.	.	PUNCT
ejpam-6087	262	1	•	•	NUM
ejpam-6087	262	2	steps	step	NOUN
ejpam-6087	262	3	of	of	ADP
ejpam-6087	262	4	pca	pca	NOUN
ejpam-6087	262	5	:	:	PUNCT
ejpam-6087	262	6	(	(	PUNCT
ejpam-6087	262	7	i	i	NOUN
ejpam-6087	262	8	)	)	PUNCT
ejpam-6087	262	9	compute	compute	NOUN
ejpam-6087	262	10	covariance	covariance	NOUN
ejpam-6087	262	11	matrix	matrix	NOUN
ejpam-6087	262	12	:	:	PUNCT
ejpam-6087	262	13	calculate	calculate	VERB
ejpam-6087	262	14	the	the	DET
ejpam-6087	262	15	covariance	covariance	NOUN
ejpam-6087	262	16	matrix	matrix	NOUN
ejpam-6087	262	17	of	of	ADP
ejpam-6087	262	18	the	the	DET
ejpam-6087	262	19	feature	feature	NOUN
ejpam-6087	262	20	set	set	NOUN
ejpam-6087	262	21	.	.	PUNCT
ejpam-6087	263	1	(	(	PUNCT
ejpam-6087	263	2	ii	ii	NOUN
ejpam-6087	263	3	)	)	PUNCT
ejpam-6087	263	4	eigen	eigen	PROPN
ejpam-6087	263	5	decomposition	decomposition	NOUN
ejpam-6087	263	6	:	:	PUNCT
ejpam-6087	263	7	perform	perform	VERB
ejpam-6087	263	8	eigenvalue	eigenvalue	NOUN
ejpam-6087	263	9	decomposition	decomposition	NOUN
ejpam-6087	263	10	to	to	PART
ejpam-6087	263	11	obtain	obtain	VERB
ejpam-6087	263	12	eigenvectors	eigenvector	NOUN
ejpam-6087	263	13	and	and	CCONJ
ejpam-6087	263	14	eigenvalues	eigenvalue	NOUN
ejpam-6087	263	15	.	.	PUNCT
ejpam-6087	264	1	(	(	PUNCT
ejpam-6087	264	2	iii	iii	NOUN
ejpam-6087	264	3	)	)	PUNCT
ejpam-6087	264	4	sort	sort	NOUN
ejpam-6087	264	5	principal	principal	ADJ
ejpam-6087	264	6	components	component	NOUN
ejpam-6087	264	7	:	:	PUNCT
ejpam-6087	264	8	sort	sort	VERB
ejpam-6087	264	9	eigenvectors	eigenvector	NOUN
ejpam-6087	264	10	by	by	ADP
ejpam-6087	264	11	the	the	DET
ejpam-6087	264	12	magnitude	magnitude	NOUN
ejpam-6087	264	13	of	of	ADP
ejpam-6087	264	14	their	their	PRON
ejpam-6087	264	15	corresponding	corresponding	ADJ
ejpam-6087	264	16	eigenvalues	eigenvalue	NOUN
ejpam-6087	264	17	.	.	PUNCT
ejpam-6087	265	1	(	(	PUNCT
ejpam-6087	265	2	iv	iv	X
ejpam-6087	265	3	)	)	PUNCT
ejpam-6087	265	4	select	select	ADJ
ejpam-6087	265	5	top	top	ADJ
ejpam-6087	265	6	components	component	NOUN
ejpam-6087	265	7	:	:	PUNCT
ejpam-6087	265	8	choose	choose	VERB
ejpam-6087	265	9	the	the	DET
ejpam-6087	265	10	top	top	ADJ
ejpam-6087	265	11	k	k	PROPN
ejpam-6087	265	12	eigenvectors	eigenvector	NOUN
ejpam-6087	265	13	to	to	PART
ejpam-6087	265	14	form	form	VERB
ejpam-6087	265	15	the	the	DET
ejpam-6087	265	16	principal	principal	ADJ
ejpam-6087	265	17	components	component	NOUN
ejpam-6087	265	18	.	.	PUNCT
ejpam-6087	266	1	(	(	PUNCT
ejpam-6087	266	2	v	v	NOUN
ejpam-6087	266	3	)	)	PUNCT
ejpam-6087	266	4	project	project	NOUN
ejpam-6087	266	5	data	datum	NOUN
ejpam-6087	266	6	:	:	PUNCT
ejpam-6087	266	7	transform	transform	VERB
ejpam-6087	266	8	the	the	DET
ejpam-6087	266	9	original	original	ADJ
ejpam-6087	266	10	data	datum	NOUN
ejpam-6087	266	11	into	into	ADP
ejpam-6087	266	12	the	the	DET
ejpam-6087	266	13	new	new	ADJ
ejpam-6087	266	14	feature	feature	NOUN
ejpam-6087	266	15	space	space	NOUN
ejpam-6087	266	16	spanned	span	VERB
ejpam-6087	266	17	by	by	ADP
ejpam-6087	266	18	the	the	DET
ejpam-6087	266	19	selected	select	VERB
ejpam-6087	266	20	principal	principal	ADJ
ejpam-6087	266	21	components	component	NOUN
ejpam-6087	266	22	.	.	PUNCT
ejpam-6087	267	1	•	•	NUM
ejpam-6087	267	2	mathematical	mathematical	ADJ
ejpam-6087	267	3	formulation	formulation	NOUN
ejpam-6087	267	4	:	:	PUNCT
ejpam-6087	267	5	let	let	VERB
ejpam-6087	267	6	x	x	PRON
ejpam-6087	267	7	be	be	AUX
ejpam-6087	267	8	the	the	DET
ejpam-6087	267	9	feature	feature	NOUN
ejpam-6087	267	10	matrix	matrix	NOUN
ejpam-6087	267	11	with	with	ADP
ejpam-6087	267	12	n	n	NOUN
ejpam-6087	267	13	samples	sample	NOUN
ejpam-6087	267	14	and	and	CCONJ
ejpam-6087	267	15	m	m	NOUN
ejpam-6087	267	16	features	feature	NOUN
ejpam-6087	267	17	.	.	PUNCT
ejpam-6087	268	1	(	(	PUNCT
ejpam-6087	268	2	i	i	NOUN
ejpam-6087	268	3	)	)	PUNCT
ejpam-6087	268	4	compute	compute	NOUN
ejpam-6087	268	5	covariance	covariance	NOUN
ejpam-6087	268	6	matrix	matrix	NOUN
ejpam-6087	268	7	:	:	PUNCT
ejpam-6087	269	1	σ	σ	NOUN
ejpam-6087	269	2	=	=	SYM
ejpam-6087	269	3	1	1	NUM
ejpam-6087	269	4	n−	n−	PROPN
ejpam-6087	269	5	1	1	NUM
ejpam-6087	269	6	xtx	xtx	PROPN
ejpam-6087	269	7	(	(	PUNCT
ejpam-6087	269	8	7	7	NUM
ejpam-6087	269	9	)	)	PUNCT
ejpam-6087	269	10	(	(	PUNCT
ejpam-6087	269	11	ii	ii	NOUN
ejpam-6087	269	12	)	)	PUNCT
ejpam-6087	269	13	perform	perform	VERB
ejpam-6087	269	14	eigen	eigen	NOUN
ejpam-6087	269	15	decomposition	decomposition	NOUN
ejpam-6087	269	16	:	:	PUNCT
ejpam-6087	269	17	find	find	VERB
ejpam-6087	269	18	eigenvectors	eigenvector	NOUN
ejpam-6087	269	19	w	w	NOUN
ejpam-6087	269	20	and	and	CCONJ
ejpam-6087	269	21	eigenvalues	eigenvalue	VERB
ejpam-6087	269	22	λ	λ	NOUN
ejpam-6087	269	23	from	from	ADP
ejpam-6087	269	24	the	the	DET
ejpam-6087	269	25	covariance	covariance	NOUN
ejpam-6087	269	26	matrix	matrix	NOUN
ejpam-6087	269	27	:	:	PUNCT
ejpam-6087	269	28	σw	σw	PROPN
ejpam-6087	269	29	=	=	PUNCT
ejpam-6087	269	30	wλ	wλ	PROPN
ejpam-6087	269	31	(	(	PUNCT
ejpam-6087	269	32	8)	8)	NUM
ejpam-6087	269	33	(	(	PUNCT
ejpam-6087	269	34	iii	iii	NOUN
ejpam-6087	269	35	)	)	PUNCT
ejpam-6087	269	36	sort	sort	NOUN
ejpam-6087	269	37	eigenvectors	eigenvector	NOUN
ejpam-6087	269	38	:	:	PUNCT
ejpam-6087	269	39	sort	sort	NOUN
ejpam-6087	269	40	eigenvectors	eigenvector	NOUN
ejpam-6087	269	41	based	base	VERB
ejpam-6087	269	42	on	on	ADP
ejpam-6087	269	43	the	the	DET
ejpam-6087	269	44	eigenvalues	eigenvalue	NOUN
ejpam-6087	269	45	in	in	ADP
ejpam-6087	269	46	descending	descend	VERB
ejpam-6087	269	47	order	order	NOUN
ejpam-6087	269	48	.	.	PUNCT
ejpam-6087	270	1	(	(	PUNCT
ejpam-6087	270	2	iv	iv	X
ejpam-6087	270	3	)	)	PUNCT
ejpam-6087	270	4	select	select	VERB
ejpam-6087	270	5	the	the	DET
ejpam-6087	270	6	top	top	ADJ
ejpam-6087	270	7	k	k	PROPN
ejpam-6087	270	8	components	component	NOUN
ejpam-6087	270	9	:	:	PUNCT
ejpam-6087	270	10	choose	choose	VERB
ejpam-6087	270	11	the	the	DET
ejpam-6087	270	12	top	top	ADJ
ejpam-6087	270	13	k	k	PROPN
ejpam-6087	270	14	eigenvectors	eigenvector	NOUN
ejpam-6087	270	15	to	to	PART
ejpam-6087	270	16	form	form	VERB
ejpam-6087	270	17	the	the	DET
ejpam-6087	270	18	new	new	ADJ
ejpam-6087	270	19	feature	feature	NOUN
ejpam-6087	270	20	space	space	NOUN
ejpam-6087	270	21	.	.	PUNCT
ejpam-6087	271	1	(	(	PUNCT
ejpam-6087	271	2	v	v	NOUN
ejpam-6087	271	3	)	)	PUNCT
ejpam-6087	271	4	project	project	NOUN
ejpam-6087	271	5	data	datum	NOUN
ejpam-6087	271	6	:	:	PUNCT
ejpam-6087	271	7	project	project	VERB
ejpam-6087	271	8	the	the	DET
ejpam-6087	271	9	original	original	ADJ
ejpam-6087	271	10	data	datum	NOUN
ejpam-6087	271	11	onto	onto	ADP
ejpam-6087	271	12	the	the	DET
ejpam-6087	271	13	new	new	ADJ
ejpam-6087	271	14	space	space	NOUN
ejpam-6087	271	15	:	:	PUNCT
ejpam-6087	271	16	xnew	xnew	PROPN
ejpam-6087	271	17	=	=	SYM
ejpam-6087	271	18	xw	xw	PROPN
ejpam-6087	271	19	(	(	PUNCT
ejpam-6087	271	20	9	9	NUM
ejpam-6087	271	21	)	)	PUNCT
ejpam-6087	271	22	m.	m.	NOUN
ejpam-6087	271	23	al	al	PROPN
ejpam-6087	271	24	-	-	PUNCT
ejpam-6087	271	25	tarawneh	tarawneh	PROPN
ejpam-6087	271	26	et	et	PROPN
ejpam-6087	271	27	al	al	PROPN
ejpam-6087	271	28	.	.	PUNCT
ejpam-6087	271	29	/	/	SYM
ejpam-6087	271	30	eur	eur	PROPN
ejpam-6087	271	31	.	.	PUNCT
ejpam-6087	272	1	j.	j.	PROPN
ejpam-6087	272	2	pure	pure	PROPN
ejpam-6087	272	3	appl	appl	PROPN
ejpam-6087	272	4	.	.	PROPN
ejpam-6087	272	5	math	math	PROPN
ejpam-6087	272	6	,	,	PUNCT
ejpam-6087	272	7	18	18	NUM
ejpam-6087	272	8	(	(	PUNCT
ejpam-6087	272	9	2	2	NUM
ejpam-6087	272	10	)	)	PUNCT
ejpam-6087	272	11	(	(	PUNCT
ejpam-6087	272	12	2025	2025	NUM
ejpam-6087	272	13	)	)	PUNCT
ejpam-6087	272	14	,	,	PUNCT
ejpam-6087	272	15	6087	6087	NUM
ejpam-6087	272	16	15	15	NUM
ejpam-6087	272	17	of	of	ADP
ejpam-6087	272	18	39	39	NUM
ejpam-6087	272	19	3.4.4	3.4.4	NUM
ejpam-6087	272	20	.	.	PUNCT
ejpam-6087	273	1	genetic	genetic	ADJ
ejpam-6087	273	2	algorithm	algorithm	NOUN
ejpam-6087	273	3	gas	gas	NOUN
ejpam-6087	273	4	simulate	simulate	VERB
ejpam-6087	273	5	the	the	DET
ejpam-6087	273	6	process	process	NOUN
ejpam-6087	273	7	of	of	ADP
ejpam-6087	273	8	natural	natural	ADJ
ejpam-6087	273	9	evolution	evolution	NOUN
ejpam-6087	273	10	to	to	PART
ejpam-6087	273	11	solve	solve	VERB
ejpam-6087	273	12	optimization	optimization	NOUN
ejpam-6087	273	13	problems	problem	NOUN
ejpam-6087	273	14	.	.	PUNCT
ejpam-6087	274	1	they	they	PRON
ejpam-6087	274	2	operate	operate	VERB
ejpam-6087	274	3	on	on	ADP
ejpam-6087	274	4	a	a	DET
ejpam-6087	274	5	population	population	NOUN
ejpam-6087	274	6	of	of	ADP
ejpam-6087	274	7	potential	potential	ADJ
ejpam-6087	274	8	solutions	solution	NOUN
ejpam-6087	274	9	and	and	CCONJ
ejpam-6087	274	10	use	use	VERB
ejpam-6087	274	11	evolutionary	evolutionary	ADJ
ejpam-6087	274	12	processes	process	NOUN
ejpam-6087	274	13	such	such	ADJ
ejpam-6087	274	14	as	as	ADP
ejpam-6087	274	15	selection	selection	NOUN
ejpam-6087	274	16	,	,	PUNCT
ejpam-6087	274	17	crossover	crossover	NOUN
ejpam-6087	274	18	,	,	PUNCT
ejpam-6087	274	19	and	and	CCONJ
ejpam-6087	274	20	mutation	mutation	NOUN
ejpam-6087	274	21	to	to	PART
ejpam-6087	274	22	evolve	evolve	VERB
ejpam-6087	274	23	better	well	ADJ
ejpam-6087	274	24	solutions	solution	NOUN
ejpam-6087	274	25	over	over	ADP
ejpam-6087	274	26	generations	generation	NOUN
ejpam-6087	274	27	.	.	PUNCT
ejpam-6087	275	1	•	•	NUM
ejpam-6087	275	2	mathematical	mathematical	ADJ
ejpam-6087	275	3	formulation	formulation	NOUN
ejpam-6087	275	4	the	the	DET
ejpam-6087	275	5	ga	ga	NOUN
ejpam-6087	275	6	-	-	PUNCT
ejpam-6087	275	7	based	base	VERB
ejpam-6087	275	8	feature	feature	NOUN
ejpam-6087	275	9	selection	selection	NOUN
ejpam-6087	275	10	process	process	NOUN
ejpam-6087	275	11	involves	involve	VERB
ejpam-6087	275	12	the	the	DET
ejpam-6087	275	13	following	follow	VERB
ejpam-6087	275	14	components	component	NOUN
ejpam-6087	275	15	:	:	PUNCT
ejpam-6087	275	16	•	•	NUM
ejpam-6087	275	17	representation	representation	NOUN
ejpam-6087	275	18	each	each	DET
ejpam-6087	275	19	individual	individual	NOUN
ejpam-6087	275	20	in	in	ADP
ejpam-6087	275	21	the	the	DET
ejpam-6087	275	22	ga	ga	PROPN
ejpam-6087	275	23	population	population	NOUN
ejpam-6087	275	24	represents	represent	VERB
ejpam-6087	275	25	a	a	DET
ejpam-6087	275	26	subset	subset	NOUN
ejpam-6087	275	27	of	of	ADP
ejpam-6087	275	28	features	feature	NOUN
ejpam-6087	275	29	using	use	VERB
ejpam-6087	275	30	a	a	DET
ejpam-6087	275	31	binary	binary	ADJ
ejpam-6087	275	32	vector	vector	NOUN
ejpam-6087	275	33	.	.	PUNCT
ejpam-6087	276	1	if	if	SCONJ
ejpam-6087	276	2	the	the	DET
ejpam-6087	276	3	i	i	NOUN
ejpam-6087	276	4	-	-	PUNCT
ejpam-6087	276	5	th	th	X
ejpam-6087	276	6	bit	bit	NOUN
ejpam-6087	276	7	of	of	ADP
ejpam-6087	276	8	the	the	DET
ejpam-6087	276	9	vector	vector	NOUN
ejpam-6087	276	10	is	be	AUX
ejpam-6087	276	11	1	1	NUM
ejpam-6087	276	12	,	,	PUNCT
ejpam-6087	276	13	the	the	DET
ejpam-6087	276	14	corresponding	corresponding	ADJ
ejpam-6087	276	15	feature	feature	NOUN
ejpam-6087	276	16	is	be	AUX
ejpam-6087	276	17	included	include	VERB
ejpam-6087	276	18	in	in	ADP
ejpam-6087	276	19	the	the	DET
ejpam-6087	276	20	subset	subset	NOUN
ejpam-6087	276	21	;	;	PUNCT
ejpam-6087	276	22	otherwise	otherwise	ADV
ejpam-6087	276	23	,	,	PUNCT
ejpam-6087	276	24	it	it	PRON
ejpam-6087	276	25	is	be	AUX
ejpam-6087	276	26	excluded	exclude	VERB
ejpam-6087	276	27	.	.	PUNCT
ejpam-6087	277	1	let	let	VERB
ejpam-6087	277	2	x	x	PRON
ejpam-6087	277	3	be	be	AUX
ejpam-6087	277	4	the	the	DET
ejpam-6087	277	5	feature	feature	NOUN
ejpam-6087	277	6	matrix	matrix	NOUN
ejpam-6087	277	7	with	with	ADP
ejpam-6087	277	8	n	n	NOUN
ejpam-6087	277	9	features	feature	NOUN
ejpam-6087	277	10	and	and	CCONJ
ejpam-6087	277	11	y	y	X
ejpam-6087	277	12	the	the	DET
ejpam-6087	277	13	target	target	NOUN
ejpam-6087	277	14	variable	variable	NOUN
ejpam-6087	277	15	.	.	PUNCT
ejpam-6087	278	1	each	each	DET
ejpam-6087	278	2	individual	individual	NOUN
ejpam-6087	278	3	x	x	X
ejpam-6087	278	4	can	can	AUX
ejpam-6087	278	5	be	be	AUX
ejpam-6087	278	6	represented	represent	VERB
ejpam-6087	278	7	as	as	ADP
ejpam-6087	278	8	:	:	PUNCT
ejpam-6087	278	9	x	x	X
ejpam-6087	278	10	=	=	PUNCT
ejpam-6087	279	1	[	[	X
ejpam-6087	279	2	x1	x1	X
ejpam-6087	279	3	,	,	PUNCT
ejpam-6087	279	4	x2	x2	PROPN
ejpam-6087	279	5	,	,	PUNCT
ejpam-6087	279	6	.	.	PUNCT
ejpam-6087	279	7	.	.	PUNCT
ejpam-6087	280	1	.	.	PUNCT
ejpam-6087	281	1	,	,	PUNCT
ejpam-6087	281	2	xn	xn	PROPN
ejpam-6087	281	3	]	]	X
ejpam-6087	281	4	(	(	PUNCT
ejpam-6087	281	5	10	10	NUM
ejpam-6087	281	6	)	)	PUNCT
ejpam-6087	281	7	where	where	SCONJ
ejpam-6087	281	8	xi	xi	PROPN
ejpam-6087	281	9	∈	∈	PROPN
ejpam-6087	281	10	{	{	PUNCT
ejpam-6087	281	11	0	0	NUM
ejpam-6087	281	12	,	,	PUNCT
ejpam-6087	281	13	1	1	NUM
ejpam-6087	281	14	}	}	PUNCT
ejpam-6087	281	15	.	.	PUNCT
ejpam-6087	282	1	•	•	NUM
ejpam-6087	282	2	fitness	fitness	NOUN
ejpam-6087	282	3	function	function	NOUN
ejpam-6087	282	4	the	the	DET
ejpam-6087	282	5	fitness	fitness	NOUN
ejpam-6087	282	6	function	function	NOUN
ejpam-6087	282	7	evaluates	evaluate	VERB
ejpam-6087	282	8	the	the	DET
ejpam-6087	282	9	quality	quality	NOUN
ejpam-6087	282	10	of	of	ADP
ejpam-6087	282	11	the	the	DET
ejpam-6087	282	12	subset	subset	NOUN
ejpam-6087	282	13	of	of	ADP
ejpam-6087	282	14	features	feature	NOUN
ejpam-6087	282	15	.	.	PUNCT
ejpam-6087	283	1	in	in	ADP
ejpam-6087	283	2	our	our	PRON
ejpam-6087	283	3	case	case	NOUN
ejpam-6087	283	4	,	,	PUNCT
ejpam-6087	283	5	the	the	DET
ejpam-6087	283	6	fitness	fitness	NOUN
ejpam-6087	283	7	function	function	NOUN
ejpam-6087	283	8	is	be	AUX
ejpam-6087	283	9	the	the	DET
ejpam-6087	283	10	accuracy	accuracy	NOUN
ejpam-6087	283	11	of	of	ADP
ejpam-6087	283	12	a	a	DET
ejpam-6087	283	13	classifier	classifier	NOUN
ejpam-6087	283	14	(	(	PUNCT
ejpam-6087	283	15	random	random	ADJ
ejpam-6087	283	16	forest	forest	NOUN
ejpam-6087	283	17	in	in	ADP
ejpam-6087	283	18	this	this	DET
ejpam-6087	283	19	work	work	NOUN
ejpam-6087	283	20	)	)	PUNCT
ejpam-6087	283	21	trained	train	VERB
ejpam-6087	283	22	on	on	ADP
ejpam-6087	283	23	the	the	DET
ejpam-6087	283	24	selected	select	VERB
ejpam-6087	283	25	features	feature	NOUN
ejpam-6087	283	26	:	:	PUNCT
ejpam-6087	283	27	f(x	f(x	PROPN
ejpam-6087	283	28	)	)	PUNCT
ejpam-6087	283	29	=	=	SYM
ejpam-6087	284	1	accuracy(classifier(xx	accuracy(classifier(xx	PROPN
ejpam-6087	284	2	,	,	PUNCT
ejpam-6087	284	3	y	y	NOUN
ejpam-6087	284	4	)	)	PUNCT
ejpam-6087	284	5	)	)	PUNCT
ejpam-6087	284	6	(	(	PUNCT
ejpam-6087	284	7	11	11	NUM
ejpam-6087	284	8	)	)	PUNCT
ejpam-6087	284	9	where	where	SCONJ
ejpam-6087	284	10	xx	xx	PRON
ejpam-6087	284	11	denotes	denote	VERB
ejpam-6087	284	12	the	the	DET
ejpam-6087	284	13	feature	feature	NOUN
ejpam-6087	284	14	matrix	matrix	NOUN
ejpam-6087	284	15	with	with	ADP
ejpam-6087	284	16	features	feature	NOUN
ejpam-6087	284	17	selected	select	VERB
ejpam-6087	284	18	according	accord	VERB
ejpam-6087	284	19	to	to	ADP
ejpam-6087	284	20	x.	x.	NOUN
ejpam-6087	284	21	•	•	NOUN
ejpam-6087	284	22	genetic	genetic	ADJ
ejpam-6087	284	23	algorithm	algorithm	NOUN
ejpam-6087	284	24	operations	operation	NOUN
ejpam-6087	284	25	the	the	DET
ejpam-6087	284	26	ga	ga	PROPN
ejpam-6087	284	27	operations	operation	NOUN
ejpam-6087	284	28	are	be	AUX
ejpam-6087	284	29	as	as	SCONJ
ejpam-6087	284	30	follows	follow	VERB
ejpam-6087	284	31	:	:	PUNCT
ejpam-6087	284	32	(	(	PUNCT
ejpam-6087	284	33	i	i	NOUN
ejpam-6087	284	34	)	)	PUNCT
ejpam-6087	284	35	initialization	initialization	NOUN
ejpam-6087	284	36	:	:	PUNCT
ejpam-6087	284	37	generate	generate	VERB
ejpam-6087	284	38	an	an	DET
ejpam-6087	284	39	initial	initial	ADJ
ejpam-6087	284	40	population	population	NOUN
ejpam-6087	284	41	of	of	ADP
ejpam-6087	284	42	binary	binary	ADJ
ejpam-6087	284	43	vectors	vector	NOUN
ejpam-6087	284	44	.	.	PUNCT
ejpam-6087	285	1	(	(	PUNCT
ejpam-6087	285	2	ii	ii	NOUN
ejpam-6087	285	3	)	)	PUNCT
ejpam-6087	285	4	selection	selection	NOUN
ejpam-6087	285	5	:	:	PUNCT
ejpam-6087	285	6	select	select	ADJ
ejpam-6087	285	7	individuals	individual	NOUN
ejpam-6087	285	8	based	base	VERB
ejpam-6087	285	9	on	on	ADP
ejpam-6087	285	10	their	their	PRON
ejpam-6087	285	11	fitness	fitness	NOUN
ejpam-6087	285	12	using	use	VERB
ejpam-6087	285	13	tournament	tournament	NOUN
ejpam-6087	285	14	selection	selection	NOUN
ejpam-6087	285	15	.	.	PUNCT
ejpam-6087	286	1	(	(	PUNCT
ejpam-6087	286	2	iii	iii	X
ejpam-6087	286	3	)	)	PUNCT
ejpam-6087	286	4	crossover	crossover	NOUN
ejpam-6087	286	5	:	:	PUNCT
ejpam-6087	286	6	create	create	VERB
ejpam-6087	286	7	new	new	ADJ
ejpam-6087	286	8	individuals	individual	NOUN
ejpam-6087	286	9	by	by	ADP
ejpam-6087	286	10	exchanging	exchange	VERB
ejpam-6087	286	11	segments	segment	NOUN
ejpam-6087	286	12	of	of	ADP
ejpam-6087	286	13	binary	binary	ADJ
ejpam-6087	286	14	vectors	vector	NOUN
ejpam-6087	286	15	between	between	ADP
ejpam-6087	286	16	parents	parent	NOUN
ejpam-6087	286	17	.	.	PUNCT
ejpam-6087	287	1	(	(	PUNCT
ejpam-6087	287	2	iv	iv	X
ejpam-6087	287	3	)	)	PUNCT
ejpam-6087	287	4	mutation	mutation	NOUN
ejpam-6087	287	5	:	:	PUNCT
ejpam-6087	287	6	introduce	introduce	VERB
ejpam-6087	287	7	variability	variability	NOUN
ejpam-6087	287	8	by	by	ADP
ejpam-6087	287	9	flipping	flip	VERB
ejpam-6087	287	10	bits	bit	NOUN
ejpam-6087	287	11	in	in	ADP
ejpam-6087	287	12	the	the	DET
ejpam-6087	287	13	binary	binary	ADJ
ejpam-6087	287	14	vectors	vector	NOUN
ejpam-6087	287	15	.	.	PUNCT
ejpam-6087	288	1	(	(	PUNCT
ejpam-6087	288	2	v	v	NOUN
ejpam-6087	288	3	)	)	PUNCT
ejpam-6087	288	4	replacement	replacement	NOUN
ejpam-6087	288	5	:	:	PUNCT
ejpam-6087	288	6	form	form	VERB
ejpam-6087	288	7	the	the	DET
ejpam-6087	288	8	next	next	ADJ
ejpam-6087	288	9	generation	generation	NOUN
ejpam-6087	288	10	of	of	ADP
ejpam-6087	288	11	the	the	DET
ejpam-6087	288	12	population	population	NOUN
ejpam-6087	288	13	.	.	PUNCT
ejpam-6087	289	1	m.	m.	NOUN
ejpam-6087	289	2	al	al	PROPN
ejpam-6087	289	3	-	-	PUNCT
ejpam-6087	289	4	tarawneh	tarawneh	PROPN
ejpam-6087	289	5	et	et	PROPN
ejpam-6087	289	6	al	al	PROPN
ejpam-6087	289	7	.	.	PUNCT
ejpam-6087	289	8	/	/	SYM
ejpam-6087	289	9	eur	eur	PROPN
ejpam-6087	289	10	.	.	PUNCT
ejpam-6087	290	1	j.	j.	PROPN
ejpam-6087	290	2	pure	pure	PROPN
ejpam-6087	290	3	appl	appl	PROPN
ejpam-6087	290	4	.	.	PROPN
ejpam-6087	290	5	math	math	PROPN
ejpam-6087	290	6	,	,	PUNCT
ejpam-6087	290	7	18	18	NUM
ejpam-6087	290	8	(	(	PUNCT
ejpam-6087	290	9	2	2	NUM
ejpam-6087	290	10	)	)	PUNCT
ejpam-6087	290	11	(	(	PUNCT
ejpam-6087	290	12	2025	2025	NUM
ejpam-6087	290	13	)	)	PUNCT
ejpam-6087	290	14	,	,	PUNCT
ejpam-6087	290	15	6087	6087	NUM
ejpam-6087	290	16	16	16	NUM
ejpam-6087	290	17	of	of	ADP
ejpam-6087	290	18	39	39	NUM
ejpam-6087	290	19	3.5	3.5	NUM
ejpam-6087	290	20	.	.	PUNCT
ejpam-6087	291	1	performance	performance	NOUN
ejpam-6087	291	2	measures	measure	NOUN
ejpam-6087	291	3	in	in	ADP
ejpam-6087	291	4	assessing	assess	VERB
ejpam-6087	291	5	the	the	DET
ejpam-6087	291	6	performance	performance	NOUN
ejpam-6087	291	7	of	of	ADP
ejpam-6087	291	8	the	the	DET
ejpam-6087	291	9	previously	previously	ADV
ejpam-6087	291	10	outlined	outline	VERB
ejpam-6087	291	11	classification	classification	NOUN
ejpam-6087	291	12	ml	ml	NOUN
ejpam-6087	291	13	methods	method	NOUN
ejpam-6087	291	14	for	for	ADP
ejpam-6087	291	15	the	the	DET
ejpam-6087	291	16	detection	detection	NOUN
ejpam-6087	291	17	of	of	ADP
ejpam-6087	291	18	fake	fake	ADJ
ejpam-6087	291	19	news	news	NOUN
ejpam-6087	291	20	in	in	ADP
ejpam-6087	291	21	tweets	tweet	NOUN
ejpam-6087	291	22	,	,	PUNCT
ejpam-6087	291	23	it	it	PRON
ejpam-6087	291	24	is	be	AUX
ejpam-6087	291	25	crucial	crucial	ADJ
ejpam-6087	291	26	to	to	PART
ejpam-6087	291	27	evaluate	evaluate	VERB
ejpam-6087	291	28	their	their	PRON
ejpam-6087	291	29	accuracy	accuracy	NOUN
ejpam-6087	291	30	,	,	PUNCT
ejpam-6087	291	31	recall	recall	NOUN
ejpam-6087	291	32	,	,	PUNCT
ejpam-6087	291	33	precision	precision	NOUN
ejpam-6087	291	34	,	,	PUNCT
ejpam-6087	291	35	f1	f1	NOUN
ejpam-6087	291	36	score	score	NOUN
ejpam-6087	291	37	and	and	CCONJ
ejpam-6087	291	38	auc	auc	NOUN
ejpam-6087	291	39	-	-	PUNCT
ejpam-6087	291	40	roc	roc	NOUN
ejpam-6087	291	41	.	.	PUNCT
ejpam-6087	292	1	these	these	DET
ejpam-6087	292	2	metrics	metric	NOUN
ejpam-6087	292	3	provide	provide	VERB
ejpam-6087	292	4	valuable	valuable	ADJ
ejpam-6087	292	5	insights	insight	NOUN
ejpam-6087	292	6	into	into	ADP
ejpam-6087	292	7	the	the	DET
ejpam-6087	292	8	model	model	NOUN
ejpam-6087	292	9	’s	’s	PART
ejpam-6087	292	10	ability	ability	NOUN
ejpam-6087	292	11	to	to	PART
ejpam-6087	292	12	correctly	correctly	ADV
ejpam-6087	292	13	classify	classify	VERB
ejpam-6087	292	14	tweets	tweet	NOUN
ejpam-6087	292	15	as	as	ADP
ejpam-6087	292	16	either	either	CCONJ
ejpam-6087	292	17	real	real	ADJ
ejpam-6087	292	18	or	or	CCONJ
ejpam-6087	292	19	fake	fake	ADJ
ejpam-6087	292	20	,	,	PUNCT
ejpam-6087	292	21	detect	detect	VERB
ejpam-6087	292	22	relevant	relevant	ADJ
ejpam-6087	292	23	instances	instance	NOUN
ejpam-6087	292	24	,	,	PUNCT
ejpam-6087	292	25	balance	balance	NOUN
ejpam-6087	292	26	between	between	ADP
ejpam-6087	292	27	precision	precision	NOUN
ejpam-6087	292	28	and	and	CCONJ
ejpam-6087	292	29	recall	recall	NOUN
ejpam-6087	292	30	,	,	PUNCT
ejpam-6087	292	31	and	and	CCONJ
ejpam-6087	292	32	evaluate	evaluate	VERB
ejpam-6087	292	33	the	the	DET
ejpam-6087	292	34	overall	overall	ADJ
ejpam-6087	292	35	performance	performance	NOUN
ejpam-6087	292	36	of	of	ADP
ejpam-6087	292	37	the	the	DET
ejpam-6087	292	38	classification	classification	NOUN
ejpam-6087	292	39	model	model	NOUN
ejpam-6087	292	40	.	.	PUNCT
ejpam-6087	293	1	given	give	VERB
ejpam-6087	293	2	the	the	DET
ejpam-6087	293	3	potential	potential	ADJ
ejpam-6087	293	4	impact	impact	NOUN
ejpam-6087	293	5	of	of	ADP
ejpam-6087	293	6	false	false	ADJ
ejpam-6087	293	7	positives	positive	NOUN
ejpam-6087	293	8	where	where	SCONJ
ejpam-6087	293	9	legitimate	legitimate	ADJ
ejpam-6087	293	10	news	news	NOUN
ejpam-6087	293	11	can	can	AUX
ejpam-6087	293	12	be	be	AUX
ejpam-6087	293	13	incorrectly	incorrectly	ADV
ejpam-6087	293	14	labeled	label	VERB
ejpam-6087	293	15	fake	fake	ADJ
ejpam-6087	293	16	it	it	PRON
ejpam-6087	293	17	is	be	AUX
ejpam-6087	293	18	particularly	particularly	ADV
ejpam-6087	293	19	important	important	ADJ
ejpam-6087	293	20	to	to	PART
ejpam-6087	293	21	evaluate	evaluate	VERB
ejpam-6087	293	22	models	model	NOUN
ejpam-6087	293	23	not	not	PART
ejpam-6087	293	24	only	only	ADV
ejpam-6087	293	25	based	base	VERB
ejpam-6087	293	26	on	on	ADP
ejpam-6087	293	27	overall	overall	ADJ
ejpam-6087	293	28	accuracy	accuracy	NOUN
ejpam-6087	293	29	,	,	PUNCT
ejpam-6087	293	30	but	but	CCONJ
ejpam-6087	293	31	also	also	ADV
ejpam-6087	293	32	in	in	ADP
ejpam-6087	293	33	terms	term	NOUN
ejpam-6087	293	34	of	of	ADP
ejpam-6087	293	35	their	their	PRON
ejpam-6087	293	36	precision	precision	NOUN
ejpam-6087	293	37	and	and	CCONJ
ejpam-6087	293	38	recall	recall	NOUN
ejpam-6087	293	39	.	.	PUNCT
ejpam-6087	294	1	high	high	ADJ
ejpam-6087	294	2	precision	precision	NOUN
ejpam-6087	294	3	helps	help	VERB
ejpam-6087	294	4	minimize	minimize	VERB
ejpam-6087	294	5	false	false	ADJ
ejpam-6087	294	6	positives	positive	NOUN
ejpam-6087	294	7	,	,	PUNCT
ejpam-6087	294	8	which	which	PRON
ejpam-6087	294	9	is	be	AUX
ejpam-6087	294	10	crucial	crucial	ADJ
ejpam-6087	294	11	for	for	ADP
ejpam-6087	294	12	maintaining	maintain	VERB
ejpam-6087	294	13	trust	trust	NOUN
ejpam-6087	294	14	and	and	CCONJ
ejpam-6087	294	15	preventing	prevent	VERB
ejpam-6087	294	16	the	the	DET
ejpam-6087	294	17	suppression	suppression	NOUN
ejpam-6087	294	18	of	of	ADP
ejpam-6087	294	19	valid	valid	ADJ
ejpam-6087	294	20	content	content	NOUN
ejpam-6087	294	21	.	.	PUNCT
ejpam-6087	295	1	in	in	ADP
ejpam-6087	295	2	contrast	contrast	NOUN
ejpam-6087	295	3	,	,	PUNCT
ejpam-6087	295	4	high	high	ADJ
ejpam-6087	295	5	recall	recall	NOUN
ejpam-6087	295	6	ensures	ensure	VERB
ejpam-6087	295	7	that	that	SCONJ
ejpam-6087	295	8	actual	actual	ADJ
ejpam-6087	295	9	instances	instance	NOUN
ejpam-6087	295	10	of	of	ADP
ejpam-6087	295	11	fake	fake	ADJ
ejpam-6087	295	12	news	news	NOUN
ejpam-6087	295	13	are	be	AUX
ejpam-6087	295	14	effectively	effectively	ADV
ejpam-6087	295	15	identified	identify	VERB
ejpam-6087	295	16	.	.	PUNCT
ejpam-6087	296	1	the	the	DET
ejpam-6087	296	2	combination	combination	NOUN
ejpam-6087	296	3	of	of	ADP
ejpam-6087	296	4	these	these	DET
ejpam-6087	296	5	metrics	metric	NOUN
ejpam-6087	296	6	supports	support	VERB
ejpam-6087	296	7	a	a	DET
ejpam-6087	296	8	balanced	balanced	ADJ
ejpam-6087	296	9	and	and	CCONJ
ejpam-6087	296	10	context	context	NOUN
ejpam-6087	296	11	-	-	PUNCT
ejpam-6087	296	12	aware	aware	ADJ
ejpam-6087	296	13	interpretation	interpretation	NOUN
ejpam-6087	296	14	of	of	ADP
ejpam-6087	296	15	the	the	DET
ejpam-6087	296	16	behavior	behavior	NOUN
ejpam-6087	296	17	of	of	ADP
ejpam-6087	296	18	the	the	DET
ejpam-6087	296	19	model	model	NOUN
ejpam-6087	296	20	,	,	PUNCT
ejpam-6087	296	21	allowing	allow	VERB
ejpam-6087	296	22	stakeholders	stakeholder	NOUN
ejpam-6087	296	23	to	to	PART
ejpam-6087	296	24	prioritize	prioritize	VERB
ejpam-6087	296	25	detection	detection	NOUN
ejpam-6087	296	26	objectives	objective	NOUN
ejpam-6087	296	27	based	base	VERB
ejpam-6087	296	28	on	on	ADP
ejpam-6087	296	29	specific	specific	ADJ
ejpam-6087	296	30	deployment	deployment	NOUN
ejpam-6087	296	31	needs	need	NOUN
ejpam-6087	296	32	.	.	PUNCT
ejpam-6087	297	1	•	•	NUM
ejpam-6087	297	2	accuracy	accuracy	NOUN
ejpam-6087	297	3	is	be	AUX
ejpam-6087	297	4	a	a	DET
ejpam-6087	297	5	measure	measure	NOUN
ejpam-6087	297	6	of	of	ADP
ejpam-6087	297	7	how	how	SCONJ
ejpam-6087	297	8	well	well	ADV
ejpam-6087	297	9	a	a	DET
ejpam-6087	297	10	model	model	NOUN
ejpam-6087	297	11	is	be	AUX
ejpam-6087	297	12	able	able	ADJ
ejpam-6087	297	13	to	to	PART
ejpam-6087	297	14	correctly	correctly	ADV
ejpam-6087	297	15	classify	classify	VERB
ejpam-6087	297	16	tweets	tweet	NOUN
ejpam-6087	297	17	as	as	ADP
ejpam-6087	297	18	real	real	ADJ
ejpam-6087	297	19	or	or	CCONJ
ejpam-6087	297	20	fake	fake	ADJ
ejpam-6087	297	21	.	.	PUNCT
ejpam-6087	298	1	it	it	PRON
ejpam-6087	298	2	is	be	AUX
ejpam-6087	298	3	calculated	calculate	VERB
ejpam-6087	298	4	as	as	ADP
ejpam-6087	298	5	the	the	DET
ejpam-6087	298	6	proportion	proportion	NOUN
ejpam-6087	298	7	of	of	ADP
ejpam-6087	298	8	correctly	correctly	ADV
ejpam-6087	298	9	classified	classified	ADJ
ejpam-6087	298	10	instances	instance	NOUN
ejpam-6087	298	11	(	(	PUNCT
ejpam-6087	298	12	both	both	CCONJ
ejpam-6087	298	13	real	real	ADJ
ejpam-6087	298	14	and	and	CCONJ
ejpam-6087	298	15	fake	fake	ADJ
ejpam-6087	298	16	tweets	tweet	NOUN
ejpam-6087	298	17	)	)	PUNCT
ejpam-6087	298	18	out	out	ADP
ejpam-6087	298	19	of	of	ADP
ejpam-6087	298	20	the	the	DET
ejpam-6087	298	21	total	total	ADJ
ejpam-6087	298	22	number	number	NOUN
ejpam-6087	298	23	of	of	ADP
ejpam-6087	298	24	instances	instance	NOUN
ejpam-6087	298	25	.	.	PUNCT
ejpam-6087	299	1	accuracy	accuracy	NOUN
ejpam-6087	299	2	=	=	PRON
ejpam-6087	299	3	tp	tp	X
ejpam-6087	299	4	+	+	CCONJ
ejpam-6087	299	5	tn	tn	NOUN
ejpam-6087	299	6	tp	tp	ADP
ejpam-6087	299	7	+	+	CCONJ
ejpam-6087	299	8	tn	tn	PROPN
ejpam-6087	300	1	+	+	NUM
ejpam-6087	300	2	fp	fp	PROPN
ejpam-6087	300	3	+	+	NUM
ejpam-6087	300	4	fn	fn	PROPN
ejpam-6087	300	5	(	(	PUNCT
ejpam-6087	300	6	12	12	NUM
ejpam-6087	300	7	)	)	PUNCT
ejpam-6087	301	1	where	where	SCONJ
ejpam-6087	301	2	:	:	PUNCT
ejpam-6087	301	3	–	–	PUNCT
ejpam-6087	301	4	tp	tp	ADP
ejpam-6087	301	5	(	(	PUNCT
ejpam-6087	301	6	true	true	ADJ
ejpam-6087	301	7	positives	positive	NOUN
ejpam-6087	301	8	)	)	PUNCT
ejpam-6087	301	9	refers	refer	VERB
ejpam-6087	301	10	to	to	ADP
ejpam-6087	301	11	the	the	DET
ejpam-6087	301	12	number	number	NOUN
ejpam-6087	301	13	of	of	ADP
ejpam-6087	301	14	fake	fake	ADJ
ejpam-6087	301	15	tweets	tweet	NOUN
ejpam-6087	301	16	correctly	correctly	ADV
ejpam-6087	301	17	classified	classify	VERB
ejpam-6087	301	18	as	as	ADP
ejpam-6087	301	19	fake	fake	ADJ
ejpam-6087	301	20	.	.	PUNCT
ejpam-6087	302	1	–	–	PUNCT
ejpam-6087	302	2	tn	tn	PROPN
ejpam-6087	302	3	(	(	PUNCT
ejpam-6087	302	4	true	true	ADJ
ejpam-6087	302	5	negatives	negative	NOUN
ejpam-6087	302	6	)	)	PUNCT
ejpam-6087	302	7	refers	refer	VERB
ejpam-6087	302	8	to	to	ADP
ejpam-6087	302	9	the	the	DET
ejpam-6087	302	10	number	number	NOUN
ejpam-6087	302	11	of	of	ADP
ejpam-6087	302	12	real	real	ADJ
ejpam-6087	302	13	tweets	tweet	NOUN
ejpam-6087	302	14	correctly	correctly	ADV
ejpam-6087	302	15	classified	classify	VERB
ejpam-6087	302	16	as	as	ADP
ejpam-6087	302	17	real	real	ADJ
ejpam-6087	302	18	.	.	PUNCT
ejpam-6087	303	1	–	–	PUNCT
ejpam-6087	303	2	fp	fp	X
ejpam-6087	303	3	(	(	PUNCT
ejpam-6087	303	4	false	false	ADJ
ejpam-6087	303	5	positives	positive	NOUN
ejpam-6087	303	6	)	)	PUNCT
ejpam-6087	303	7	refers	refer	VERB
ejpam-6087	303	8	to	to	ADP
ejpam-6087	303	9	the	the	DET
ejpam-6087	303	10	number	number	NOUN
ejpam-6087	303	11	of	of	ADP
ejpam-6087	303	12	real	real	ADJ
ejpam-6087	303	13	tweets	tweet	NOUN
ejpam-6087	303	14	incorrectly	incorrectly	ADV
ejpam-6087	303	15	classified	classify	VERB
ejpam-6087	303	16	as	as	ADP
ejpam-6087	303	17	fake	fake	ADJ
ejpam-6087	303	18	.	.	PUNCT
ejpam-6087	304	1	–	–	PUNCT
ejpam-6087	304	2	fn	fn	NOUN
ejpam-6087	304	3	(	(	PUNCT
ejpam-6087	304	4	false	false	ADJ
ejpam-6087	304	5	negatives	negative	NOUN
ejpam-6087	304	6	)	)	PUNCT
ejpam-6087	304	7	refers	refer	VERB
ejpam-6087	304	8	to	to	ADP
ejpam-6087	304	9	the	the	DET
ejpam-6087	304	10	number	number	NOUN
ejpam-6087	304	11	of	of	ADP
ejpam-6087	304	12	fake	fake	ADJ
ejpam-6087	304	13	tweets	tweet	NOUN
ejpam-6087	304	14	incorrectly	incorrectly	ADV
ejpam-6087	304	15	classified	classify	VERB
ejpam-6087	304	16	as	as	ADP
ejpam-6087	304	17	real	real	ADJ
ejpam-6087	304	18	.	.	PUNCT
ejpam-6087	305	1	•	•	NUM
ejpam-6087	305	2	precision	precision	NOUN
ejpam-6087	305	3	measures	measure	VERB
ejpam-6087	305	4	the	the	DET
ejpam-6087	305	5	model	model	NOUN
ejpam-6087	305	6	’s	’s	PART
ejpam-6087	305	7	ability	ability	NOUN
ejpam-6087	305	8	to	to	PART
ejpam-6087	305	9	avoid	avoid	VERB
ejpam-6087	305	10	false	false	ADJ
ejpam-6087	305	11	positives	positive	NOUN
ejpam-6087	305	12	,	,	PUNCT
ejpam-6087	305	13	focusing	focus	VERB
ejpam-6087	305	14	on	on	ADP
ejpam-6087	305	15	how	how	SCONJ
ejpam-6087	305	16	many	many	ADJ
ejpam-6087	305	17	of	of	ADP
ejpam-6087	305	18	the	the	DET
ejpam-6087	305	19	tweets	tweet	NOUN
ejpam-6087	305	20	classified	classify	VERB
ejpam-6087	305	21	as	as	ADP
ejpam-6087	305	22	fake	fake	ADJ
ejpam-6087	305	23	are	be	AUX
ejpam-6087	305	24	indeed	indeed	ADV
ejpam-6087	305	25	fake	fake	ADJ
ejpam-6087	305	26	.	.	PUNCT
ejpam-6087	306	1	it	it	PRON
ejpam-6087	306	2	is	be	AUX
ejpam-6087	306	3	calculated	calculate	VERB
ejpam-6087	306	4	as	as	ADP
ejpam-6087	306	5	the	the	DET
ejpam-6087	306	6	proportion	proportion	NOUN
ejpam-6087	306	7	of	of	ADP
ejpam-6087	306	8	true	true	ADJ
ejpam-6087	306	9	positives	positive	NOUN
ejpam-6087	306	10	out	out	ADP
ejpam-6087	306	11	of	of	ADP
ejpam-6087	306	12	the	the	DET
ejpam-6087	306	13	total	total	ADJ
ejpam-6087	306	14	number	number	NOUN
ejpam-6087	306	15	of	of	ADP
ejpam-6087	306	16	instances	instance	NOUN
ejpam-6087	306	17	classified	classify	VERB
ejpam-6087	306	18	as	as	ADP
ejpam-6087	306	19	fake	fake	ADJ
ejpam-6087	306	20	.	.	PUNCT
ejpam-6087	307	1	precision	precision	NOUN
ejpam-6087	307	2	=	=	PUNCT
ejpam-6087	308	1	tp	tp	NOUN
ejpam-6087	308	2	tp	tp	ADP
ejpam-6087	308	3	+	+	CCONJ
ejpam-6087	308	4	fp	fp	X
ejpam-6087	308	5	(	(	PUNCT
ejpam-6087	308	6	13	13	NUM
ejpam-6087	308	7	)	)	PUNCT
ejpam-6087	308	8	in	in	ADP
ejpam-6087	308	9	the	the	DET
ejpam-6087	308	10	context	context	NOUN
ejpam-6087	308	11	of	of	ADP
ejpam-6087	308	12	fake	fake	ADJ
ejpam-6087	308	13	news	news	NOUN
ejpam-6087	308	14	detection	detection	NOUN
ejpam-6087	308	15	in	in	ADP
ejpam-6087	308	16	tweets	tweet	NOUN
ejpam-6087	308	17	,	,	PUNCT
ejpam-6087	308	18	precision	precision	NOUN
ejpam-6087	308	19	indicates	indicate	VERB
ejpam-6087	308	20	the	the	DET
ejpam-6087	308	21	model	model	NOUN
ejpam-6087	308	22	’s	’s	PART
ejpam-6087	308	23	effectiveness	effectiveness	NOUN
ejpam-6087	308	24	in	in	ADP
ejpam-6087	308	25	identifying	identify	VERB
ejpam-6087	308	26	actual	actual	ADJ
ejpam-6087	308	27	fake	fake	ADJ
ejpam-6087	308	28	tweets	tweet	NOUN
ejpam-6087	308	29	among	among	ADP
ejpam-6087	308	30	those	those	PRON
ejpam-6087	308	31	it	it	PRON
ejpam-6087	308	32	classifies	classify	VERB
ejpam-6087	308	33	as	as	ADV
ejpam-6087	308	34	fake	fake	ADJ
ejpam-6087	308	35	.	.	PUNCT
ejpam-6087	309	1	m.	m.	NOUN
ejpam-6087	309	2	al	al	PROPN
ejpam-6087	309	3	-	-	PUNCT
ejpam-6087	309	4	tarawneh	tarawneh	PROPN
ejpam-6087	309	5	et	et	PROPN
ejpam-6087	309	6	al	al	PROPN
ejpam-6087	309	7	.	.	PUNCT
ejpam-6087	309	8	/	/	SYM
ejpam-6087	309	9	eur	eur	PROPN
ejpam-6087	309	10	.	.	PUNCT
ejpam-6087	310	1	j.	j.	PROPN
ejpam-6087	310	2	pure	pure	PROPN
ejpam-6087	310	3	appl	appl	PROPN
ejpam-6087	310	4	.	.	PROPN
ejpam-6087	310	5	math	math	PROPN
ejpam-6087	310	6	,	,	PUNCT
ejpam-6087	310	7	18	18	NUM
ejpam-6087	310	8	(	(	PUNCT
ejpam-6087	310	9	2	2	NUM
ejpam-6087	310	10	)	)	PUNCT
ejpam-6087	310	11	(	(	PUNCT
ejpam-6087	310	12	2025	2025	NUM
ejpam-6087	310	13	)	)	PUNCT
ejpam-6087	310	14	,	,	PUNCT
ejpam-6087	310	15	6087	6087	NUM
ejpam-6087	310	16	17	17	NUM
ejpam-6087	310	17	of	of	ADP
ejpam-6087	310	18	39	39	NUM
ejpam-6087	310	19	•	•	NOUN
ejpam-6087	310	20	recall	recall	NOUN
ejpam-6087	310	21	measures	measure	NOUN
ejpam-6087	310	22	the	the	DET
ejpam-6087	310	23	proportion	proportion	NOUN
ejpam-6087	310	24	of	of	ADP
ejpam-6087	310	25	actual	actual	ADJ
ejpam-6087	310	26	fake	fake	ADJ
ejpam-6087	310	27	tweets	tweet	NOUN
ejpam-6087	310	28	that	that	PRON
ejpam-6087	310	29	are	be	AUX
ejpam-6087	310	30	correctly	correctly	ADV
ejpam-6087	310	31	identified	identify	VERB
ejpam-6087	310	32	by	by	ADP
ejpam-6087	310	33	the	the	DET
ejpam-6087	310	34	model	model	NOUN
ejpam-6087	310	35	.	.	PUNCT
ejpam-6087	311	1	it	it	PRON
ejpam-6087	311	2	is	be	AUX
ejpam-6087	311	3	calculated	calculate	VERB
ejpam-6087	311	4	by	by	ADP
ejpam-6087	311	5	dividing	divide	VERB
ejpam-6087	311	6	the	the	DET
ejpam-6087	311	7	number	number	NOUN
ejpam-6087	311	8	of	of	ADP
ejpam-6087	311	9	true	true	ADJ
ejpam-6087	311	10	positives	positive	NOUN
ejpam-6087	311	11	by	by	ADP
ejpam-6087	311	12	the	the	DET
ejpam-6087	311	13	sum	sum	NOUN
ejpam-6087	311	14	of	of	ADP
ejpam-6087	311	15	true	true	ADJ
ejpam-6087	311	16	positives	positive	NOUN
ejpam-6087	311	17	and	and	CCONJ
ejpam-6087	311	18	false	false	ADJ
ejpam-6087	311	19	negatives	negative	NOUN
ejpam-6087	311	20	.	.	PUNCT
ejpam-6087	312	1	recall	recall	NOUN
ejpam-6087	312	2	=	=	PROPN
ejpam-6087	313	1	tp	tp	NOUN
ejpam-6087	313	2	tp	tp	ADP
ejpam-6087	313	3	+	+	CCONJ
ejpam-6087	314	1	fn	fn	PROPN
ejpam-6087	314	2	(	(	PUNCT
ejpam-6087	314	3	14	14	NUM
ejpam-6087	314	4	)	)	PUNCT
ejpam-6087	314	5	in	in	ADP
ejpam-6087	314	6	the	the	DET
ejpam-6087	314	7	context	context	NOUN
ejpam-6087	314	8	of	of	ADP
ejpam-6087	314	9	fake	fake	ADJ
ejpam-6087	314	10	news	news	NOUN
ejpam-6087	314	11	detection	detection	NOUN
ejpam-6087	314	12	in	in	ADP
ejpam-6087	314	13	tweets	tweet	NOUN
ejpam-6087	314	14	,	,	PUNCT
ejpam-6087	314	15	recall	recall	PROPN
ejpam-6087	314	16	indicates	indicate	VERB
ejpam-6087	314	17	the	the	DET
ejpam-6087	314	18	model	model	NOUN
ejpam-6087	314	19	’s	’s	PART
ejpam-6087	314	20	ability	ability	NOUN
ejpam-6087	314	21	to	to	PART
ejpam-6087	314	22	detect	detect	VERB
ejpam-6087	314	23	all	all	DET
ejpam-6087	314	24	the	the	DET
ejpam-6087	314	25	fake	fake	ADJ
ejpam-6087	314	26	tweets	tweet	NOUN
ejpam-6087	314	27	that	that	PRON
ejpam-6087	314	28	are	be	AUX
ejpam-6087	314	29	actually	actually	ADV
ejpam-6087	314	30	present	present	ADJ
ejpam-6087	314	31	.	.	PUNCT
ejpam-6087	315	1	•	•	NUM
ejpam-6087	315	2	f1	f1	NOUN
ejpam-6087	315	3	score	score	NOUN
ejpam-6087	315	4	is	be	AUX
ejpam-6087	315	5	a	a	DET
ejpam-6087	315	6	harmonic	harmonic	ADJ
ejpam-6087	315	7	mean	mean	NOUN
ejpam-6087	315	8	of	of	ADP
ejpam-6087	315	9	precision	precision	NOUN
ejpam-6087	315	10	and	and	CCONJ
ejpam-6087	315	11	recall	recall	NOUN
ejpam-6087	315	12	,	,	PUNCT
ejpam-6087	315	13	providing	provide	VERB
ejpam-6087	315	14	a	a	DET
ejpam-6087	315	15	balanced	balanced	ADJ
ejpam-6087	315	16	measure	measure	NOUN
ejpam-6087	315	17	of	of	ADP
ejpam-6087	315	18	the	the	DET
ejpam-6087	315	19	model	model	NOUN
ejpam-6087	315	20	’s	’s	PART
ejpam-6087	315	21	performance	performance	NOUN
ejpam-6087	315	22	in	in	ADP
ejpam-6087	315	23	detecting	detect	VERB
ejpam-6087	315	24	fake	fake	ADJ
ejpam-6087	315	25	tweets	tweet	NOUN
ejpam-6087	315	26	.	.	PUNCT
ejpam-6087	316	1	it	it	PRON
ejpam-6087	316	2	accounts	account	VERB
ejpam-6087	316	3	for	for	ADP
ejpam-6087	316	4	both	both	CCONJ
ejpam-6087	316	5	the	the	DET
ejpam-6087	316	6	precision	precision	NOUN
ejpam-6087	316	7	of	of	ADP
ejpam-6087	316	8	the	the	DET
ejpam-6087	316	9	model	model	NOUN
ejpam-6087	316	10	’s	’s	PART
ejpam-6087	316	11	predictions	prediction	NOUN
ejpam-6087	316	12	and	and	CCONJ
ejpam-6087	316	13	its	its	PRON
ejpam-6087	316	14	ability	ability	NOUN
ejpam-6087	316	15	to	to	PART
ejpam-6087	316	16	capture	capture	VERB
ejpam-6087	316	17	all	all	DET
ejpam-6087	316	18	relevant	relevant	ADJ
ejpam-6087	316	19	instances	instance	NOUN
ejpam-6087	316	20	.	.	PUNCT
ejpam-6087	317	1	f1	f1	NOUN
ejpam-6087	317	2	=	=	SYM
ejpam-6087	317	3	2	2	NUM
ejpam-6087	317	4	·	·	PUNCT
ejpam-6087	317	5	precision	precision	NOUN
ejpam-6087	317	6	·	·	PUNCT
ejpam-6087	317	7	recall	recall	NOUN
ejpam-6087	317	8	precision	precision	NOUN
ejpam-6087	317	9	+	+	CCONJ
ejpam-6087	317	10	recall	recall	NOUN
ejpam-6087	317	11	(	(	PUNCT
ejpam-6087	317	12	15	15	NUM
ejpam-6087	317	13	)	)	PUNCT
ejpam-6087	317	14	the	the	DET
ejpam-6087	317	15	f1	f1	PROPN
ejpam-6087	317	16	score	score	NOUN
ejpam-6087	317	17	is	be	AUX
ejpam-6087	317	18	particularly	particularly	ADV
ejpam-6087	317	19	useful	useful	ADJ
ejpam-6087	317	20	in	in	ADP
ejpam-6087	317	21	the	the	DET
ejpam-6087	317	22	detection	detection	NOUN
ejpam-6087	317	23	of	of	ADP
ejpam-6087	317	24	fake	fake	ADJ
ejpam-6087	317	25	news	news	NOUN
ejpam-6087	317	26	for	for	ADP
ejpam-6087	317	27	tweets	tweet	NOUN
ejpam-6087	317	28	where	where	SCONJ
ejpam-6087	317	29	both	both	DET
ejpam-6087	317	30	false	false	ADJ
ejpam-6087	317	31	positives	positive	NOUN
ejpam-6087	317	32	and	and	CCONJ
ejpam-6087	317	33	false	false	ADJ
ejpam-6087	317	34	negatives	negative	NOUN
ejpam-6087	317	35	have	have	VERB
ejpam-6087	317	36	significant	significant	ADJ
ejpam-6087	317	37	consequences	consequence	NOUN
ejpam-6087	317	38	,	,	PUNCT
ejpam-6087	317	39	and	and	CCONJ
ejpam-6087	317	40	the	the	DET
ejpam-6087	317	41	balance	balance	NOUN
ejpam-6087	317	42	between	between	ADP
ejpam-6087	317	43	precision	precision	NOUN
ejpam-6087	317	44	and	and	CCONJ
ejpam-6087	317	45	recall	recall	NOUN
ejpam-6087	317	46	is	be	AUX
ejpam-6087	317	47	crucial	crucial	ADJ
ejpam-6087	317	48	.	.	PUNCT
ejpam-6087	318	1	•	•	NUM
ejpam-6087	318	2	auc	auc	NOUN
ejpam-6087	318	3	-	-	PUNCT
ejpam-6087	318	4	roc	roc	NOUN
ejpam-6087	318	5	(	(	PUNCT
ejpam-6087	318	6	area	area	NOUN
ejpam-6087	318	7	under	under	ADP
ejpam-6087	318	8	the	the	DET
ejpam-6087	318	9	receiver	receiver	NOUN
ejpam-6087	318	10	operating	operate	VERB
ejpam-6087	318	11	characteristic	characteristic	ADJ
ejpam-6087	318	12	curve	curve	NOUN
ejpam-6087	318	13	)	)	PUNCT
ejpam-6087	318	14	is	be	AUX
ejpam-6087	318	15	a	a	DET
ejpam-6087	318	16	performance	performance	NOUN
ejpam-6087	318	17	measurement	measurement	NOUN
ejpam-6087	318	18	for	for	ADP
ejpam-6087	318	19	classification	classification	NOUN
ejpam-6087	318	20	problems	problem	NOUN
ejpam-6087	318	21	at	at	ADP
ejpam-6087	318	22	various	various	ADJ
ejpam-6087	318	23	thresholds	threshold	NOUN
ejpam-6087	318	24	.	.	PUNCT
ejpam-6087	319	1	it	it	PRON
ejpam-6087	319	2	plots	plot	VERB
ejpam-6087	319	3	the	the	DET
ejpam-6087	319	4	true	true	ADJ
ejpam-6087	319	5	positive	positive	ADJ
ejpam-6087	319	6	rate	rate	NOUN
ejpam-6087	319	7	(	(	PUNCT
ejpam-6087	319	8	recall	recall	NOUN
ejpam-6087	319	9	)	)	PUNCT
ejpam-6087	319	10	against	against	ADP
ejpam-6087	319	11	the	the	DET
ejpam-6087	319	12	false	false	ADJ
ejpam-6087	319	13	positive	positive	ADJ
ejpam-6087	319	14	rate	rate	NOUN
ejpam-6087	319	15	and	and	CCONJ
ejpam-6087	319	16	measures	measure	NOUN
ejpam-6087	319	17	the	the	DET
ejpam-6087	319	18	area	area	NOUN
ejpam-6087	319	19	under	under	ADP
ejpam-6087	319	20	this	this	DET
ejpam-6087	319	21	curve	curve	NOUN
ejpam-6087	319	22	.	.	PUNCT
ejpam-6087	320	1	auc	auc	NOUN
ejpam-6087	320	2	-	-	PUNCT
ejpam-6087	320	3	roc	roc	NOUN
ejpam-6087	320	4	=	=	SYM
ejpam-6087	320	5	∫	∫	PROPN
ejpam-6087	321	1	+	+	PROPN
ejpam-6087	321	2	∞	∞	PROPN
ejpam-6087	321	3	−∞	−∞	X
ejpam-6087	321	4	tpr(x	tpr(x	PROPN
ejpam-6087	321	5	)	)	PUNCT
ejpam-6087	321	6	dfpr(x	dfpr(x	PROPN
ejpam-6087	321	7	)	)	PUNCT
ejpam-6087	321	8	(	(	PUNCT
ejpam-6087	321	9	16	16	NUM
ejpam-6087	321	10	)	)	PUNCT
ejpam-6087	321	11	where	where	SCONJ
ejpam-6087	321	12	:	:	PUNCT
ejpam-6087	321	13	–	–	PUNCT
ejpam-6087	321	14	tpr	tpr	NOUN
ejpam-6087	321	15	(	(	PUNCT
ejpam-6087	321	16	true	true	ADJ
ejpam-6087	321	17	positive	positive	ADJ
ejpam-6087	321	18	rate	rate	NOUN
ejpam-6087	321	19	)	)	PUNCT
ejpam-6087	321	20	is	be	AUX
ejpam-6087	321	21	the	the	DET
ejpam-6087	321	22	recall	recall	NOUN
ejpam-6087	321	23	of	of	ADP
ejpam-6087	321	24	the	the	DET
ejpam-6087	321	25	model	model	NOUN
ejpam-6087	321	26	.	.	PUNCT
ejpam-6087	322	1	–	–	PUNCT
ejpam-6087	322	2	fpr	fpr	INTJ
ejpam-6087	322	3	(	(	PUNCT
ejpam-6087	322	4	false	false	ADJ
ejpam-6087	322	5	positive	positive	ADJ
ejpam-6087	322	6	rate	rate	NOUN
ejpam-6087	322	7	)	)	PUNCT
ejpam-6087	322	8	is	be	AUX
ejpam-6087	322	9	the	the	DET
ejpam-6087	322	10	proportion	proportion	NOUN
ejpam-6087	322	11	of	of	ADP
ejpam-6087	322	12	actual	actual	ADJ
ejpam-6087	322	13	negatives	negative	NOUN
ejpam-6087	322	14	that	that	PRON
ejpam-6087	322	15	are	be	AUX
ejpam-6087	322	16	incorrectly	incorrectly	ADV
ejpam-6087	322	17	classified	classify	VERB
ejpam-6087	322	18	as	as	ADP
ejpam-6087	322	19	positives	positive	NOUN
ejpam-6087	322	20	.	.	PUNCT
ejpam-6087	323	1	in	in	ADP
ejpam-6087	323	2	the	the	DET
ejpam-6087	323	3	context	context	NOUN
ejpam-6087	323	4	of	of	ADP
ejpam-6087	323	5	fake	fake	ADJ
ejpam-6087	323	6	news	news	NOUN
ejpam-6087	323	7	detection	detection	NOUN
ejpam-6087	323	8	in	in	ADP
ejpam-6087	323	9	tweets	tweet	NOUN
ejpam-6087	323	10	,	,	PUNCT
ejpam-6087	323	11	the	the	DET
ejpam-6087	323	12	auc	auc	NOUN
ejpam-6087	323	13	-	-	PUNCT
ejpam-6087	323	14	roc	roc	NOUN
ejpam-6087	323	15	score	score	NOUN
ejpam-6087	323	16	provides	provide	VERB
ejpam-6087	323	17	an	an	DET
ejpam-6087	323	18	aggregate	aggregate	ADJ
ejpam-6087	323	19	measure	measure	NOUN
ejpam-6087	323	20	of	of	ADP
ejpam-6087	323	21	the	the	DET
ejpam-6087	323	22	model	model	NOUN
ejpam-6087	323	23	’s	’s	PART
ejpam-6087	323	24	ability	ability	NOUN
ejpam-6087	323	25	to	to	PART
ejpam-6087	323	26	discriminate	discriminate	VERB
ejpam-6087	323	27	between	between	ADP
ejpam-6087	323	28	fake	fake	ADJ
ejpam-6087	323	29	and	and	CCONJ
ejpam-6087	323	30	real	real	ADJ
ejpam-6087	323	31	tweets	tweet	NOUN
ejpam-6087	323	32	across	across	ADP
ejpam-6087	323	33	all	all	DET
ejpam-6087	323	34	possible	possible	ADJ
ejpam-6087	323	35	classification	classification	NOUN
ejpam-6087	323	36	thresholds	threshold	NOUN
ejpam-6087	323	37	.	.	PUNCT
ejpam-6087	324	1	a	a	DET
ejpam-6087	324	2	higher	high	ADJ
ejpam-6087	324	3	auc	auc	NOUN
ejpam-6087	324	4	-	-	PUNCT
ejpam-6087	324	5	roc	roc	NOUN
ejpam-6087	324	6	score	score	NOUN
ejpam-6087	324	7	indicates	indicate	VERB
ejpam-6087	324	8	a	a	DET
ejpam-6087	324	9	better	well	ADJ
ejpam-6087	324	10	overall	overall	ADJ
ejpam-6087	324	11	performance	performance	NOUN
ejpam-6087	324	12	in	in	ADP
ejpam-6087	324	13	distinguishing	distinguish	VERB
ejpam-6087	324	14	between	between	ADP
ejpam-6087	324	15	fake	fake	ADJ
ejpam-6087	324	16	and	and	CCONJ
ejpam-6087	324	17	real	real	ADJ
ejpam-6087	324	18	tweets	tweet	NOUN
ejpam-6087	324	19	.	.	PUNCT
ejpam-6087	325	1	4	4	X
ejpam-6087	325	2	.	.	NOUN
ejpam-6087	325	3	results	result	NOUN
ejpam-6087	325	4	and	and	CCONJ
ejpam-6087	325	5	analysis	analysis	NOUN
ejpam-6087	325	6	4.1	4.1	NUM
ejpam-6087	325	7	.	.	PUNCT
ejpam-6087	326	1	individual	individual	ADJ
ejpam-6087	326	2	classifiers	classifier	NOUN
ejpam-6087	326	3	results	result	VERB
ejpam-6087	326	4	table	table	NOUN
ejpam-6087	326	5	4	4	NUM
ejpam-6087	326	6	demonstrates	demonstrate	VERB
ejpam-6087	326	7	the	the	DET
ejpam-6087	326	8	performance	performance	NOUN
ejpam-6087	326	9	of	of	ADP
ejpam-6087	326	10	various	various	ADJ
ejpam-6087	326	11	individual	individual	ADJ
ejpam-6087	326	12	ml	ml	NOUN
ejpam-6087	326	13	models	model	NOUN
ejpam-6087	326	14	in	in	ADP
ejpam-6087	326	15	detecting	detect	VERB
ejpam-6087	326	16	fake	fake	ADJ
ejpam-6087	326	17	news	news	NOUN
ejpam-6087	326	18	using	use	VERB
ejpam-6087	326	19	the	the	DET
ejpam-6087	326	20	truthseeker	truthseeker	NOUN
ejpam-6087	326	21	dataset	dataset	NOUN
ejpam-6087	326	22	.	.	PUNCT
ejpam-6087	327	1	the	the	DET
ejpam-6087	327	2	results	result	NOUN
ejpam-6087	327	3	shown	show	VERB
ejpam-6087	327	4	in	in	ADP
ejpam-6087	327	5	this	this	DET
ejpam-6087	327	6	table	table	NOUN
ejpam-6087	327	7	reveal	reveal	VERB
ejpam-6087	327	8	notable	notable	ADJ
ejpam-6087	327	9	m.	m.	NOUN
ejpam-6087	327	10	al	al	PROPN
ejpam-6087	327	11	-	-	PUNCT
ejpam-6087	327	12	tarawneh	tarawneh	PROPN
ejpam-6087	327	13	et	et	PROPN
ejpam-6087	327	14	al	al	PROPN
ejpam-6087	327	15	.	.	PUNCT
ejpam-6087	327	16	/	/	SYM
ejpam-6087	327	17	eur	eur	PROPN
ejpam-6087	327	18	.	.	PUNCT
ejpam-6087	328	1	j.	j.	PROPN
ejpam-6087	328	2	pure	pure	PROPN
ejpam-6087	328	3	appl	appl	PROPN
ejpam-6087	328	4	.	.	PROPN
ejpam-6087	328	5	math	math	PROPN
ejpam-6087	328	6	,	,	PUNCT
ejpam-6087	328	7	18	18	NUM
ejpam-6087	328	8	(	(	PUNCT
ejpam-6087	328	9	2	2	NUM
ejpam-6087	328	10	)	)	PUNCT
ejpam-6087	328	11	(	(	PUNCT
ejpam-6087	328	12	2025	2025	NUM
ejpam-6087	328	13	)	)	PUNCT
ejpam-6087	328	14	,	,	PUNCT
ejpam-6087	328	15	6087	6087	NUM
ejpam-6087	328	16	18	18	NUM
ejpam-6087	328	17	of	of	ADP
ejpam-6087	328	18	39	39	NUM
ejpam-6087	328	19	insights	insight	NOUN
ejpam-6087	328	20	into	into	ADP
ejpam-6087	328	21	the	the	DET
ejpam-6087	328	22	effectiveness	effectiveness	NOUN
ejpam-6087	328	23	of	of	ADP
ejpam-6087	328	24	these	these	DET
ejpam-6087	328	25	models	model	NOUN
ejpam-6087	328	26	in	in	ADP
ejpam-6087	328	27	detecting	detect	VERB
ejpam-6087	328	28	fake	fake	ADJ
ejpam-6087	328	29	news	news	NOUN
ejpam-6087	328	30	.	.	PUNCT
ejpam-6087	329	1	each	each	DET
ejpam-6087	329	2	model	model	NOUN
ejpam-6087	329	3	was	be	AUX
ejpam-6087	329	4	assessed	assess	VERB
ejpam-6087	329	5	based	base	VERB
ejpam-6087	329	6	on	on	ADP
ejpam-6087	329	7	five	five	NUM
ejpam-6087	329	8	key	key	ADJ
ejpam-6087	329	9	metrics	metric	NOUN
ejpam-6087	329	10	:	:	PUNCT
ejpam-6087	329	11	accuracy	accuracy	NOUN
ejpam-6087	329	12	,	,	PUNCT
ejpam-6087	329	13	precision	precision	NOUN
ejpam-6087	329	14	,	,	PUNCT
ejpam-6087	329	15	recall	recall	NOUN
ejpam-6087	329	16	,	,	PUNCT
ejpam-6087	329	17	f1	f1	NOUN
ejpam-6087	329	18	score	score	NOUN
ejpam-6087	329	19	,	,	PUNCT
ejpam-6087	329	20	and	and	CCONJ
ejpam-6087	329	21	auc	auc	NOUN
ejpam-6087	329	22	-	-	PUNCT
ejpam-6087	329	23	roc	roc	NOUN
ejpam-6087	329	24	,	,	PUNCT
ejpam-6087	329	25	which	which	PRON
ejpam-6087	329	26	together	together	ADV
ejpam-6087	329	27	provide	provide	VERB
ejpam-6087	329	28	a	a	DET
ejpam-6087	329	29	comprehensive	comprehensive	ADJ
ejpam-6087	329	30	view	view	NOUN
ejpam-6087	329	31	of	of	ADP
ejpam-6087	329	32	their	their	PRON
ejpam-6087	329	33	predictive	predictive	ADJ
ejpam-6087	329	34	capabilities	capability	NOUN
ejpam-6087	329	35	.	.	PUNCT
ejpam-6087	330	1	in	in	ADP
ejpam-6087	330	2	this	this	DET
ejpam-6087	330	3	regard	regard	NOUN
ejpam-6087	330	4	:	:	PUNCT
ejpam-6087	330	5	table	table	NOUN
ejpam-6087	330	6	4	4	NUM
ejpam-6087	330	7	:	:	PUNCT
ejpam-6087	330	8	performance	performance	NOUN
ejpam-6087	330	9	of	of	ADP
ejpam-6087	330	10	individual	individual	ADJ
ejpam-6087	330	11	classifiers	classifier	NOUN
ejpam-6087	330	12	on	on	ADP
ejpam-6087	330	13	the	the	DET
ejpam-6087	330	14	truthseeker	truthseeker	NOUN
ejpam-6087	330	15	dataset	dataset	NOUN
ejpam-6087	330	16	classifier	classifier	PROPN
ejpam-6087	330	17	accuracy	accuracy	PROPN
ejpam-6087	330	18	precision	precision	NOUN
ejpam-6087	330	19	recall	recall	VERB
ejpam-6087	330	20	f1	f1	NOUN
ejpam-6087	330	21	-	-	PUNCT
ejpam-6087	330	22	score	score	NOUN
ejpam-6087	330	23	auc	auc	NOUN
ejpam-6087	330	24	-	-	PUNCT
ejpam-6087	330	25	roc	roc	NOUN
ejpam-6087	330	26	nb	nb	PROPN
ejpam-6087	330	27	92.969	92.969	NUM
ejpam-6087	330	28	%	%	NOUN
ejpam-6087	330	29	92.996	92.996	NUM
ejpam-6087	330	30	%	%	NOUN
ejpam-6087	330	31	92.969	92.969	NUM
ejpam-6087	330	32	%	%	NOUN
ejpam-6087	330	33	92.971	92.971	NUM
ejpam-6087	330	34	%	%	NOUN
ejpam-6087	330	35	93.414	93.414	NUM
ejpam-6087	330	36	%	%	NOUN
ejpam-6087	330	37	lr	lr	NOUN
ejpam-6087	330	38	95.801	95.801	NUM
ejpam-6087	330	39	%	%	NOUN
ejpam-6087	330	40	95.801	95.801	NUM
ejpam-6087	330	41	%	%	NOUN
ejpam-6087	330	42	95.801	95.801	NUM
ejpam-6087	330	43	%	%	NOUN
ejpam-6087	330	44	95.801	95.801	NUM
ejpam-6087	330	45	%	%	NOUN
ejpam-6087	330	46	95.889	95.889	NUM
ejpam-6087	330	47	%	%	NOUN
ejpam-6087	330	48	knn	knn	NOUN
ejpam-6087	330	49	92.414	92.414	NUM
ejpam-6087	330	50	%	%	NOUN
ejpam-6087	330	51	92.414	92.414	NUM
ejpam-6087	330	52	%	%	NOUN
ejpam-6087	330	53	92.414	92.414	NUM
ejpam-6087	330	54	%	%	NOUN
ejpam-6087	330	55	92.414	92.414	NUM
ejpam-6087	330	56	%	%	NOUN
ejpam-6087	330	57	95.029	95.029	NUM
ejpam-6087	330	58	%	%	NOUN
ejpam-6087	330	59	mlp	mlp	NOUN
ejpam-6087	330	60	95.387	95.387	NUM
ejpam-6087	330	61	%	%	NOUN
ejpam-6087	330	62	95.387	95.387	NUM
ejpam-6087	330	63	%	%	NOUN
ejpam-6087	330	64	95.387	95.387	NUM
ejpam-6087	330	65	%	%	NOUN
ejpam-6087	330	66	95.387	95.387	NUM
ejpam-6087	330	67	%	%	NOUN
ejpam-6087	330	68	95.847	95.847	NUM
ejpam-6087	330	69	%	%	NOUN
ejpam-6087	330	70	svm	svm	NOUN
ejpam-6087	330	71	95.704	95.704	NUM
ejpam-6087	330	72	%	%	NOUN
ejpam-6087	330	73	95.704	95.704	NUM
ejpam-6087	330	74	%	%	NOUN
ejpam-6087	330	75	95.704	95.704	NUM
ejpam-6087	330	76	%	%	NOUN
ejpam-6087	330	77	95.704	95.704	NUM
ejpam-6087	330	78	%	%	NOUN
ejpam-6087	330	79	95.902	95.902	NUM
ejpam-6087	330	80	%	%	NOUN
ejpam-6087	330	81	dt	dt	NOUN
ejpam-6087	330	82	91.006	91.006	NUM
ejpam-6087	330	83	%	%	NOUN
ejpam-6087	330	84	91.007	91.007	NUM
ejpam-6087	330	85	%	%	NOUN
ejpam-6087	330	86	91.006	91.006	NUM
ejpam-6087	330	87	%	%	NOUN
ejpam-6087	330	88	91.006	91.006	NUM
ejpam-6087	330	89	%	%	NOUN
ejpam-6087	330	90	91.002	91.002	NUM
ejpam-6087	330	91	%	%	NOUN
ejpam-6087	330	92	rf	rf	NOUN
ejpam-6087	330	93	95.801	95.801	NUM
ejpam-6087	330	94	%	%	NOUN
ejpam-6087	330	95	95.801	95.801	NUM
ejpam-6087	330	96	%	%	NOUN
ejpam-6087	330	97	95.801	95.801	NUM
ejpam-6087	330	98	%	%	NOUN
ejpam-6087	330	99	95.801	95.801	NUM
ejpam-6087	330	100	%	%	NOUN
ejpam-6087	330	101	95.892	95.892	NUM
ejpam-6087	330	102	%	%	NOUN
ejpam-6087	330	103	•	•	ADP
ejpam-6087	330	104	naive	naive	ADJ
ejpam-6087	330	105	bayes	bayes	NOUN
ejpam-6087	330	106	:	:	PUNCT
ejpam-6087	330	107	nb	nb	PROPN
ejpam-6087	330	108	achieved	achieve	VERB
ejpam-6087	330	109	an	an	DET
ejpam-6087	330	110	accuracy	accuracy	NOUN
ejpam-6087	330	111	of	of	ADP
ejpam-6087	330	112	92.969	92.969	NUM
ejpam-6087	330	113	%	%	NOUN
ejpam-6087	330	114	,	,	PUNCT
ejpam-6087	330	115	with	with	ADP
ejpam-6087	330	116	precision	precision	NOUN
ejpam-6087	330	117	and	and	CCONJ
ejpam-6087	330	118	recall	recall	VERB
ejpam-6087	330	119	close	close	ADV
ejpam-6087	330	120	to	to	ADP
ejpam-6087	330	121	93	93	NUM
ejpam-6087	330	122	%	%	NOUN
ejpam-6087	330	123	,	,	PUNCT
ejpam-6087	330	124	indicating	indicate	VERB
ejpam-6087	330	125	its	its	PRON
ejpam-6087	330	126	ability	ability	NOUN
ejpam-6087	330	127	to	to	PART
ejpam-6087	330	128	consistently	consistently	ADV
ejpam-6087	330	129	identify	identify	VERB
ejpam-6087	330	130	both	both	DET
ejpam-6087	330	131	true	true	ADJ
ejpam-6087	330	132	positives	positive	NOUN
ejpam-6087	330	133	and	and	CCONJ
ejpam-6087	330	134	true	true	ADJ
ejpam-6087	330	135	negatives	negative	NOUN
ejpam-6087	330	136	.	.	PUNCT
ejpam-6087	331	1	its	its	PRON
ejpam-6087	331	2	auc	auc	NOUN
ejpam-6087	331	3	-	-	PUNCT
ejpam-6087	331	4	roc	roc	NOUN
ejpam-6087	331	5	of	of	ADP
ejpam-6087	331	6	93.414	93.414	NUM
ejpam-6087	331	7	%	%	NOUN
ejpam-6087	331	8	reflects	reflect	VERB
ejpam-6087	331	9	a	a	DET
ejpam-6087	331	10	robust	robust	ADJ
ejpam-6087	331	11	performance	performance	NOUN
ejpam-6087	331	12	in	in	ADP
ejpam-6087	331	13	distinguishing	distinguish	VERB
ejpam-6087	331	14	between	between	ADP
ejpam-6087	331	15	fake	fake	ADJ
ejpam-6087	331	16	and	and	CCONJ
ejpam-6087	331	17	real	real	ADJ
ejpam-6087	331	18	news	news	NOUN
ejpam-6087	331	19	.	.	PUNCT
ejpam-6087	332	1	•	•	NUM
ejpam-6087	332	2	logistic	logistic	ADJ
ejpam-6087	332	3	regression	regression	NOUN
ejpam-6087	332	4	:	:	PUNCT
ejpam-6087	332	5	lr	lr	NOUN
ejpam-6087	332	6	demonstrated	demonstrate	VERB
ejpam-6087	332	7	superior	superior	ADJ
ejpam-6087	332	8	performance	performance	NOUN
ejpam-6087	332	9	with	with	ADP
ejpam-6087	332	10	accuracy	accuracy	NOUN
ejpam-6087	332	11	,	,	PUNCT
ejpam-6087	332	12	precision	precision	NOUN
ejpam-6087	332	13	,	,	PUNCT
ejpam-6087	332	14	recall	recall	NOUN
ejpam-6087	332	15	,	,	PUNCT
ejpam-6087	332	16	and	and	CCONJ
ejpam-6087	332	17	f1	f1	ADJ
ejpam-6087	332	18	score	score	NOUN
ejpam-6087	332	19	of	of	ADP
ejpam-6087	332	20	95.801	95.801	NUM
ejpam-6087	332	21	%	%	NOUN
ejpam-6087	332	22	.	.	PUNCT
ejpam-6087	333	1	the	the	DET
ejpam-6087	333	2	auc	auc	NOUN
ejpam-6087	333	3	-	-	PUNCT
ejpam-6087	333	4	roc	roc	NOUN
ejpam-6087	333	5	score	score	NOUN
ejpam-6087	333	6	of	of	ADP
ejpam-6087	333	7	95.889	95.889	NUM
ejpam-6087	333	8	%	%	NOUN
ejpam-6087	333	9	further	far	ADV
ejpam-6087	333	10	emphasizes	emphasize	VERB
ejpam-6087	333	11	its	its	PRON
ejpam-6087	333	12	strong	strong	ADJ
ejpam-6087	333	13	discriminatory	discriminatory	ADJ
ejpam-6087	333	14	power	power	NOUN
ejpam-6087	333	15	,	,	PUNCT
ejpam-6087	333	16	making	make	VERB
ejpam-6087	333	17	it	it	PRON
ejpam-6087	333	18	one	one	NUM
ejpam-6087	333	19	of	of	ADP
ejpam-6087	333	20	the	the	DET
ejpam-6087	333	21	most	most	ADV
ejpam-6087	333	22	reliable	reliable	ADJ
ejpam-6087	333	23	classifiers	classifier	NOUN
ejpam-6087	333	24	in	in	ADP
ejpam-6087	333	25	this	this	DET
ejpam-6087	333	26	evaluation	evaluation	NOUN
ejpam-6087	333	27	.	.	PUNCT
ejpam-6087	334	1	•	•	NUM
ejpam-6087	335	1	k	k	X
ejpam-6087	335	2	-	-	PUNCT
ejpam-6087	335	3	nearest	near	ADJ
ejpam-6087	335	4	neighbors	neighbor	NOUN
ejpam-6087	336	1	:	:	PUNCT
ejpam-6087	336	2	knn	knn	PROPN
ejpam-6087	336	3	showed	show	VERB
ejpam-6087	336	4	a	a	DET
ejpam-6087	336	5	slightly	slightly	ADV
ejpam-6087	336	6	lower	low	ADJ
ejpam-6087	336	7	performance	performance	NOUN
ejpam-6087	336	8	with	with	ADP
ejpam-6087	336	9	an	an	DET
ejpam-6087	336	10	accuracy	accuracy	NOUN
ejpam-6087	336	11	of	of	ADP
ejpam-6087	336	12	92.414	92.414	NUM
ejpam-6087	336	13	%	%	NOUN
ejpam-6087	336	14	.	.	PUNCT
ejpam-6087	337	1	despite	despite	SCONJ
ejpam-6087	337	2	its	its	PRON
ejpam-6087	337	3	consistent	consistent	ADJ
ejpam-6087	337	4	scores	score	NOUN
ejpam-6087	337	5	in	in	ADP
ejpam-6087	337	6	precision	precision	NOUN
ejpam-6087	337	7	,	,	PUNCT
ejpam-6087	337	8	recall	recall	NOUN
ejpam-6087	337	9	,	,	PUNCT
ejpam-6087	337	10	and	and	CCONJ
ejpam-6087	337	11	f1	f1	PROPN
ejpam-6087	337	12	score	score	NOUN
ejpam-6087	337	13	,	,	PUNCT
ejpam-6087	337	14	its	its	PRON
ejpam-6087	337	15	auc	auc	NOUN
ejpam-6087	337	16	-	-	PUNCT
ejpam-6087	337	17	roc	roc	NOUN
ejpam-6087	337	18	of	of	ADP
ejpam-6087	337	19	95.029	95.029	NUM
ejpam-6087	337	20	%	%	NOUN
ejpam-6087	337	21	suggests	suggest	VERB
ejpam-6087	337	22	that	that	SCONJ
ejpam-6087	337	23	knn	knn	PROPN
ejpam-6087	337	24	can	can	AUX
ejpam-6087	337	25	effectively	effectively	ADV
ejpam-6087	337	26	handle	handle	VERB
ejpam-6087	337	27	the	the	DET
ejpam-6087	337	28	classification	classification	NOUN
ejpam-6087	337	29	task	task	NOUN
ejpam-6087	337	30	,	,	PUNCT
ejpam-6087	337	31	although	although	SCONJ
ejpam-6087	337	32	not	not	PART
ejpam-6087	337	33	as	as	ADV
ejpam-6087	337	34	efficiently	efficiently	ADV
ejpam-6087	337	35	as	as	ADP
ejpam-6087	337	36	some	some	DET
ejpam-6087	337	37	other	other	ADJ
ejpam-6087	337	38	methods	method	NOUN
ejpam-6087	337	39	.	.	PUNCT
ejpam-6087	338	1	•	•	NUM
ejpam-6087	338	2	multilayer	multilayer	PROPN
ejpam-6087	338	3	perceptron	perceptron	PROPN
ejpam-6087	338	4	:	:	PUNCT
ejpam-6087	339	1	mlp	mlp	PROPN
ejpam-6087	339	2	achieved	achieve	VERB
ejpam-6087	339	3	an	an	DET
ejpam-6087	339	4	accuracy	accuracy	NOUN
ejpam-6087	339	5	of	of	ADP
ejpam-6087	339	6	95.387	95.387	NUM
ejpam-6087	339	7	%	%	NOUN
ejpam-6087	339	8	,	,	PUNCT
ejpam-6087	339	9	with	with	ADP
ejpam-6087	339	10	matching	match	VERB
ejpam-6087	339	11	precision	precision	NOUN
ejpam-6087	339	12	,	,	PUNCT
ejpam-6087	339	13	recall	recall	NOUN
ejpam-6087	339	14	,	,	PUNCT
ejpam-6087	339	15	and	and	CCONJ
ejpam-6087	339	16	f1	f1	ADJ
ejpam-6087	339	17	score	score	NOUN
ejpam-6087	339	18	values	value	NOUN
ejpam-6087	339	19	.	.	PUNCT
ejpam-6087	340	1	its	its	PRON
ejpam-6087	340	2	auc	auc	NOUN
ejpam-6087	340	3	-	-	PUNCT
ejpam-6087	340	4	roc	roc	NOUN
ejpam-6087	340	5	of	of	ADP
ejpam-6087	340	6	95.847	95.847	NUM
ejpam-6087	340	7	%	%	NOUN
ejpam-6087	340	8	indicates	indicate	VERB
ejpam-6087	340	9	strong	strong	ADJ
ejpam-6087	340	10	predictive	predictive	ADJ
ejpam-6087	340	11	performance	performance	NOUN
ejpam-6087	340	12	,	,	PUNCT
ejpam-6087	340	13	making	make	VERB
ejpam-6087	340	14	it	it	PRON
ejpam-6087	340	15	a	a	DET
ejpam-6087	340	16	competitive	competitive	ADJ
ejpam-6087	340	17	choice	choice	NOUN
ejpam-6087	340	18	for	for	ADP
ejpam-6087	340	19	the	the	DET
ejpam-6087	340	20	detection	detection	NOUN
ejpam-6087	340	21	of	of	ADP
ejpam-6087	340	22	fake	fake	ADJ
ejpam-6087	340	23	news	news	NOUN
ejpam-6087	340	24	.	.	PUNCT
ejpam-6087	341	1	•	•	NUM
ejpam-6087	341	2	support	support	NOUN
ejpam-6087	341	3	vector	vector	NOUN
ejpam-6087	341	4	machine	machine	NOUN
ejpam-6087	341	5	:	:	PUNCT
ejpam-6087	341	6	svm	svm	PROPN
ejpam-6087	341	7	stood	stand	VERB
ejpam-6087	341	8	out	out	ADP
ejpam-6087	341	9	with	with	ADP
ejpam-6087	341	10	an	an	DET
ejpam-6087	341	11	accuracy	accuracy	NOUN
ejpam-6087	341	12	of	of	ADP
ejpam-6087	341	13	95.704	95.704	NUM
ejpam-6087	341	14	%	%	NOUN
ejpam-6087	341	15	and	and	CCONJ
ejpam-6087	341	16	uniformly	uniformly	ADV
ejpam-6087	341	17	high	high	ADJ
ejpam-6087	341	18	scores	score	NOUN
ejpam-6087	341	19	for	for	ADP
ejpam-6087	341	20	precision	precision	NOUN
ejpam-6087	341	21	,	,	PUNCT
ejpam-6087	341	22	recall	recall	NOUN
ejpam-6087	341	23	,	,	PUNCT
ejpam-6087	341	24	and	and	CCONJ
ejpam-6087	341	25	f1	f1	PROPN
ejpam-6087	341	26	score	score	NOUN
ejpam-6087	341	27	.	.	PUNCT
ejpam-6087	342	1	its	its	PRON
ejpam-6087	342	2	auc	auc	NOUN
ejpam-6087	342	3	-	-	PUNCT
ejpam-6087	342	4	roc	roc	NOUN
ejpam-6087	342	5	of	of	ADP
ejpam-6087	342	6	95.902	95.902	NUM
ejpam-6087	342	7	%	%	NOUN
ejpam-6087	342	8	underscores	underscore	VERB
ejpam-6087	342	9	its	its	PRON
ejpam-6087	342	10	excellent	excellent	ADJ
ejpam-6087	342	11	ability	ability	NOUN
ejpam-6087	342	12	to	to	PART
ejpam-6087	342	13	differentiate	differentiate	VERB
ejpam-6087	342	14	between	between	ADP
ejpam-6087	342	15	classes	class	NOUN
ejpam-6087	342	16	,	,	PUNCT
ejpam-6087	342	17	proving	prove	VERB
ejpam-6087	342	18	to	to	PART
ejpam-6087	342	19	be	be	AUX
ejpam-6087	342	20	one	one	NUM
ejpam-6087	342	21	of	of	ADP
ejpam-6087	342	22	the	the	DET
ejpam-6087	342	23	best	good	ADJ
ejpam-6087	342	24	performers	performer	NOUN
ejpam-6087	342	25	.	.	PUNCT
ejpam-6087	343	1	•	•	NUM
ejpam-6087	343	2	decision	decision	NOUN
ejpam-6087	343	3	tree	tree	NOUN
ejpam-6087	343	4	:	:	PUNCT
ejpam-6087	343	5	dt	dt	PROPN
ejpam-6087	343	6	showed	show	VERB
ejpam-6087	343	7	a	a	DET
ejpam-6087	343	8	lower	low	ADJ
ejpam-6087	343	9	overall	overall	ADJ
ejpam-6087	343	10	performance	performance	NOUN
ejpam-6087	343	11	with	with	ADP
ejpam-6087	343	12	an	an	DET
ejpam-6087	343	13	accuracy	accuracy	NOUN
ejpam-6087	343	14	of	of	ADP
ejpam-6087	343	15	91.006	91.006	NUM
ejpam-6087	343	16	%	%	NOUN
ejpam-6087	343	17	.	.	PUNCT
ejpam-6087	344	1	its	its	PRON
ejpam-6087	344	2	precision	precision	NOUN
ejpam-6087	344	3	and	and	CCONJ
ejpam-6087	344	4	recall	recall	NOUN
ejpam-6087	344	5	values	value	NOUN
ejpam-6087	344	6	were	be	AUX
ejpam-6087	344	7	slightly	slightly	ADV
ejpam-6087	344	8	higher	high	ADJ
ejpam-6087	344	9	at	at	ADP
ejpam-6087	344	10	91.007	91.007	NUM
ejpam-6087	344	11	%	%	NOUN
ejpam-6087	344	12	,	,	PUNCT
ejpam-6087	344	13	but	but	CCONJ
ejpam-6087	344	14	the	the	DET
ejpam-6087	344	15	f1	f1	PROPN
ejpam-6087	344	16	score	score	NOUN
ejpam-6087	344	17	and	and	CCONJ
ejpam-6087	344	18	auc	auc	NOUN
ejpam-6087	344	19	-	-	PUNCT
ejpam-6087	344	20	roc	roc	NOUN
ejpam-6087	344	21	of	of	ADP
ejpam-6087	344	22	91.002	91.002	NUM
ejpam-6087	344	23	%	%	NOUN
ejpam-6087	344	24	reflect	reflect	VERB
ejpam-6087	344	25	its	its	PRON
ejpam-6087	344	26	limitations	limitation	NOUN
ejpam-6087	344	27	in	in	ADP
ejpam-6087	344	28	handling	handle	VERB
ejpam-6087	344	29	the	the	DET
ejpam-6087	344	30	data	datum	NOUN
ejpam-6087	344	31	set	set	VERB
ejpam-6087	344	32	as	as	ADV
ejpam-6087	344	33	effectively	effectively	ADV
ejpam-6087	344	34	as	as	ADP
ejpam-6087	344	35	other	other	ADJ
ejpam-6087	344	36	classifiers	classifier	NOUN
ejpam-6087	344	37	.	.	PUNCT
ejpam-6087	345	1	•	•	NUM
ejpam-6087	345	2	random	random	ADJ
ejpam-6087	345	3	forest	forest	NOUN
ejpam-6087	345	4	:	:	PUNCT
ejpam-6087	345	5	rf	rf	VERB
ejpam-6087	345	6	matched	match	VERB
ejpam-6087	345	7	lr	lr	NOUN
ejpam-6087	345	8	in	in	ADP
ejpam-6087	345	9	accuracy	accuracy	NOUN
ejpam-6087	345	10	,	,	PUNCT
ejpam-6087	345	11	precision	precision	NOUN
ejpam-6087	345	12	,	,	PUNCT
ejpam-6087	345	13	recall	recall	NOUN
ejpam-6087	345	14	,	,	PUNCT
ejpam-6087	345	15	and	and	CCONJ
ejpam-6087	345	16	f1	f1	NOUN
ejpam-6087	345	17	score	score	NOUN
ejpam-6087	345	18	at	at	ADP
ejpam-6087	345	19	95	95	NUM
ejpam-6087	345	20	.	.	PUNCT
ejpam-6087	346	1	801	801	NUM
ejpam-6087	346	2	%	%	NOUN
ejpam-6087	346	3	.	.	PUNCT
ejpam-6087	347	1	its	its	PRON
ejpam-6087	347	2	auc	auc	NOUN
ejpam-6087	347	3	-	-	PUNCT
ejpam-6087	347	4	roc	roc	NOUN
ejpam-6087	347	5	of	of	ADP
ejpam-6087	347	6	95.892	95.892	NUM
ejpam-6087	347	7	%	%	NOUN
ejpam-6087	347	8	confirms	confirm	VERB
ejpam-6087	347	9	its	its	PRON
ejpam-6087	347	10	reliability	reliability	NOUN
ejpam-6087	347	11	and	and	CCONJ
ejpam-6087	347	12	robustness	robustness	NOUN
ejpam-6087	347	13	,	,	PUNCT
ejpam-6087	347	14	showcasing	showcase	VERB
ejpam-6087	347	15	its	its	PRON
ejpam-6087	347	16	potential	potential	NOUN
ejpam-6087	347	17	as	as	ADP
ejpam-6087	347	18	a	a	DET
ejpam-6087	347	19	powerful	powerful	ADJ
ejpam-6087	347	20	ensemble	ensemble	ADJ
ejpam-6087	347	21	method	method	NOUN
ejpam-6087	347	22	for	for	ADP
ejpam-6087	347	23	detecting	detect	VERB
ejpam-6087	347	24	fake	fake	ADJ
ejpam-6087	347	25	news	news	NOUN
ejpam-6087	347	26	.	.	PUNCT
ejpam-6087	348	1	m.	m.	PROPN
ejpam-6087	348	2	al	al	PROPN
ejpam-6087	348	3	-	-	PUNCT
ejpam-6087	348	4	tarawneh	tarawneh	PROPN
ejpam-6087	348	5	et	et	PROPN
ejpam-6087	348	6	al	al	PROPN
ejpam-6087	348	7	.	.	PUNCT
ejpam-6087	348	8	/	/	SYM
ejpam-6087	348	9	eur	eur	PROPN
ejpam-6087	348	10	.	.	PUNCT
ejpam-6087	349	1	j.	j.	PROPN
ejpam-6087	349	2	pure	pure	PROPN
ejpam-6087	349	3	appl	appl	PROPN
ejpam-6087	349	4	.	.	PROPN
ejpam-6087	349	5	math	math	PROPN
ejpam-6087	349	6	,	,	PUNCT
ejpam-6087	349	7	18	18	NUM
ejpam-6087	349	8	(	(	PUNCT
ejpam-6087	349	9	2	2	NUM
ejpam-6087	349	10	)	)	PUNCT
ejpam-6087	349	11	(	(	PUNCT
ejpam-6087	349	12	2025	2025	NUM
ejpam-6087	349	13	)	)	PUNCT
ejpam-6087	349	14	,	,	PUNCT
ejpam-6087	349	15	6087	6087	NUM
ejpam-6087	349	16	19	19	NUM
ejpam-6087	349	17	of	of	ADP
ejpam-6087	349	18	39	39	NUM
ejpam-6087	349	19	in	in	ADP
ejpam-6087	349	20	summary	summary	NOUN
ejpam-6087	349	21	,	,	PUNCT
ejpam-6087	349	22	lr	lr	INTJ
ejpam-6087	349	23	,	,	PUNCT
ejpam-6087	349	24	svm	svm	NOUN
ejpam-6087	349	25	and	and	CCONJ
ejpam-6087	349	26	rf	rf	VERB
ejpam-6087	349	27	emerged	emerge	VERB
ejpam-6087	349	28	as	as	ADP
ejpam-6087	349	29	the	the	DET
ejpam-6087	349	30	leading	lead	VERB
ejpam-6087	349	31	classifiers	classifier	NOUN
ejpam-6087	349	32	,	,	PUNCT
ejpam-6087	349	33	demonstrating	demonstrate	VERB
ejpam-6087	349	34	high	high	ADJ
ejpam-6087	349	35	precision	precision	NOUN
ejpam-6087	349	36	and	and	CCONJ
ejpam-6087	349	37	balanced	balanced	ADJ
ejpam-6087	349	38	performance	performance	NOUN
ejpam-6087	349	39	across	across	ADP
ejpam-6087	349	40	all	all	DET
ejpam-6087	349	41	metrics	metric	NOUN
ejpam-6087	349	42	.	.	PUNCT
ejpam-6087	350	1	these	these	DET
ejpam-6087	350	2	methods	method	NOUN
ejpam-6087	350	3	provide	provide	VERB
ejpam-6087	350	4	a	a	DET
ejpam-6087	350	5	solid	solid	ADJ
ejpam-6087	350	6	foundation	foundation	NOUN
ejpam-6087	350	7	for	for	ADP
ejpam-6087	350	8	reliable	reliable	ADJ
ejpam-6087	350	9	fake	fake	ADJ
ejpam-6087	350	10	news	news	NOUN
ejpam-6087	350	11	detection	detection	NOUN
ejpam-6087	350	12	on	on	ADP
ejpam-6087	350	13	the	the	DET
ejpam-6087	350	14	truthseeker	truthseeker	NOUN
ejpam-6087	350	15	dataset	dataset	NOUN
ejpam-6087	350	16	,	,	PUNCT
ejpam-6087	350	17	highlighting	highlight	VERB
ejpam-6087	350	18	their	their	PRON
ejpam-6087	350	19	practical	practical	ADJ
ejpam-6087	350	20	applicability	applicability	NOUN
ejpam-6087	350	21	in	in	ADP
ejpam-6087	350	22	real	real	ADJ
ejpam-6087	350	23	-	-	PUNCT
ejpam-6087	350	24	world	world	NOUN
ejpam-6087	350	25	scenarios	scenario	NOUN
ejpam-6087	350	26	.	.	PUNCT
ejpam-6087	351	1	4.2	4.2	NUM
ejpam-6087	351	2	.	.	PUNCT
ejpam-6087	351	3	ensemble	ensemble	ADJ
ejpam-6087	351	4	methods	method	NOUN
ejpam-6087	351	5	results	result	VERB
ejpam-6087	351	6	table	table	VERB
ejpam-6087	351	7	5	5	NUM
ejpam-6087	351	8	presents	present	VERB
ejpam-6087	351	9	the	the	DET
ejpam-6087	351	10	performance	performance	NOUN
ejpam-6087	351	11	metrics	metric	NOUN
ejpam-6087	351	12	of	of	ADP
ejpam-6087	351	13	various	various	ADJ
ejpam-6087	351	14	ensemble	ensemble	ADJ
ejpam-6087	351	15	ml	ml	NOUN
ejpam-6087	351	16	algorithms	algorithm	NOUN
ejpam-6087	351	17	used	use	VERB
ejpam-6087	351	18	for	for	ADP
ejpam-6087	351	19	the	the	DET
ejpam-6087	351	20	detection	detection	NOUN
ejpam-6087	351	21	of	of	ADP
ejpam-6087	351	22	fake	fake	ADJ
ejpam-6087	351	23	news	news	NOUN
ejpam-6087	351	24	.	.	PUNCT
ejpam-6087	352	1	these	these	DET
ejpam-6087	352	2	algorithms	algorithm	NOUN
ejpam-6087	352	3	include	include	VERB
ejpam-6087	352	4	both	both	CCONJ
ejpam-6087	352	5	homogeneous	homogeneous	ADJ
ejpam-6087	352	6	and	and	CCONJ
ejpam-6087	352	7	heterogeneous	heterogeneous	ADJ
ejpam-6087	352	8	ensemble	ensemble	ADJ
ejpam-6087	352	9	methods	method	NOUN
ejpam-6087	352	10	,	,	PUNCT
ejpam-6087	352	11	evaluated	evaluate	VERB
ejpam-6087	352	12	across	across	ADP
ejpam-6087	352	13	five	five	NUM
ejpam-6087	352	14	key	key	ADJ
ejpam-6087	352	15	metrics	metric	NOUN
ejpam-6087	352	16	:	:	PUNCT
ejpam-6087	352	17	accuracy	accuracy	NOUN
ejpam-6087	352	18	,	,	PUNCT
ejpam-6087	352	19	precision	precision	NOUN
ejpam-6087	352	20	,	,	PUNCT
ejpam-6087	352	21	recall	recall	NOUN
ejpam-6087	352	22	,	,	PUNCT
ejpam-6087	352	23	f1	f1	NOUN
ejpam-6087	352	24	score	score	NOUN
ejpam-6087	352	25	,	,	PUNCT
ejpam-6087	352	26	and	and	CCONJ
ejpam-6087	352	27	auc	auc	NOUN
ejpam-6087	352	28	-	-	PUNCT
ejpam-6087	352	29	roc	roc	NOUN
ejpam-6087	352	30	.	.	PUNCT
ejpam-6087	353	1	the	the	DET
ejpam-6087	353	2	results	result	NOUN
ejpam-6087	353	3	depicted	depict	VERB
ejpam-6087	353	4	in	in	ADP
ejpam-6087	353	5	this	this	DET
ejpam-6087	353	6	table	table	NOUN
ejpam-6087	353	7	can	can	AUX
ejpam-6087	353	8	be	be	AUX
ejpam-6087	353	9	summarized	summarize	VERB
ejpam-6087	353	10	as	as	SCONJ
ejpam-6087	353	11	follows	follow	VERB
ejpam-6087	353	12	:	:	PUNCT
ejpam-6087	353	13	•	•	NUM
ejpam-6087	353	14	catboost	catboost	PROPN
ejpam-6087	353	15	demonstrated	demonstrate	VERB
ejpam-6087	353	16	the	the	DET
ejpam-6087	353	17	highest	high	ADJ
ejpam-6087	353	18	overall	overall	ADJ
ejpam-6087	353	19	performance	performance	NOUN
ejpam-6087	353	20	with	with	ADP
ejpam-6087	353	21	an	an	DET
ejpam-6087	353	22	accuracy	accuracy	NOUN
ejpam-6087	353	23	of	of	ADP
ejpam-6087	353	24	95.801	95.801	NUM
ejpam-6087	353	25	%	%	NOUN
ejpam-6087	353	26	and	and	CCONJ
ejpam-6087	353	27	an	an	DET
ejpam-6087	353	28	equally	equally	ADV
ejpam-6087	353	29	high	high	ADJ
ejpam-6087	353	30	precision	precision	NOUN
ejpam-6087	353	31	,	,	PUNCT
ejpam-6087	353	32	recall	recall	NOUN
ejpam-6087	353	33	,	,	PUNCT
ejpam-6087	353	34	and	and	CCONJ
ejpam-6087	353	35	f1	f1	ADJ
ejpam-6087	353	36	score	score	NOUN
ejpam-6087	353	37	of	of	ADP
ejpam-6087	353	38	95.801	95.801	NUM
ejpam-6087	353	39	%	%	NOUN
ejpam-6087	353	40	.	.	PUNCT
ejpam-6087	354	1	its	its	PRON
ejpam-6087	354	2	auc	auc	NOUN
ejpam-6087	354	3	-	-	PUNCT
ejpam-6087	354	4	roc	roc	NOUN
ejpam-6087	354	5	score	score	NOUN
ejpam-6087	354	6	of	of	ADP
ejpam-6087	354	7	95.859	95.859	NUM
ejpam-6087	354	8	%	%	NOUN
ejpam-6087	354	9	highlights	highlight	NOUN
ejpam-6087	354	10	it	it	PRON
ejpam-6087	354	11	as	as	ADP
ejpam-6087	354	12	one	one	NUM
ejpam-6087	354	13	of	of	ADP
ejpam-6087	354	14	the	the	DET
ejpam-6087	354	15	main	main	ADJ
ejpam-6087	354	16	methods	method	NOUN
ejpam-6087	354	17	to	to	PART
ejpam-6087	354	18	distinguish	distinguish	VERB
ejpam-6087	354	19	between	between	ADP
ejpam-6087	354	20	fake	fake	ADJ
ejpam-6087	354	21	and	and	CCONJ
ejpam-6087	354	22	real	real	ADJ
ejpam-6087	354	23	news	news	NOUN
ejpam-6087	354	24	.	.	PUNCT
ejpam-6087	355	1	•	•	NUM
ejpam-6087	355	2	histgradient	histgradient	NOUN
ejpam-6087	355	3	boosting	boost	VERB
ejpam-6087	355	4	also	also	ADV
ejpam-6087	355	5	achieved	achieve	VERB
ejpam-6087	355	6	a	a	DET
ejpam-6087	355	7	high	high	ADJ
ejpam-6087	355	8	accuracy	accuracy	NOUN
ejpam-6087	355	9	of	of	ADP
ejpam-6087	355	10	95.801	95.801	NUM
ejpam-6087	355	11	%	%	NOUN
ejpam-6087	355	12	with	with	ADP
ejpam-6087	355	13	consistent	consistent	ADJ
ejpam-6087	355	14	metrics	metric	NOUN
ejpam-6087	355	15	for	for	ADP
ejpam-6087	355	16	precision	precision	NOUN
ejpam-6087	355	17	,	,	PUNCT
ejpam-6087	355	18	recall	recall	NOUN
ejpam-6087	355	19	,	,	PUNCT
ejpam-6087	355	20	and	and	CCONJ
ejpam-6087	355	21	f1	f1	PROPN
ejpam-6087	355	22	score	score	NOUN
ejpam-6087	355	23	.	.	PUNCT
ejpam-6087	356	1	its	its	PRON
ejpam-6087	356	2	auc	auc	NOUN
ejpam-6087	356	3	-	-	PUNCT
ejpam-6087	356	4	roc	roc	NOUN
ejpam-6087	356	5	of	of	ADP
ejpam-6087	356	6	95.904	95.904	NUM
ejpam-6087	356	7	%	%	NOUN
ejpam-6087	356	8	is	be	AUX
ejpam-6087	356	9	the	the	DET
ejpam-6087	356	10	highest	high	ADJ
ejpam-6087	356	11	among	among	ADP
ejpam-6087	356	12	all	all	DET
ejpam-6087	356	13	methods	method	NOUN
ejpam-6087	356	14	tested	test	VERB
ejpam-6087	356	15	,	,	PUNCT
ejpam-6087	356	16	marking	mark	VERB
ejpam-6087	356	17	it	it	PRON
ejpam-6087	356	18	as	as	ADP
ejpam-6087	356	19	the	the	DET
ejpam-6087	356	20	most	most	ADV
ejpam-6087	356	21	effective	effective	ADJ
ejpam-6087	356	22	classifier	classifier	NOUN
ejpam-6087	356	23	for	for	ADP
ejpam-6087	356	24	this	this	DET
ejpam-6087	356	25	dataset	dataset	NOUN
ejpam-6087	356	26	.	.	PUNCT
ejpam-6087	357	1	•	•	NOUN
ejpam-6087	357	2	lightgbm	lightgbm	VERB
ejpam-6087	357	3	matched	match	VERB
ejpam-6087	357	4	catboost	catboost	NOUN
ejpam-6087	357	5	and	and	CCONJ
ejpam-6087	357	6	histgradient	histgradient	NOUN
ejpam-6087	357	7	boosting	boost	VERB
ejpam-6087	357	8	with	with	ADP
ejpam-6087	357	9	an	an	DET
ejpam-6087	357	10	accuracy	accuracy	NOUN
ejpam-6087	357	11	of	of	ADP
ejpam-6087	357	12	95.801	95.801	NUM
ejpam-6087	357	13	%	%	NOUN
ejpam-6087	357	14	and	and	CCONJ
ejpam-6087	357	15	consistent	consistent	ADJ
ejpam-6087	357	16	precision	precision	NOUN
ejpam-6087	357	17	,	,	PUNCT
ejpam-6087	357	18	recall	recall	NOUN
ejpam-6087	357	19	and	and	CCONJ
ejpam-6087	357	20	f1	f1	PROPN
ejpam-6087	357	21	score	score	NOUN
ejpam-6087	357	22	.	.	PUNCT
ejpam-6087	358	1	the	the	DET
ejpam-6087	358	2	auc	auc	NOUN
ejpam-6087	358	3	-	-	PUNCT
ejpam-6087	358	4	roc	roc	NOUN
ejpam-6087	358	5	score	score	NOUN
ejpam-6087	358	6	of	of	ADP
ejpam-6087	358	7	95.854	95.854	NUM
ejpam-6087	358	8	%	%	NOUN
ejpam-6087	358	9	further	far	ADV
ejpam-6087	358	10	establishes	establish	VERB
ejpam-6087	358	11	it	it	PRON
ejpam-6087	358	12	as	as	ADP
ejpam-6087	358	13	a	a	DET
ejpam-6087	358	14	highly	highly	ADV
ejpam-6087	358	15	effective	effective	ADJ
ejpam-6087	358	16	method	method	NOUN
ejpam-6087	358	17	for	for	ADP
ejpam-6087	358	18	the	the	DET
ejpam-6087	358	19	detection	detection	NOUN
ejpam-6087	358	20	of	of	ADP
ejpam-6087	358	21	fake	fake	ADJ
ejpam-6087	358	22	news	news	NOUN
ejpam-6087	358	23	.	.	PUNCT
ejpam-6087	359	1	•	•	NUM
ejpam-6087	359	2	xgboost	xgboost	PRON
ejpam-6087	359	3	achieved	achieve	VERB
ejpam-6087	359	4	an	an	DET
ejpam-6087	359	5	accuracy	accuracy	NOUN
ejpam-6087	359	6	of	of	ADP
ejpam-6087	359	7	95.790	95.790	NUM
ejpam-6087	359	8	%	%	NOUN
ejpam-6087	359	9	with	with	ADP
ejpam-6087	359	10	an	an	DET
ejpam-6087	359	11	equivalent	equivalent	ADJ
ejpam-6087	359	12	precision	precision	NOUN
ejpam-6087	359	13	,	,	PUNCT
ejpam-6087	359	14	recall	recall	NOUN
ejpam-6087	359	15	,	,	PUNCT
ejpam-6087	359	16	and	and	CCONJ
ejpam-6087	359	17	f1	f1	PROPN
ejpam-6087	359	18	score	score	NOUN
ejpam-6087	359	19	.	.	PUNCT
ejpam-6087	360	1	its	its	PRON
ejpam-6087	360	2	auc	auc	NOUN
ejpam-6087	360	3	-	-	PUNCT
ejpam-6087	360	4	roc	roc	NOUN
ejpam-6087	360	5	of	of	ADP
ejpam-6087	360	6	95.880	95.880	NUM
ejpam-6087	360	7	%	%	NOUN
ejpam-6087	360	8	reflects	reflect	VERB
ejpam-6087	360	9	strong	strong	ADJ
ejpam-6087	360	10	performance	performance	NOUN
ejpam-6087	360	11	and	and	CCONJ
ejpam-6087	360	12	makes	make	VERB
ejpam-6087	360	13	it	it	PRON
ejpam-6087	360	14	a	a	DET
ejpam-6087	360	15	competitive	competitive	ADJ
ejpam-6087	360	16	choice	choice	NOUN
ejpam-6087	360	17	among	among	ADP
ejpam-6087	360	18	the	the	DET
ejpam-6087	360	19	tested	test	VERB
ejpam-6087	360	20	classifiers	classifier	NOUN
ejpam-6087	360	21	.	.	PUNCT
ejpam-6087	361	1	•	•	NOUN
ejpam-6087	361	2	stacking	stack	VERB
ejpam-6087	361	3	classifier	classifier	NOUN
ejpam-6087	361	4	produced	produce	VERB
ejpam-6087	361	5	an	an	DET
ejpam-6087	361	6	accuracy	accuracy	NOUN
ejpam-6087	361	7	of	of	ADP
ejpam-6087	361	8	95.801	95.801	NUM
ejpam-6087	361	9	%	%	NOUN
ejpam-6087	361	10	and	and	CCONJ
ejpam-6087	361	11	matched	match	VERB
ejpam-6087	361	12	the	the	DET
ejpam-6087	361	13	metrics	metric	NOUN
ejpam-6087	361	14	for	for	ADP
ejpam-6087	361	15	precision	precision	NOUN
ejpam-6087	361	16	,	,	PUNCT
ejpam-6087	361	17	recall	recall	NOUN
ejpam-6087	361	18	,	,	PUNCT
ejpam-6087	361	19	and	and	CCONJ
ejpam-6087	361	20	f1	f1	NOUN
ejpam-6087	361	21	-	-	PUNCT
ejpam-6087	361	22	score	score	NOUN
ejpam-6087	361	23	.	.	PUNCT
ejpam-6087	362	1	with	with	ADP
ejpam-6087	362	2	an	an	DET
ejpam-6087	362	3	auc	auc	NOUN
ejpam-6087	362	4	-	-	PUNCT
ejpam-6087	362	5	roc	roc	NOUN
ejpam-6087	362	6	of	of	ADP
ejpam-6087	362	7	95.816	95.816	NUM
ejpam-6087	362	8	%	%	NOUN
ejpam-6087	362	9	,	,	PUNCT
ejpam-6087	362	10	it	it	PRON
ejpam-6087	362	11	is	be	AUX
ejpam-6087	362	12	also	also	ADV
ejpam-6087	362	13	a	a	DET
ejpam-6087	362	14	strong	strong	ADJ
ejpam-6087	362	15	performer	performer	NOUN
ejpam-6087	362	16	for	for	ADP
ejpam-6087	362	17	the	the	DET
ejpam-6087	362	18	detection	detection	NOUN
ejpam-6087	362	19	of	of	ADP
ejpam-6087	362	20	fake	fake	ADJ
ejpam-6087	362	21	news	news	NOUN
ejpam-6087	362	22	.	.	PUNCT
ejpam-6087	363	1	•	•	NOUN
ejpam-6087	363	2	voting	voting	NOUN
ejpam-6087	363	3	classifier	classifier	NOUN
ejpam-6087	363	4	showed	show	VERB
ejpam-6087	363	5	an	an	DET
ejpam-6087	363	6	accuracy	accuracy	NOUN
ejpam-6087	363	7	of	of	ADP
ejpam-6087	363	8	95.786	95.786	NUM
ejpam-6087	363	9	%	%	NOUN
ejpam-6087	363	10	with	with	ADP
ejpam-6087	363	11	consistent	consistent	ADJ
ejpam-6087	363	12	precision	precision	NOUN
ejpam-6087	363	13	,	,	PUNCT
ejpam-6087	363	14	recall	recall	NOUN
ejpam-6087	363	15	,	,	PUNCT
ejpam-6087	363	16	and	and	CCONJ
ejpam-6087	363	17	f1	f1	ADJ
ejpam-6087	363	18	-	-	PUNCT
ejpam-6087	363	19	score	score	NOUN
ejpam-6087	363	20	values	value	NOUN
ejpam-6087	363	21	.	.	PUNCT
ejpam-6087	364	1	the	the	DET
ejpam-6087	364	2	auc	auc	NOUN
ejpam-6087	364	3	-	-	PUNCT
ejpam-6087	364	4	roc	roc	NOUN
ejpam-6087	364	5	of	of	ADP
ejpam-6087	364	6	95.892	95.892	NUM
ejpam-6087	364	7	%	%	NOUN
ejpam-6087	364	8	indicates	indicate	VERB
ejpam-6087	364	9	reliable	reliable	ADJ
ejpam-6087	364	10	performance	performance	NOUN
ejpam-6087	364	11	,	,	PUNCT
ejpam-6087	364	12	although	although	SCONJ
ejpam-6087	364	13	slightly	slightly	ADV
ejpam-6087	364	14	less	less	ADV
ejpam-6087	364	15	effective	effective	ADJ
ejpam-6087	364	16	compared	compare	VERB
ejpam-6087	364	17	to	to	ADP
ejpam-6087	364	18	the	the	DET
ejpam-6087	364	19	best	good	ADJ
ejpam-6087	364	20	methods	method	NOUN
ejpam-6087	364	21	.	.	PUNCT
ejpam-6087	365	1	•	•	NOUN
ejpam-6087	365	2	bagging	bagging	NOUN
ejpam-6087	365	3	(	(	PUNCT
ejpam-6087	365	4	decision	decision	NOUN
ejpam-6087	365	5	tree	tree	NOUN
ejpam-6087	365	6	)	)	PUNCT
ejpam-6087	365	7	demonstrated	demonstrate	VERB
ejpam-6087	365	8	high	high	ADJ
ejpam-6087	365	9	performance	performance	NOUN
ejpam-6087	365	10	with	with	ADP
ejpam-6087	365	11	an	an	DET
ejpam-6087	365	12	accuracy	accuracy	NOUN
ejpam-6087	365	13	of	of	ADP
ejpam-6087	365	14	95.738	95.738	NUM
ejpam-6087	365	15	%	%	NOUN
ejpam-6087	365	16	and	and	CCONJ
ejpam-6087	365	17	uniform	uniform	ADJ
ejpam-6087	365	18	scores	score	NOUN
ejpam-6087	365	19	for	for	ADP
ejpam-6087	365	20	precision	precision	NOUN
ejpam-6087	365	21	,	,	PUNCT
ejpam-6087	365	22	recall	recall	NOUN
ejpam-6087	365	23	,	,	PUNCT
ejpam-6087	365	24	and	and	CCONJ
ejpam-6087	365	25	f1	f1	PROPN
ejpam-6087	365	26	score	score	NOUN
ejpam-6087	365	27	.	.	PUNCT
ejpam-6087	366	1	its	its	PRON
ejpam-6087	366	2	auc	auc	NOUN
ejpam-6087	366	3	-	-	PUNCT
ejpam-6087	366	4	roc	roc	NOUN
ejpam-6087	366	5	of	of	ADP
ejpam-6087	366	6	95.893	95.893	NUM
ejpam-6087	366	7	%	%	NOUN
ejpam-6087	366	8	confirms	confirm	VERB
ejpam-6087	366	9	its	its	PRON
ejpam-6087	366	10	effectiveness	effectiveness	NOUN
ejpam-6087	366	11	in	in	ADP
ejpam-6087	366	12	handling	handle	VERB
ejpam-6087	366	13	the	the	DET
ejpam-6087	366	14	classification	classification	NOUN
ejpam-6087	366	15	task	task	NOUN
ejpam-6087	366	16	.	.	PUNCT
ejpam-6087	367	1	•	•	NUM
ejpam-6087	367	2	gradient	gradient	NOUN
ejpam-6087	367	3	boosting	boosting	NOUN
ejpam-6087	367	4	provided	provide	VERB
ejpam-6087	367	5	strong	strong	ADJ
ejpam-6087	367	6	results	result	NOUN
ejpam-6087	367	7	with	with	ADP
ejpam-6087	367	8	an	an	DET
ejpam-6087	367	9	accuracy	accuracy	NOUN
ejpam-6087	367	10	of	of	ADP
ejpam-6087	367	11	95.801	95.801	NUM
ejpam-6087	367	12	%	%	NOUN
ejpam-6087	367	13	and	and	CCONJ
ejpam-6087	367	14	consistent	consistent	ADJ
ejpam-6087	367	15	metrics	metric	NOUN
ejpam-6087	367	16	for	for	ADP
ejpam-6087	367	17	precision	precision	NOUN
ejpam-6087	367	18	,	,	PUNCT
ejpam-6087	367	19	recall	recall	NOUN
ejpam-6087	367	20	,	,	PUNCT
ejpam-6087	367	21	and	and	CCONJ
ejpam-6087	367	22	f1	f1	PROPN
ejpam-6087	367	23	score	score	NOUN
ejpam-6087	367	24	.	.	PUNCT
ejpam-6087	368	1	its	its	PRON
ejpam-6087	368	2	auc	auc	NOUN
ejpam-6087	368	3	-	-	PUNCT
ejpam-6087	368	4	roc	roc	NOUN
ejpam-6087	368	5	of	of	ADP
ejpam-6087	368	6	95.827	95.827	NUM
ejpam-6087	368	7	%	%	NOUN
ejpam-6087	368	8	reflects	reflect	VERB
ejpam-6087	368	9	effective	effective	ADJ
ejpam-6087	368	10	performance	performance	NOUN
ejpam-6087	368	11	for	for	ADP
ejpam-6087	368	12	fake	fake	ADJ
ejpam-6087	368	13	news	news	NOUN
ejpam-6087	368	14	detection	detection	NOUN
ejpam-6087	368	15	.	.	PUNCT
ejpam-6087	369	1	m.	m.	NOUN
ejpam-6087	369	2	al	al	PROPN
ejpam-6087	369	3	-	-	PUNCT
ejpam-6087	369	4	tarawneh	tarawneh	PROPN
ejpam-6087	369	5	et	et	PROPN
ejpam-6087	369	6	al	al	PROPN
ejpam-6087	369	7	.	.	PUNCT
ejpam-6087	369	8	/	/	SYM
ejpam-6087	369	9	eur	eur	PROPN
ejpam-6087	369	10	.	.	PUNCT
ejpam-6087	370	1	j.	j.	PROPN
ejpam-6087	370	2	pure	pure	PROPN
ejpam-6087	370	3	appl	appl	PROPN
ejpam-6087	370	4	.	.	PROPN
ejpam-6087	370	5	math	math	PROPN
ejpam-6087	370	6	,	,	PUNCT
ejpam-6087	370	7	18	18	NUM
ejpam-6087	370	8	(	(	PUNCT
ejpam-6087	370	9	2	2	NUM
ejpam-6087	370	10	)	)	PUNCT
ejpam-6087	370	11	(	(	PUNCT
ejpam-6087	370	12	2025	2025	NUM
ejpam-6087	370	13	)	)	PUNCT
ejpam-6087	370	14	,	,	PUNCT
ejpam-6087	370	15	6087	6087	NUM
ejpam-6087	370	16	20	20	NUM
ejpam-6087	370	17	of	of	ADP
ejpam-6087	370	18	39	39	NUM
ejpam-6087	370	19	•	•	NOUN
ejpam-6087	370	20	extra	extra	ADJ
ejpam-6087	370	21	trees	tree	NOUN
ejpam-6087	370	22	achieved	achieve	VERB
ejpam-6087	370	23	an	an	DET
ejpam-6087	370	24	accuracy	accuracy	NOUN
ejpam-6087	370	25	of	of	ADP
ejpam-6087	370	26	95.756	95.756	NUM
ejpam-6087	370	27	%	%	NOUN
ejpam-6087	370	28	with	with	ADP
ejpam-6087	370	29	very	very	ADV
ejpam-6087	370	30	close	close	ADJ
ejpam-6087	370	31	values	value	NOUN
ejpam-6087	370	32	for	for	ADP
ejpam-6087	370	33	precision	precision	NOUN
ejpam-6087	370	34	,	,	PUNCT
ejpam-6087	370	35	recall	recall	NOUN
ejpam-6087	370	36	,	,	PUNCT
ejpam-6087	370	37	and	and	CCONJ
ejpam-6087	370	38	f1	f1	PROPN
ejpam-6087	370	39	score	score	NOUN
ejpam-6087	370	40	.	.	PUNCT
ejpam-6087	371	1	the	the	DET
ejpam-6087	371	2	auc	auc	NOUN
ejpam-6087	371	3	-	-	PUNCT
ejpam-6087	371	4	roc	roc	NOUN
ejpam-6087	371	5	of	of	ADP
ejpam-6087	371	6	95.779	95.779	NUM
ejpam-6087	371	7	%	%	NOUN
ejpam-6087	371	8	indicates	indicate	VERB
ejpam-6087	371	9	strong	strong	ADJ
ejpam-6087	371	10	performance	performance	NOUN
ejpam-6087	371	11	,	,	PUNCT
ejpam-6087	371	12	although	although	SCONJ
ejpam-6087	371	13	slightly	slightly	ADV
ejpam-6087	371	14	lower	low	ADJ
ejpam-6087	371	15	than	than	ADP
ejpam-6087	371	16	the	the	DET
ejpam-6087	371	17	top	top	ADJ
ejpam-6087	371	18	methods	method	NOUN
ejpam-6087	371	19	.	.	PUNCT
ejpam-6087	372	1	•	•	NOUN
ejpam-6087	372	2	adaboost	adaboost	ADV
ejpam-6087	372	3	(	(	PUNCT
ejpam-6087	372	4	decision	decision	NOUN
ejpam-6087	372	5	tree	tree	NOUN
ejpam-6087	372	6	)	)	PUNCT
ejpam-6087	372	7	showed	show	VERB
ejpam-6087	372	8	a	a	DET
ejpam-6087	372	9	lower	low	ADJ
ejpam-6087	372	10	performance	performance	NOUN
ejpam-6087	372	11	with	with	ADP
ejpam-6087	372	12	an	an	DET
ejpam-6087	372	13	accuracy	accuracy	NOUN
ejpam-6087	372	14	of	of	ADP
ejpam-6087	372	15	91.025	91.025	NUM
ejpam-6087	372	16	%	%	NOUN
ejpam-6087	372	17	and	and	CCONJ
ejpam-6087	372	18	consistent	consistent	ADJ
ejpam-6087	372	19	scores	score	NOUN
ejpam-6087	372	20	on	on	ADP
ejpam-6087	372	21	all	all	DET
ejpam-6087	372	22	metrics	metric	NOUN
ejpam-6087	372	23	.	.	PUNCT
ejpam-6087	373	1	the	the	DET
ejpam-6087	373	2	auc	auc	NOUN
ejpam-6087	373	3	-	-	PUNCT
ejpam-6087	373	4	roc	roc	NOUN
ejpam-6087	373	5	of	of	ADP
ejpam-6087	373	6	91.019	91.019	NUM
ejpam-6087	373	7	%	%	NOUN
ejpam-6087	373	8	indicates	indicate	VERB
ejpam-6087	373	9	that	that	SCONJ
ejpam-6087	373	10	it	it	PRON
ejpam-6087	373	11	is	be	AUX
ejpam-6087	373	12	less	less	ADV
ejpam-6087	373	13	effective	effective	ADJ
ejpam-6087	373	14	compared	compare	VERB
ejpam-6087	373	15	to	to	ADP
ejpam-6087	373	16	other	other	ADJ
ejpam-6087	373	17	methods	method	NOUN
ejpam-6087	373	18	.	.	PUNCT
ejpam-6087	374	1	•	•	NOUN
ejpam-6087	374	2	isolation	isolation	NOUN
ejpam-6087	374	3	forest	forest	NOUN
ejpam-6087	374	4	was	be	AUX
ejpam-6087	374	5	the	the	DET
ejpam-6087	374	6	least	least	ADJ
ejpam-6087	374	7	effective	effective	ADJ
ejpam-6087	374	8	with	with	ADP
ejpam-6087	374	9	an	an	DET
ejpam-6087	374	10	accuracy	accuracy	NOUN
ejpam-6087	374	11	of	of	ADP
ejpam-6087	374	12	51.140	51.140	NUM
ejpam-6087	374	13	%	%	NOUN
ejpam-6087	374	14	and	and	CCONJ
ejpam-6087	374	15	significantly	significantly	ADV
ejpam-6087	374	16	lower	low	ADJ
ejpam-6087	374	17	scores	score	NOUN
ejpam-6087	374	18	for	for	ADP
ejpam-6087	374	19	precision	precision	NOUN
ejpam-6087	374	20	,	,	PUNCT
ejpam-6087	374	21	recall	recall	NOUN
ejpam-6087	374	22	,	,	PUNCT
ejpam-6087	374	23	and	and	CCONJ
ejpam-6087	374	24	f1	f1	PROPN
ejpam-6087	374	25	score	score	NOUN
ejpam-6087	374	26	.	.	PUNCT
ejpam-6087	375	1	the	the	DET
ejpam-6087	375	2	auc	auc	NOUN
ejpam-6087	375	3	-	-	PUNCT
ejpam-6087	375	4	roc	roc	NOUN
ejpam-6087	375	5	of	of	ADP
ejpam-6087	375	6	47.959	47.959	NUM
ejpam-6087	375	7	%	%	NOUN
ejpam-6087	375	8	reflects	reflect	VERB
ejpam-6087	375	9	its	its	PRON
ejpam-6087	375	10	unsuitability	unsuitability	NOUN
ejpam-6087	375	11	for	for	ADP
ejpam-6087	375	12	the	the	DET
ejpam-6087	375	13	detection	detection	NOUN
ejpam-6087	375	14	of	of	ADP
ejpam-6087	375	15	fake	fake	ADJ
ejpam-6087	375	16	news	news	NOUN
ejpam-6087	375	17	,	,	PUNCT
ejpam-6087	375	18	as	as	SCONJ
ejpam-6087	375	19	it	it	PRON
ejpam-6087	375	20	is	be	AUX
ejpam-6087	375	21	designed	design	VERB
ejpam-6087	375	22	for	for	ADP
ejpam-6087	375	23	the	the	DET
ejpam-6087	375	24	detection	detection	NOUN
ejpam-6087	375	25	of	of	ADP
ejpam-6087	375	26	anomalies	anomaly	NOUN
ejpam-6087	375	27	rather	rather	ADV
ejpam-6087	375	28	than	than	ADP
ejpam-6087	375	29	classification	classification	NOUN
ejpam-6087	375	30	.	.	PUNCT
ejpam-6087	376	1	table	table	NOUN
ejpam-6087	376	2	5	5	NUM
ejpam-6087	376	3	:	:	PUNCT
ejpam-6087	376	4	performance	performance	NOUN
ejpam-6087	376	5	of	of	ADP
ejpam-6087	376	6	ensemble	ensemble	ADJ
ejpam-6087	376	7	classifiers	classifier	NOUN
ejpam-6087	376	8	on	on	ADP
ejpam-6087	376	9	the	the	DET
ejpam-6087	376	10	truthseeker	truthseeker	NOUN
ejpam-6087	377	1	dataset	dataset	NOUN
ejpam-6087	377	2	classifier	classifier	PROPN
ejpam-6087	377	3	accuracy	accuracy	PROPN
ejpam-6087	377	4	precision	precision	NOUN
ejpam-6087	377	5	recall	recall	VERB
ejpam-6087	377	6	f1	f1	NOUN
ejpam-6087	377	7	-	-	PUNCT
ejpam-6087	377	8	score	score	NOUN
ejpam-6087	377	9	auc	auc	NOUN
ejpam-6087	377	10	-	-	PUNCT
ejpam-6087	377	11	roc	roc	NOUN
ejpam-6087	377	12	adaboost	adaboost	ADJ
ejpam-6087	377	13	(	(	PUNCT
ejpam-6087	377	14	decision	decision	NOUN
ejpam-6087	377	15	tree	tree	NOUN
ejpam-6087	377	16	)	)	PUNCT
ejpam-6087	378	1	91.025	91.025	NUM
ejpam-6087	378	2	%	%	NOUN
ejpam-6087	378	3	91.025	91.025	NUM
ejpam-6087	378	4	%	%	NOUN
ejpam-6087	378	5	91.025	91.025	NUM
ejpam-6087	378	6	%	%	NOUN
ejpam-6087	378	7	91.025	91.025	NUM
ejpam-6087	378	8	%	%	NOUN
ejpam-6087	378	9	91.019	91.019	NUM
ejpam-6087	378	10	%	%	NOUN
ejpam-6087	378	11	bagging	bagging	NOUN
ejpam-6087	378	12	(	(	PUNCT
ejpam-6087	378	13	decision	decision	NOUN
ejpam-6087	378	14	tree	tree	NOUN
ejpam-6087	378	15	)	)	PUNCT
ejpam-6087	378	16	95.738	95.738	NUM
ejpam-6087	378	17	%	%	NOUN
ejpam-6087	378	18	95.738	95.738	NUM
ejpam-6087	378	19	%	%	NOUN
ejpam-6087	378	20	95.738	95.738	NUM
ejpam-6087	378	21	%	%	NOUN
ejpam-6087	378	22	95.738	95.738	NUM
ejpam-6087	378	23	%	%	NOUN
ejpam-6087	378	24	95.893	95.893	NUM
ejpam-6087	378	25	%	%	NOUN
ejpam-6087	378	26	catboost	catboost	NOUN
ejpam-6087	378	27	95.801	95.801	NUM
ejpam-6087	378	28	%	%	NOUN
ejpam-6087	378	29	95.801	95.801	NUM
ejpam-6087	378	30	%	%	NOUN
ejpam-6087	378	31	95.801	95.801	NUM
ejpam-6087	378	32	%	%	NOUN
ejpam-6087	378	33	95.801	95.801	NUM
ejpam-6087	378	34	%	%	NOUN
ejpam-6087	378	35	95.859	95.859	NUM
ejpam-6087	378	36	%	%	NOUN
ejpam-6087	378	37	extra	extra	ADJ
ejpam-6087	378	38	trees	tree	NOUN
ejpam-6087	378	39	95.756	95.756	NUM
ejpam-6087	378	40	%	%	NOUN
ejpam-6087	378	41	95.757	95.757	NUM
ejpam-6087	378	42	%	%	NOUN
ejpam-6087	378	43	95.756	95.756	NUM
ejpam-6087	378	44	%	%	NOUN
ejpam-6087	378	45	95.756	95.756	NUM
ejpam-6087	378	46	%	%	NOUN
ejpam-6087	378	47	95.779	95.779	NUM
ejpam-6087	378	48	%	%	NOUN
ejpam-6087	378	49	gradient	gradient	NOUN
ejpam-6087	378	50	boosting	boost	VERB
ejpam-6087	378	51	95.801	95.801	NUM
ejpam-6087	378	52	%	%	NOUN
ejpam-6087	378	53	95.801	95.801	NUM
ejpam-6087	378	54	%	%	NOUN
ejpam-6087	378	55	95.801	95.801	NUM
ejpam-6087	378	56	%	%	NOUN
ejpam-6087	378	57	95.801	95.801	NUM
ejpam-6087	378	58	%	%	NOUN
ejpam-6087	378	59	95.827	95.827	NUM
ejpam-6087	378	60	%	%	NOUN
ejpam-6087	378	61	histgradient	histgradient	NOUN
ejpam-6087	378	62	boosting	boost	VERB
ejpam-6087	378	63	95.801	95.801	NUM
ejpam-6087	378	64	%	%	NOUN
ejpam-6087	378	65	95.801	95.801	NUM
ejpam-6087	378	66	%	%	NOUN
ejpam-6087	378	67	95.801	95.801	NUM
ejpam-6087	378	68	%	%	NOUN
ejpam-6087	378	69	95.801	95.801	NUM
ejpam-6087	378	70	%	%	NOUN
ejpam-6087	378	71	95.904	95.904	NUM
ejpam-6087	378	72	%	%	NOUN
ejpam-6087	378	73	isolation	isolation	NOUN
ejpam-6087	378	74	forest	forest	NOUN
ejpam-6087	378	75	51.140	51.140	NUM
ejpam-6087	378	76	%	%	NOUN
ejpam-6087	378	77	26.376	26.376	NUM
ejpam-6087	378	78	%	%	NOUN
ejpam-6087	378	79	51.140	51.140	NUM
ejpam-6087	378	80	%	%	NOUN
ejpam-6087	378	81	34.802	34.802	NUM
ejpam-6087	378	82	%	%	NOUN
ejpam-6087	378	83	47.959	47.959	NUM
ejpam-6087	378	84	%	%	NOUN
ejpam-6087	378	85	lightgbm	lightgbm	VERB
ejpam-6087	378	86	95.801	95.801	NUM
ejpam-6087	378	87	%	%	NOUN
ejpam-6087	378	88	95.801	95.801	NUM
ejpam-6087	378	89	%	%	NOUN
ejpam-6087	378	90	95.801	95.801	NUM
ejpam-6087	378	91	%	%	NOUN
ejpam-6087	378	92	95.801	95.801	NUM
ejpam-6087	378	93	%	%	NOUN
ejpam-6087	378	94	95.854	95.854	NUM
ejpam-6087	378	95	%	%	NOUN
ejpam-6087	378	96	stacking	stacking	NOUN
ejpam-6087	378	97	classifier	classifier	NOUN
ejpam-6087	378	98	95.801	95.801	NUM
ejpam-6087	378	99	%	%	NOUN
ejpam-6087	378	100	95.801	95.801	NUM
ejpam-6087	378	101	%	%	NOUN
ejpam-6087	378	102	95.801	95.801	NUM
ejpam-6087	378	103	%	%	NOUN
ejpam-6087	378	104	95.801	95.801	NUM
ejpam-6087	378	105	%	%	NOUN
ejpam-6087	378	106	95.816	95.816	NUM
ejpam-6087	378	107	%	%	NOUN
ejpam-6087	378	108	voting	voting	NOUN
ejpam-6087	378	109	classifier	classifier	NOUN
ejpam-6087	378	110	95.786	95.786	NUM
ejpam-6087	378	111	%	%	NOUN
ejpam-6087	378	112	95.786	95.786	NUM
ejpam-6087	378	113	%	%	NOUN
ejpam-6087	378	114	95.786	95.786	NUM
ejpam-6087	378	115	%	%	NOUN
ejpam-6087	378	116	95.786	95.786	NUM
ejpam-6087	378	117	%	%	NOUN
ejpam-6087	378	118	95.892	95.892	NUM
ejpam-6087	378	119	%	%	NOUN
ejpam-6087	378	120	xgboost	xgboost	ADP
ejpam-6087	378	121	95.790	95.790	NUM
ejpam-6087	378	122	%	%	NOUN
ejpam-6087	378	123	95.790	95.790	NUM
ejpam-6087	378	124	%	%	NOUN
ejpam-6087	378	125	95.790	95.790	NUM
ejpam-6087	378	126	%	%	NOUN
ejpam-6087	378	127	95.790	95.790	NUM
ejpam-6087	378	128	%	%	NOUN
ejpam-6087	378	129	95.880	95.880	NUM
ejpam-6087	378	130	%	%	NOUN
ejpam-6087	378	131	in	in	ADP
ejpam-6087	378	132	summary	summary	NOUN
ejpam-6087	378	133	,	,	PUNCT
ejpam-6087	378	134	the	the	DET
ejpam-6087	378	135	best	well	ADV
ejpam-6087	378	136	-	-	PUNCT
ejpam-6087	378	137	performing	perform	VERB
ejpam-6087	378	138	methods	method	NOUN
ejpam-6087	378	139	for	for	ADP
ejpam-6087	378	140	fake	fake	ADJ
ejpam-6087	378	141	news	news	NOUN
ejpam-6087	378	142	detection	detection	NOUN
ejpam-6087	378	143	are	be	AUX
ejpam-6087	378	144	catboost	catboost	ADJ
ejpam-6087	378	145	,	,	PUNCT
ejpam-6087	378	146	histgradient	histgradient	NOUN
ejpam-6087	378	147	boosting	boosting	NOUN
ejpam-6087	378	148	,	,	PUNCT
ejpam-6087	378	149	lightgbm	lightgbm	ADJ
ejpam-6087	378	150	,	,	PUNCT
ejpam-6087	378	151	and	and	CCONJ
ejpam-6087	378	152	xgboost	xgboost	X
ejpam-6087	378	153	.	.	PUNCT
ejpam-6087	379	1	among	among	ADP
ejpam-6087	379	2	these	these	PRON
ejpam-6087	379	3	,	,	PUNCT
ejpam-6087	379	4	histgradient	histgradient	NOUN
ejpam-6087	379	5	boosting	boosting	NOUN
ejpam-6087	379	6	stands	stand	VERB
ejpam-6087	379	7	out	out	ADP
ejpam-6087	379	8	with	with	ADP
ejpam-6087	379	9	the	the	DET
ejpam-6087	379	10	highest	high	ADJ
ejpam-6087	379	11	auc	auc	NOUN
ejpam-6087	379	12	-	-	PUNCT
ejpam-6087	379	13	roc	roc	NOUN
ejpam-6087	379	14	score	score	NOUN
ejpam-6087	379	15	,	,	PUNCT
ejpam-6087	379	16	while	while	SCONJ
ejpam-6087	379	17	catboost	catboost	ADJ
ejpam-6087	379	18	and	and	CCONJ
ejpam-6087	379	19	lightgbm	lightgbm	ADJ
ejpam-6087	379	20	also	also	ADV
ejpam-6087	379	21	demonstrate	demonstrate	VERB
ejpam-6087	379	22	exceptional	exceptional	ADJ
ejpam-6087	379	23	performance	performance	NOUN
ejpam-6087	379	24	across	across	ADP
ejpam-6087	379	25	all	all	DET
ejpam-6087	379	26	metrics	metric	NOUN
ejpam-6087	379	27	.	.	PUNCT
ejpam-6087	380	1	the	the	DET
ejpam-6087	380	2	stacking	stacking	NOUN
ejpam-6087	380	3	classifier	classifier	NOUN
ejpam-6087	380	4	and	and	CCONJ
ejpam-6087	380	5	voting	voting	NOUN
ejpam-6087	380	6	classifier	classifier	NOUN
ejpam-6087	380	7	are	be	AUX
ejpam-6087	380	8	also	also	ADV
ejpam-6087	380	9	effective	effective	ADJ
ejpam-6087	380	10	but	but	CCONJ
ejpam-6087	380	11	slightly	slightly	ADV
ejpam-6087	380	12	less	less	ADV
ejpam-6087	380	13	so	so	ADV
ejpam-6087	380	14	compared	compare	VERB
ejpam-6087	380	15	to	to	ADP
ejpam-6087	380	16	the	the	DET
ejpam-6087	380	17	aforementioned	aforementioned	ADJ
ejpam-6087	380	18	top	top	ADJ
ejpam-6087	380	19	methods	method	NOUN
ejpam-6087	380	20	.	.	PUNCT
ejpam-6087	381	1	bagging	bagging	NOUN
ejpam-6087	381	2	(	(	PUNCT
ejpam-6087	381	3	decision	decision	NOUN
ejpam-6087	381	4	tree	tree	NOUN
ejpam-6087	381	5	)	)	PUNCT
ejpam-6087	381	6	and	and	CCONJ
ejpam-6087	381	7	gradient	gradient	NOUN
ejpam-6087	381	8	boosting	boost	VERB
ejpam-6087	381	9	offer	offer	VERB
ejpam-6087	381	10	strong	strong	ADJ
ejpam-6087	381	11	performance	performance	NOUN
ejpam-6087	381	12	,	,	PUNCT
ejpam-6087	381	13	though	though	SCONJ
ejpam-6087	381	14	they	they	PRON
ejpam-6087	381	15	fall	fall	VERB
ejpam-6087	381	16	short	short	ADV
ejpam-6087	381	17	of	of	ADP
ejpam-6087	381	18	the	the	DET
ejpam-6087	381	19	best	good	ADJ
ejpam-6087	381	20	methods	method	NOUN
ejpam-6087	381	21	.	.	PUNCT
ejpam-6087	382	1	extra	extra	ADJ
ejpam-6087	382	2	trees	tree	NOUN
ejpam-6087	382	3	and	and	CCONJ
ejpam-6087	382	4	adaboost	adaboost	ADJ
ejpam-6087	382	5	(	(	PUNCT
ejpam-6087	382	6	decision	decision	NOUN
ejpam-6087	382	7	tree	tree	NOUN
ejpam-6087	382	8	)	)	PUNCT
ejpam-6087	382	9	show	show	VERB
ejpam-6087	382	10	solid	solid	ADJ
ejpam-6087	382	11	results	result	NOUN
ejpam-6087	382	12	but	but	CCONJ
ejpam-6087	382	13	are	be	AUX
ejpam-6087	382	14	not	not	PART
ejpam-6087	382	15	as	as	ADV
ejpam-6087	382	16	competitive	competitive	ADJ
ejpam-6087	382	17	,	,	PUNCT
ejpam-6087	382	18	while	while	SCONJ
ejpam-6087	382	19	isolation	isolation	NOUN
ejpam-6087	382	20	forest	forest	NOUN
ejpam-6087	382	21	proves	prove	VERB
ejpam-6087	382	22	to	to	PART
ejpam-6087	382	23	be	be	AUX
ejpam-6087	382	24	inadequate	inadequate	ADJ
ejpam-6087	382	25	for	for	ADP
ejpam-6087	382	26	this	this	DET
ejpam-6087	382	27	specific	specific	ADJ
ejpam-6087	382	28	classification	classification	NOUN
ejpam-6087	382	29	task	task	NOUN
ejpam-6087	382	30	.	.	PUNCT
ejpam-6087	383	1	given	give	VERB
ejpam-6087	383	2	that	that	DET
ejpam-6087	383	3	individual	individual	ADJ
ejpam-6087	383	4	classifiers	classifier	NOUN
ejpam-6087	383	5	and	and	CCONJ
ejpam-6087	383	6	ensemble	ensemble	ADJ
ejpam-6087	383	7	methods	method	NOUN
ejpam-6087	383	8	demonstrated	demonstrate	VERB
ejpam-6087	383	9	comparable	comparable	ADJ
ejpam-6087	383	10	levels	level	NOUN
ejpam-6087	383	11	of	of	ADP
ejpam-6087	383	12	performance	performance	NOUN
ejpam-6087	383	13	and	and	CCONJ
ejpam-6087	383	14	considering	consider	VERB
ejpam-6087	383	15	the	the	DET
ejpam-6087	383	16	reduced	reduced	ADJ
ejpam-6087	383	17	computational	computational	ADJ
ejpam-6087	383	18	complexity	complexity	NOUN
ejpam-6087	383	19	associated	associate	VERB
ejpam-6087	383	20	with	with	ADP
ejpam-6087	383	21	individual	individual	ADJ
ejpam-6087	383	22	models	model	NOUN
ejpam-6087	383	23	,	,	PUNCT
ejpam-6087	383	24	the	the	DET
ejpam-6087	383	25	subsequent	subsequent	ADJ
ejpam-6087	383	26	analysis	analysis	NOUN
ejpam-6087	383	27	focuses	focus	VERB
ejpam-6087	383	28	on	on	ADP
ejpam-6087	383	29	evaluating	evaluate	VERB
ejpam-6087	383	30	the	the	DET
ejpam-6087	383	31	impact	impact	NOUN
ejpam-6087	383	32	of	of	ADP
ejpam-6087	383	33	feature	feature	NOUN
ejpam-6087	383	34	selection	selection	NOUN
ejpam-6087	383	35	techniques	technique	NOUN
ejpam-6087	383	36	on	on	ADP
ejpam-6087	383	37	individual	individual	ADJ
ejpam-6087	383	38	classifiers	classifier	NOUN
ejpam-6087	383	39	.	.	PUNCT
ejpam-6087	384	1	in	in	ADP
ejpam-6087	384	2	particular	particular	ADJ
ejpam-6087	384	3	,	,	PUNCT
ejpam-6087	384	4	feature	feature	NOUN
ejpam-6087	384	5	selection	selection	NOUN
ejpam-6087	384	6	methods	method	NOUN
ejpam-6087	384	7	such	such	ADJ
ejpam-6087	384	8	as	as	ADP
ejpam-6087	384	9	rfe	rfe	PROPN
ejpam-6087	384	10	and	and	CCONJ
ejpam-6087	384	11	ga	ga	PROPN
ejpam-6087	384	12	were	be	AUX
ejpam-6087	384	13	applied	apply	VERB
ejpam-6087	384	14	exclusively	exclusively	ADV
ejpam-6087	384	15	to	to	ADP
ejpam-6087	384	16	individual	individual	ADJ
ejpam-6087	384	17	classifiers	classifier	NOUN
ejpam-6087	384	18	in	in	ADP
ejpam-6087	384	19	this	this	DET
ejpam-6087	384	20	study	study	NOUN
ejpam-6087	384	21	,	,	PUNCT
ejpam-6087	384	22	and	and	CCONJ
ejpam-6087	384	23	not	not	PART
ejpam-6087	384	24	to	to	PART
ejpam-6087	384	25	ensemble	ensemble	VERB
ejpam-6087	384	26	models	model	NOUN
ejpam-6087	384	27	.	.	PUNCT
ejpam-6087	385	1	this	this	DET
ejpam-6087	385	2	methodological	methodological	ADJ
ejpam-6087	385	3	choice	choice	NOUN
ejpam-6087	385	4	was	be	AUX
ejpam-6087	385	5	made	make	VERB
ejpam-6087	385	6	to	to	PART
ejpam-6087	385	7	isolate	isolate	VERB
ejpam-6087	385	8	and	and	CCONJ
ejpam-6087	385	9	assess	assess	VERB
ejpam-6087	385	10	the	the	DET
ejpam-6087	385	11	contribution	contribution	NOUN
ejpam-6087	385	12	of	of	ADP
ejpam-6087	385	13	feature	feature	NOUN
ejpam-6087	385	14	selection	selection	NOUN
ejpam-6087	385	15	strategies	strategy	NOUN
ejpam-6087	385	16	in	in	ADP
ejpam-6087	385	17	improving	improve	VERB
ejpam-6087	385	18	model	model	NOUN
ejpam-6087	385	19	performance	performance	NOUN
ejpam-6087	385	20	while	while	SCONJ
ejpam-6087	385	21	maintaining	maintain	VERB
ejpam-6087	385	22	computational	computational	ADJ
ejpam-6087	385	23	efficiency	efficiency	NOUN
ejpam-6087	385	24	.	.	PUNCT
ejpam-6087	386	1	the	the	DET
ejpam-6087	386	2	resulting	result	VERB
ejpam-6087	386	3	analysis	analysis	NOUN
ejpam-6087	386	4	offers	offer	VERB
ejpam-6087	386	5	valuable	valuable	ADJ
ejpam-6087	386	6	information	information	NOUN
ejpam-6087	386	7	on	on	ADP
ejpam-6087	386	8	the	the	DET
ejpam-6087	386	9	potential	potential	NOUN
ejpam-6087	386	10	of	of	ADP
ejpam-6087	386	11	optimized	optimize	VERB
ejpam-6087	386	12	individual	individual	ADJ
ejpam-6087	386	13	classifiers	classifier	NOUN
ejpam-6087	386	14	to	to	PART
ejpam-6087	386	15	deliver	deliver	VERB
ejpam-6087	386	16	competitive	competitive	ADJ
ejpam-6087	386	17	performance	performance	NOUN
ejpam-6087	386	18	with	with	ADP
ejpam-6087	386	19	reduced	reduce	VERB
ejpam-6087	386	20	m.	m.	NOUN
ejpam-6087	386	21	al	al	PROPN
ejpam-6087	386	22	-	-	PUNCT
ejpam-6087	386	23	tarawneh	tarawneh	PROPN
ejpam-6087	386	24	et	et	PROPN
ejpam-6087	386	25	al	al	PROPN
ejpam-6087	386	26	.	.	PUNCT
ejpam-6087	386	27	/	/	SYM
ejpam-6087	386	28	eur	eur	PROPN
ejpam-6087	386	29	.	.	PUNCT
ejpam-6087	387	1	j.	j.	PROPN
ejpam-6087	387	2	pure	pure	PROPN
ejpam-6087	387	3	appl	appl	PROPN
ejpam-6087	387	4	.	.	PROPN
ejpam-6087	387	5	math	math	PROPN
ejpam-6087	387	6	,	,	PUNCT
ejpam-6087	387	7	18	18	NUM
ejpam-6087	387	8	(	(	PUNCT
ejpam-6087	387	9	2	2	NUM
ejpam-6087	387	10	)	)	PUNCT
ejpam-6087	387	11	(	(	PUNCT
ejpam-6087	387	12	2025	2025	NUM
ejpam-6087	387	13	)	)	PUNCT
ejpam-6087	387	14	,	,	PUNCT
ejpam-6087	387	15	6087	6087	NUM
ejpam-6087	387	16	21	21	NUM
ejpam-6087	387	17	of	of	ADP
ejpam-6087	387	18	39	39	NUM
ejpam-6087	387	19	training	training	NOUN
ejpam-6087	387	20	overhead	overhead	ADV
ejpam-6087	387	21	,	,	PUNCT
ejpam-6087	387	22	thus	thus	ADV
ejpam-6087	387	23	supporting	support	VERB
ejpam-6087	387	24	their	their	PRON
ejpam-6087	387	25	applicability	applicability	NOUN
ejpam-6087	387	26	in	in	ADP
ejpam-6087	387	27	resource	resource	NOUN
ejpam-6087	387	28	-	-	PUNCT
ejpam-6087	387	29	constrained	constrain	VERB
ejpam-6087	387	30	or	or	CCONJ
ejpam-6087	387	31	real	real	ADJ
ejpam-6087	387	32	-	-	PUNCT
ejpam-6087	387	33	time	time	NOUN
ejpam-6087	387	34	environments	environment	NOUN
ejpam-6087	387	35	.	.	PUNCT
ejpam-6087	388	1	4.3	4.3	NUM
ejpam-6087	388	2	.	.	PUNCT
ejpam-6087	388	3	impact	impact	NOUN
ejpam-6087	388	4	of	of	ADP
ejpam-6087	388	5	feature	feature	NOUN
ejpam-6087	388	6	selection	selection	NOUN
ejpam-6087	388	7	on	on	ADP
ejpam-6087	388	8	machine	machine	NOUN
ejpam-6087	388	9	learning	learn	VERB
ejpam-6087	388	10	algorithms	algorithm	NOUN
ejpam-6087	388	11	feature	feature	NOUN
ejpam-6087	388	12	selection	selection	NOUN
ejpam-6087	388	13	plays	play	VERB
ejpam-6087	388	14	a	a	DET
ejpam-6087	388	15	crucial	crucial	ADJ
ejpam-6087	388	16	role	role	NOUN
ejpam-6087	388	17	in	in	ADP
ejpam-6087	388	18	enhancing	enhance	VERB
ejpam-6087	388	19	the	the	DET
ejpam-6087	388	20	performance	performance	NOUN
ejpam-6087	388	21	and	and	CCONJ
ejpam-6087	388	22	efficiency	efficiency	NOUN
ejpam-6087	388	23	of	of	ADP
ejpam-6087	388	24	ml	ml	NOUN
ejpam-6087	388	25	models	model	NOUN
ejpam-6087	388	26	.	.	PUNCT
ejpam-6087	389	1	this	this	DET
ejpam-6087	389	2	section	section	NOUN
ejpam-6087	389	3	explores	explore	VERB
ejpam-6087	389	4	the	the	DET
ejpam-6087	389	5	impact	impact	NOUN
ejpam-6087	389	6	of	of	ADP
ejpam-6087	389	7	different	different	ADJ
ejpam-6087	389	8	feature	feature	NOUN
ejpam-6087	389	9	selection	selection	NOUN
ejpam-6087	389	10	techniques	technique	NOUN
ejpam-6087	389	11	on	on	ADP
ejpam-6087	389	12	the	the	DET
ejpam-6087	389	13	performance	performance	NOUN
ejpam-6087	389	14	of	of	ADP
ejpam-6087	389	15	various	various	ADJ
ejpam-6087	389	16	classifiers	classifier	NOUN
ejpam-6087	389	17	in	in	ADP
ejpam-6087	389	18	the	the	DET
ejpam-6087	389	19	context	context	NOUN
ejpam-6087	389	20	of	of	ADP
ejpam-6087	389	21	fake	fake	ADJ
ejpam-6087	389	22	news	news	NOUN
ejpam-6087	389	23	detection	detection	NOUN
ejpam-6087	389	24	.	.	PUNCT
ejpam-6087	390	1	specifically	specifically	ADV
ejpam-6087	390	2	,	,	PUNCT
ejpam-6087	390	3	the	the	DET
ejpam-6087	390	4	section	section	NOUN
ejpam-6087	390	5	investigates	investigate	VERB
ejpam-6087	390	6	how	how	SCONJ
ejpam-6087	390	7	the	the	DET
ejpam-6087	390	8	performance	performance	NOUN
ejpam-6087	390	9	of	of	ADP
ejpam-6087	390	10	ml	ml	ADP
ejpam-6087	390	11	algorithms	algorithm	NOUN
ejpam-6087	390	12	is	be	AUX
ejpam-6087	390	13	affected	affect	VERB
ejpam-6087	390	14	by	by	ADP
ejpam-6087	390	15	selecting	select	VERB
ejpam-6087	390	16	different	different	ADJ
ejpam-6087	390	17	subsets	subset	NOUN
ejpam-6087	390	18	of	of	ADP
ejpam-6087	390	19	features	feature	NOUN
ejpam-6087	390	20	and	and	CCONJ
ejpam-6087	390	21	reducing	reduce	VERB
ejpam-6087	390	22	dimensionality	dimensionality	NOUN
ejpam-6087	390	23	.	.	PUNCT
ejpam-6087	391	1	the	the	DET
ejpam-6087	391	2	analysis	analysis	NOUN
ejpam-6087	391	3	begins	begin	VERB
ejpam-6087	391	4	by	by	ADP
ejpam-6087	391	5	evaluating	evaluate	VERB
ejpam-6087	391	6	the	the	DET
ejpam-6087	391	7	effectiveness	effectiveness	NOUN
ejpam-6087	391	8	of	of	ADP
ejpam-6087	391	9	feature	feature	NOUN
ejpam-6087	391	10	selection	selection	NOUN
ejpam-6087	391	11	methods	method	NOUN
ejpam-6087	391	12	such	such	ADJ
ejpam-6087	391	13	as	as	ADP
ejpam-6087	391	14	rfe	rfe	NOUN
ejpam-6087	391	15	and	and	CCONJ
ejpam-6087	391	16	selectkbest	selectkb	ADJ
ejpam-6087	391	17	.	.	PUNCT
ejpam-6087	392	1	these	these	DET
ejpam-6087	392	2	methods	method	NOUN
ejpam-6087	392	3	are	be	AUX
ejpam-6087	392	4	examined	examine	VERB
ejpam-6087	392	5	by	by	ADP
ejpam-6087	392	6	varying	vary	VERB
ejpam-6087	392	7	the	the	DET
ejpam-6087	392	8	number	number	NOUN
ejpam-6087	392	9	of	of	ADP
ejpam-6087	392	10	features	feature	NOUN
ejpam-6087	392	11	from	from	ADP
ejpam-6087	392	12	5	5	NUM
ejpam-6087	392	13	to	to	PART
ejpam-6087	392	14	30	30	NUM
ejpam-6087	392	15	and	and	CCONJ
ejpam-6087	392	16	assessing	assess	VERB
ejpam-6087	392	17	the	the	DET
ejpam-6087	392	18	impact	impact	NOUN
ejpam-6087	392	19	on	on	ADP
ejpam-6087	392	20	the	the	DET
ejpam-6087	392	21	performance	performance	NOUN
ejpam-6087	392	22	metrics	metric	NOUN
ejpam-6087	392	23	of	of	ADP
ejpam-6087	392	24	classifiers	classifier	NOUN
ejpam-6087	392	25	including	include	VERB
ejpam-6087	392	26	lr	lr	PROPN
ejpam-6087	392	27	,	,	PUNCT
ejpam-6087	392	28	nbs	nbs	PROPN
ejpam-6087	392	29	.	.	PUNCT
ejpam-6087	393	1	performance	performance	NOUN
ejpam-6087	393	2	metrics	metric	NOUN
ejpam-6087	393	3	such	such	ADJ
ejpam-6087	393	4	as	as	ADP
ejpam-6087	393	5	accuracy	accuracy	NOUN
ejpam-6087	393	6	,	,	PUNCT
ejpam-6087	393	7	precision	precision	NOUN
ejpam-6087	393	8	,	,	PUNCT
ejpam-6087	393	9	recall	recall	NOUN
ejpam-6087	393	10	,	,	PUNCT
ejpam-6087	393	11	f1	f1	NOUN
ejpam-6087	393	12	score	score	NOUN
ejpam-6087	393	13	,	,	PUNCT
ejpam-6087	393	14	and	and	CCONJ
ejpam-6087	393	15	auc	auc	NOUN
ejpam-6087	393	16	-	-	PUNCT
ejpam-6087	393	17	roc	roc	NOUN
ejpam-6087	393	18	are	be	AUX
ejpam-6087	393	19	used	use	VERB
ejpam-6087	393	20	to	to	PART
ejpam-6087	393	21	gauge	gauge	VERB
ejpam-6087	393	22	how	how	SCONJ
ejpam-6087	393	23	the	the	DET
ejpam-6087	393	24	increasing	increase	VERB
ejpam-6087	393	25	number	number	NOUN
ejpam-6087	393	26	of	of	ADP
ejpam-6087	393	27	features	feature	NOUN
ejpam-6087	393	28	influences	influence	VERB
ejpam-6087	393	29	the	the	DET
ejpam-6087	393	30	effectiveness	effectiveness	NOUN
ejpam-6087	393	31	of	of	ADP
ejpam-6087	393	32	these	these	DET
ejpam-6087	393	33	models	model	NOUN
ejpam-6087	393	34	.	.	PUNCT
ejpam-6087	394	1	subsequently	subsequently	ADV
ejpam-6087	394	2	,	,	PUNCT
ejpam-6087	394	3	pca	pca	PROPN
ejpam-6087	394	4	is	be	AUX
ejpam-6087	394	5	explored	explore	VERB
ejpam-6087	394	6	as	as	ADP
ejpam-6087	394	7	a	a	DET
ejpam-6087	394	8	technique	technique	NOUN
ejpam-6087	394	9	for	for	ADP
ejpam-6087	394	10	dimensionality	dimensionality	NOUN
ejpam-6087	394	11	reduction	reduction	NOUN
ejpam-6087	394	12	.	.	PUNCT
ejpam-6087	395	1	the	the	DET
ejpam-6087	395	2	performance	performance	NOUN
ejpam-6087	395	3	of	of	ADP
ejpam-6087	395	4	classifiers	classifier	NOUN
ejpam-6087	395	5	is	be	AUX
ejpam-6087	395	6	analyzed	analyze	VERB
ejpam-6087	395	7	as	as	SCONJ
ejpam-6087	395	8	the	the	DET
ejpam-6087	395	9	number	number	NOUN
ejpam-6087	395	10	of	of	ADP
ejpam-6087	395	11	principal	principal	ADJ
ejpam-6087	395	12	components	component	NOUN
ejpam-6087	395	13	is	be	AUX
ejpam-6087	395	14	varied	varied	ADJ
ejpam-6087	395	15	from	from	ADP
ejpam-6087	395	16	5	5	NUM
ejpam-6087	395	17	to	to	PART
ejpam-6087	395	18	30	30	NUM
ejpam-6087	395	19	,	,	PUNCT
ejpam-6087	395	20	providing	provide	VERB
ejpam-6087	395	21	insights	insight	NOUN
ejpam-6087	395	22	into	into	ADP
ejpam-6087	395	23	how	how	SCONJ
ejpam-6087	395	24	dimensionality	dimensionality	NOUN
ejpam-6087	395	25	reduction	reduction	NOUN
ejpam-6087	395	26	affects	affect	VERB
ejpam-6087	395	27	model	model	NOUN
ejpam-6087	395	28	performance	performance	NOUN
ejpam-6087	395	29	.	.	PUNCT
ejpam-6087	396	1	finally	finally	ADV
ejpam-6087	396	2	,	,	PUNCT
ejpam-6087	396	3	the	the	DET
ejpam-6087	396	4	section	section	NOUN
ejpam-6087	396	5	covers	cover	VERB
ejpam-6087	396	6	the	the	DET
ejpam-6087	396	7	use	use	NOUN
ejpam-6087	396	8	of	of	ADP
ejpam-6087	396	9	ga	ga	PROPN
ejpam-6087	396	10	for	for	ADP
ejpam-6087	396	11	feature	feature	NOUN
ejpam-6087	396	12	selection	selection	NOUN
ejpam-6087	396	13	,	,	PUNCT
ejpam-6087	396	14	focusing	focus	VERB
ejpam-6087	396	15	on	on	ADP
ejpam-6087	396	16	how	how	SCONJ
ejpam-6087	396	17	the	the	DET
ejpam-6087	396	18	optimal	optimal	ADJ
ejpam-6087	396	19	feature	feature	NOUN
ejpam-6087	396	20	subset	subset	NOUN
ejpam-6087	396	21	identified	identify	VERB
ejpam-6087	396	22	by	by	ADP
ejpam-6087	396	23	the	the	DET
ejpam-6087	396	24	ga	ga	PROPN
ejpam-6087	396	25	affects	affect	VERB
ejpam-6087	396	26	the	the	DET
ejpam-6087	396	27	performance	performance	NOUN
ejpam-6087	396	28	of	of	ADP
ejpam-6087	396	29	ml	ml	NOUN
ejpam-6087	396	30	algorithms	algorithm	NOUN
ejpam-6087	396	31	.	.	PUNCT
ejpam-6087	397	1	the	the	DET
ejpam-6087	397	2	ga	ga	PROPN
ejpam-6087	397	3	approach	approach	NOUN
ejpam-6087	397	4	is	be	AUX
ejpam-6087	397	5	evaluated	evaluate	VERB
ejpam-6087	397	6	on	on	ADP
ejpam-6087	397	7	the	the	DET
ejpam-6087	397	8	basis	basis	NOUN
ejpam-6087	397	9	of	of	ADP
ejpam-6087	397	10	the	the	DET
ejpam-6087	397	11	optimal	optimal	ADJ
ejpam-6087	397	12	number	number	NOUN
ejpam-6087	397	13	of	of	ADP
ejpam-6087	397	14	features	feature	NOUN
ejpam-6087	397	15	selected	select	VERB
ejpam-6087	397	16	and	and	CCONJ
ejpam-6087	397	17	compared	compare	VERB
ejpam-6087	397	18	with	with	ADP
ejpam-6087	397	19	the	the	DET
ejpam-6087	397	20	results	result	NOUN
ejpam-6087	397	21	obtained	obtain	VERB
ejpam-6087	397	22	from	from	ADP
ejpam-6087	397	23	rfe	rfe	NOUN
ejpam-6087	397	24	,	,	PUNCT
ejpam-6087	397	25	selectkbest	selectkb	ADJ
ejpam-6087	397	26	,	,	PUNCT
ejpam-6087	397	27	and	and	CCONJ
ejpam-6087	397	28	pca	pca	PROPN
ejpam-6087	397	29	.	.	PUNCT
ejpam-6087	398	1	through	through	ADP
ejpam-6087	398	2	this	this	DET
ejpam-6087	398	3	examination	examination	NOUN
ejpam-6087	398	4	,	,	PUNCT
ejpam-6087	398	5	the	the	DET
ejpam-6087	398	6	purpose	purpose	NOUN
ejpam-6087	398	7	of	of	ADP
ejpam-6087	398	8	this	this	DET
ejpam-6087	398	9	section	section	NOUN
ejpam-6087	398	10	is	be	AUX
ejpam-6087	398	11	to	to	PART
ejpam-6087	398	12	provide	provide	VERB
ejpam-6087	398	13	a	a	DET
ejpam-6087	398	14	comprehensive	comprehensive	ADJ
ejpam-6087	398	15	understanding	understanding	NOUN
ejpam-6087	398	16	of	of	ADP
ejpam-6087	398	17	how	how	SCONJ
ejpam-6087	398	18	different	different	ADJ
ejpam-6087	398	19	feature	feature	NOUN
ejpam-6087	398	20	selection	selection	NOUN
ejpam-6087	398	21	strategies	strategy	NOUN
ejpam-6087	398	22	and	and	CCONJ
ejpam-6087	398	23	feature	feature	NOUN
ejpam-6087	398	24	dimensions	dimension	NOUN
ejpam-6087	398	25	impact	impact	VERB
ejpam-6087	398	26	the	the	DET
ejpam-6087	398	27	performance	performance	NOUN
ejpam-6087	398	28	of	of	ADP
ejpam-6087	398	29	ml	ml	NOUN
ejpam-6087	398	30	algorithms	algorithm	NOUN
ejpam-6087	398	31	.	.	PUNCT
ejpam-6087	399	1	4.3.1	4.3.1	X
ejpam-6087	399	2	.	.	PUNCT
ejpam-6087	399	3	recursive	recursive	ADJ
ejpam-6087	399	4	feature	feature	NOUN
ejpam-6087	399	5	elimination	elimination	NOUN
ejpam-6087	399	6	results	result	NOUN
ejpam-6087	399	7	figure	figure	VERB
ejpam-6087	399	8	1	1	NUM
ejpam-6087	399	9	shows	show	VERB
ejpam-6087	399	10	the	the	DET
ejpam-6087	399	11	performance	performance	NOUN
ejpam-6087	399	12	of	of	ADP
ejpam-6087	399	13	various	various	ADJ
ejpam-6087	399	14	machine	machine	NOUN
ejpam-6087	399	15	algorithms	algorithm	NOUN
ejpam-6087	399	16	as	as	ADP
ejpam-6087	399	17	the	the	DET
ejpam-6087	399	18	number	number	NOUN
ejpam-6087	399	19	of	of	ADP
ejpam-6087	399	20	selected	select	VERB
ejpam-6087	399	21	features	feature	NOUN
ejpam-6087	399	22	increases	increase	NOUN
ejpam-6087	399	23	from	from	ADP
ejpam-6087	399	24	5	5	NUM
ejpam-6087	399	25	to	to	PART
ejpam-6087	399	26	30	30	NUM
ejpam-6087	399	27	using	use	VERB
ejpam-6087	399	28	the	the	DET
ejpam-6087	399	29	rfe	rfe	NOUN
ejpam-6087	399	30	feature	feature	NOUN
ejpam-6087	399	31	selection	selection	NOUN
ejpam-6087	399	32	method	method	NOUN
ejpam-6087	399	33	.	.	PUNCT
ejpam-6087	400	1	for	for	ADP
ejpam-6087	400	2	example	example	NOUN
ejpam-6087	400	3	,	,	PUNCT
ejpam-6087	400	4	figure	figure	NOUN
ejpam-6087	400	5	1a	1a	PROPN
ejpam-6087	400	6	illustrates	illustrate	VERB
ejpam-6087	400	7	the	the	DET
ejpam-6087	400	8	accuracy	accuracy	NOUN
ejpam-6087	400	9	of	of	ADP
ejpam-6087	400	10	the	the	DET
ejpam-6087	400	11	model	model	NOUN
ejpam-6087	400	12	as	as	ADP
ejpam-6087	400	13	the	the	DET
ejpam-6087	400	14	number	number	NOUN
ejpam-6087	400	15	of	of	ADP
ejpam-6087	400	16	features	feature	NOUN
ejpam-6087	400	17	increases	increase	NOUN
ejpam-6087	400	18	from	from	ADP
ejpam-6087	400	19	5	5	NUM
ejpam-6087	400	20	to	to	PART
ejpam-6087	400	21	30	30	NUM
ejpam-6087	400	22	.	.	PUNCT
ejpam-6087	401	1	the	the	DET
ejpam-6087	401	2	accuracy	accuracy	NOUN
ejpam-6087	401	3	results	result	VERB
ejpam-6087	401	4	for	for	ADP
ejpam-6087	401	5	different	different	ADJ
ejpam-6087	401	6	models	model	NOUN
ejpam-6087	401	7	using	use	VERB
ejpam-6087	401	8	rfe	rfe	NOUN
ejpam-6087	401	9	exhibit	exhibit	VERB
ejpam-6087	401	10	various	various	ADJ
ejpam-6087	401	11	trends	trend	NOUN
ejpam-6087	401	12	as	as	ADP
ejpam-6087	401	13	the	the	DET
ejpam-6087	401	14	number	number	NOUN
ejpam-6087	401	15	of	of	ADP
ejpam-6087	401	16	features	feature	NOUN
ejpam-6087	401	17	increases	increase	NOUN
ejpam-6087	401	18	.	.	PUNCT
ejpam-6087	402	1	lr	lr	PROPN
ejpam-6087	402	2	maintains	maintain	VERB
ejpam-6087	402	3	a	a	DET
ejpam-6087	402	4	high	high	ADJ
ejpam-6087	402	5	accuracy	accuracy	NOUN
ejpam-6087	402	6	of	of	ADP
ejpam-6087	402	7	95.80	95.80	NUM
ejpam-6087	402	8	%	%	NOUN
ejpam-6087	402	9	with	with	ADP
ejpam-6087	402	10	5	5	NUM
ejpam-6087	402	11	features	feature	NOUN
ejpam-6087	402	12	,	,	PUNCT
ejpam-6087	402	13	which	which	PRON
ejpam-6087	402	14	remains	remain	VERB
ejpam-6087	402	15	constant	constant	ADJ
ejpam-6087	402	16	up	up	ADP
ejpam-6087	402	17	to	to	PART
ejpam-6087	402	18	10	10	NUM
ejpam-6087	402	19	features	feature	NOUN
ejpam-6087	402	20	,	,	PUNCT
ejpam-6087	402	21	but	but	CCONJ
ejpam-6087	402	22	then	then	ADV
ejpam-6087	402	23	decreases	decrease	VERB
ejpam-6087	402	24	to	to	ADP
ejpam-6087	402	25	60.19	60.19	NUM
ejpam-6087	402	26	%	%	NOUN
ejpam-6087	402	27	with	with	ADP
ejpam-6087	402	28	30	30	NUM
ejpam-6087	402	29	features	feature	NOUN
ejpam-6087	402	30	.	.	PUNCT
ejpam-6087	403	1	nb	nb	PROPN
ejpam-6087	403	2	shows	show	VERB
ejpam-6087	403	3	a	a	DET
ejpam-6087	403	4	low	low	ADJ
ejpam-6087	403	5	and	and	CCONJ
ejpam-6087	403	6	relatively	relatively	ADV
ejpam-6087	403	7	constant	constant	ADJ
ejpam-6087	403	8	accuracy	accuracy	NOUN
ejpam-6087	403	9	in	in	ADP
ejpam-6087	403	10	all	all	DET
ejpam-6087	403	11	feature	feature	NOUN
ejpam-6087	403	12	sets	set	NOUN
ejpam-6087	403	13	,	,	PUNCT
ejpam-6087	403	14	with	with	ADP
ejpam-6087	403	15	52.28	52.28	NUM
ejpam-6087	403	16	%	%	NOUN
ejpam-6087	403	17	at	at	ADP
ejpam-6087	403	18	5	5	NUM
ejpam-6087	403	19	features	feature	NOUN
ejpam-6087	403	20	and	and	CCONJ
ejpam-6087	403	21	49.88	49.88	NUM
ejpam-6087	403	22	%	%	NOUN
ejpam-6087	403	23	at	at	ADP
ejpam-6087	403	24	30	30	NUM
ejpam-6087	403	25	features	feature	NOUN
ejpam-6087	403	26	.	.	PUNCT
ejpam-6087	404	1	knn	knn	PROPN
ejpam-6087	404	2	starts	start	VERB
ejpam-6087	404	3	with	with	ADP
ejpam-6087	404	4	a	a	DET
ejpam-6087	404	5	higher	high	ADJ
ejpam-6087	404	6	accuracy	accuracy	NOUN
ejpam-6087	404	7	of	of	ADP
ejpam-6087	404	8	72.84	72.84	NUM
ejpam-6087	404	9	%	%	NOUN
ejpam-6087	404	10	at	at	ADP
ejpam-6087	404	11	five	five	NUM
ejpam-6087	404	12	features	feature	NOUN
ejpam-6087	404	13	and	and	CCONJ
ejpam-6087	404	14	then	then	ADV
ejpam-6087	404	15	stabilizes	stabilize	VERB
ejpam-6087	404	16	at	at	ADP
ejpam-6087	404	17	58.73	58.73	NUM
ejpam-6087	404	18	%	%	NOUN
ejpam-6087	404	19	from	from	ADP
ejpam-6087	404	20	10	10	NUM
ejpam-6087	404	21	features	feature	NOUN
ejpam-6087	404	22	onward	onward	ADV
ejpam-6087	404	23	.	.	PUNCT
ejpam-6087	405	1	svm	svm	PROPN
ejpam-6087	405	2	shows	show	VERB
ejpam-6087	405	3	a	a	DET
ejpam-6087	405	4	consistently	consistently	ADV
ejpam-6087	405	5	low	low	ADJ
ejpam-6087	405	6	accuracy	accuracy	NOUN
ejpam-6087	405	7	of	of	ADP
ejpam-6087	405	8	around	around	ADP
ejpam-6087	405	9	53.49	53.49	NUM
ejpam-6087	405	10	%	%	NOUN
ejpam-6087	405	11	in	in	ADP
ejpam-6087	405	12	all	all	DET
ejpam-6087	405	13	feature	feature	NOUN
ejpam-6087	405	14	sets	set	NOUN
ejpam-6087	405	15	,	,	PUNCT
ejpam-6087	405	16	indicating	indicate	VERB
ejpam-6087	405	17	a	a	DET
ejpam-6087	405	18	minimal	minimal	ADJ
ejpam-6087	405	19	sensitivity	sensitivity	NOUN
ejpam-6087	405	20	to	to	ADP
ejpam-6087	405	21	the	the	DET
ejpam-6087	405	22	number	number	NOUN
ejpam-6087	405	23	of	of	ADP
ejpam-6087	405	24	features	feature	NOUN
ejpam-6087	405	25	.	.	PUNCT
ejpam-6087	406	1	mlp	mlp	NOUN
ejpam-6087	406	2	starts	start	VERB
ejpam-6087	406	3	with	with	ADP
ejpam-6087	406	4	a	a	DET
ejpam-6087	406	5	high	high	ADJ
ejpam-6087	406	6	accuracy	accuracy	NOUN
ejpam-6087	406	7	of	of	ADP
ejpam-6087	406	8	95.73	95.73	NUM
ejpam-6087	406	9	%	%	NOUN
ejpam-6087	406	10	at	at	ADP
ejpam-6087	406	11	5	5	NUM
ejpam-6087	406	12	features	feature	NOUN
ejpam-6087	406	13	,	,	PUNCT
ejpam-6087	406	14	decreases	decrease	VERB
ejpam-6087	406	15	to	to	ADP
ejpam-6087	406	16	79.09	79.09	NUM
ejpam-6087	406	17	%	%	NOUN
ejpam-6087	406	18	at	at	ADP
ejpam-6087	406	19	10	10	NUM
ejpam-6087	406	20	features	feature	NOUN
ejpam-6087	406	21	,	,	PUNCT
ejpam-6087	406	22	and	and	CCONJ
ejpam-6087	406	23	fluctuates	fluctuate	NOUN
ejpam-6087	406	24	before	before	ADP
ejpam-6087	406	25	increasing	increase	VERB
ejpam-6087	406	26	to	to	ADP
ejpam-6087	406	27	86.53	86.53	NUM
ejpam-6087	406	28	%	%	NOUN
ejpam-6087	406	29	at	at	ADP
ejpam-6087	406	30	30	30	NUM
ejpam-6087	406	31	features	feature	NOUN
ejpam-6087	406	32	.	.	PUNCT
ejpam-6087	407	1	dt	dt	PROPN
ejpam-6087	407	2	exhibits	exhibit	VERB
ejpam-6087	407	3	high	high	ADJ
ejpam-6087	407	4	and	and	CCONJ
ejpam-6087	407	5	stable	stable	ADJ
ejpam-6087	407	6	accuracy	accuracy	NOUN
ejpam-6087	407	7	,	,	PUNCT
ejpam-6087	407	8	ranging	range	VERB
ejpam-6087	407	9	from	from	ADP
ejpam-6087	407	10	91.36	91.36	NUM
ejpam-6087	407	11	%	%	NOUN
ejpam-6087	407	12	at	at	ADP
ejpam-6087	407	13	5	5	NUM
ejpam-6087	407	14	features	feature	NOUN
ejpam-6087	407	15	to	to	ADP
ejpam-6087	407	16	90.80	90.80	NUM
ejpam-6087	407	17	%	%	NOUN
ejpam-6087	407	18	at	at	ADP
ejpam-6087	407	19	30	30	NUM
ejpam-6087	407	20	features	feature	NOUN
ejpam-6087	407	21	.	.	PUNCT
ejpam-6087	408	1	finally	finally	ADV
ejpam-6087	408	2	,	,	PUNCT
ejpam-6087	408	3	rf	rf	NOUN
ejpam-6087	408	4	maintains	maintain	VERB
ejpam-6087	408	5	a	a	DET
ejpam-6087	408	6	high	high	ADJ
ejpam-6087	408	7	and	and	CCONJ
ejpam-6087	408	8	consistent	consistent	ADJ
ejpam-6087	408	9	accuracy	accuracy	NOUN
ejpam-6087	408	10	from	from	ADP
ejpam-6087	408	11	95.70	95.70	NUM
ejpam-6087	408	12	%	%	NOUN
ejpam-6087	408	13	at	at	ADP
ejpam-6087	408	14	5	5	NUM
ejpam-6087	408	15	features	feature	NOUN
ejpam-6087	408	16	to	to	ADP
ejpam-6087	408	17	95.80	95.80	NUM
ejpam-6087	408	18	%	%	NOUN
ejpam-6087	408	19	at	at	ADP
ejpam-6087	408	20	30	30	NUM
ejpam-6087	408	21	features	feature	NOUN
ejpam-6087	408	22	,	,	PUNCT
ejpam-6087	408	23	demonstrating	demonstrate	VERB
ejpam-6087	408	24	the	the	DET
ejpam-6087	408	25	most	most	ADV
ejpam-6087	408	26	stable	stable	ADJ
ejpam-6087	408	27	performance	performance	NOUN
ejpam-6087	408	28	.	.	PUNCT
ejpam-6087	409	1	on	on	ADP
ejpam-6087	409	2	the	the	DET
ejpam-6087	409	3	other	other	ADJ
ejpam-6087	409	4	hand	hand	NOUN
ejpam-6087	409	5	,	,	PUNCT
ejpam-6087	409	6	figure	figure	VERB
ejpam-6087	409	7	1b	1b	NUM
ejpam-6087	409	8	shows	show	VERB
ejpam-6087	409	9	precision	precision	NOUN
ejpam-6087	409	10	values	value	NOUN
ejpam-6087	409	11	as	as	ADP
ejpam-6087	409	12	the	the	DET
ejpam-6087	409	13	number	number	NOUN
ejpam-6087	409	14	of	of	ADP
ejpam-6087	409	15	features	feature	NOUN
ejpam-6087	409	16	increases	increase	NOUN
ejpam-6087	409	17	.	.	PUNCT
ejpam-6087	410	1	m.	m.	NOUN
ejpam-6087	410	2	al	al	PROPN
ejpam-6087	410	3	-	-	PUNCT
ejpam-6087	410	4	tarawneh	tarawneh	PROPN
ejpam-6087	410	5	et	et	PROPN
ejpam-6087	410	6	al	al	PROPN
ejpam-6087	410	7	.	.	PUNCT
ejpam-6087	410	8	/	/	SYM
ejpam-6087	410	9	eur	eur	PROPN
ejpam-6087	410	10	.	.	PUNCT
ejpam-6087	411	1	j.	j.	PROPN
ejpam-6087	411	2	pure	pure	PROPN
ejpam-6087	411	3	appl	appl	PROPN
ejpam-6087	411	4	.	.	PROPN
ejpam-6087	411	5	math	math	PROPN
ejpam-6087	411	6	,	,	PUNCT
ejpam-6087	411	7	18	18	NUM
ejpam-6087	411	8	(	(	PUNCT
ejpam-6087	411	9	2	2	NUM
ejpam-6087	411	10	)	)	PUNCT
ejpam-6087	411	11	(	(	PUNCT
ejpam-6087	411	12	2025	2025	NUM
ejpam-6087	411	13	)	)	PUNCT
ejpam-6087	411	14	,	,	PUNCT
ejpam-6087	411	15	6087	6087	NUM
ejpam-6087	411	16	22	22	NUM
ejpam-6087	411	17	of	of	ADP
ejpam-6087	411	18	39	39	NUM
ejpam-6087	411	19	5	5	NUM
ejpam-6087	411	20	10	10	NUM
ejpam-6087	411	21	15	15	NUM
ejpam-6087	411	22	20	20	NUM
ejpam-6087	411	23	25	25	NUM
ejpam-6087	411	24	30	30	NUM
ejpam-6087	411	25	number	number	NOUN
ejpam-6087	411	26	of	of	ADP
ejpam-6087	411	27	features	feature	NOUN
ejpam-6087	411	28	45	45	NUM
ejpam-6087	411	29	50	50	NUM
ejpam-6087	411	30	55	55	NUM
ejpam-6087	411	31	60	60	NUM
ejpam-6087	411	32	65	65	NUM
ejpam-6087	411	33	70	70	NUM
ejpam-6087	411	34	75	75	NUM
ejpam-6087	411	35	80	80	NUM
ejpam-6087	411	36	85	85	NUM
ejpam-6087	411	37	90	90	NUM
ejpam-6087	411	38	95	95	NUM
ejpam-6087	411	39	100	100	NUM
ejpam-6087	411	40	a	a	PRON
ejpam-6087	411	41	cc	cc	ADP
ejpam-6087	411	42	ur	ur	INTJ
ejpam-6087	411	43	ac	ac	PROPN
ejpam-6087	411	44	y	y	PROPN
ejpam-6087	411	45	(	(	PUNCT
ejpam-6087	411	46	%	%	NOUN
ejpam-6087	411	47	)	)	PUNCT
ejpam-6087	411	48	lr	lr	PROPN
ejpam-6087	411	49	nb	nb	PROPN
ejpam-6087	411	50	knn	knn	PROPN
ejpam-6087	411	51	svm	svm	PROPN
ejpam-6087	411	52	mlp	mlp	PROPN
ejpam-6087	411	53	dt	dt	PROPN
ejpam-6087	411	54	rf	rf	ADP
ejpam-6087	411	55	a	a	DET
ejpam-6087	411	56	accuracy	accuracy	NOUN
ejpam-6087	411	57	5	5	NUM
ejpam-6087	411	58	10	10	NUM
ejpam-6087	411	59	15	15	NUM
ejpam-6087	411	60	20	20	NUM
ejpam-6087	411	61	25	25	NUM
ejpam-6087	411	62	30	30	NUM
ejpam-6087	411	63	number	number	NOUN
ejpam-6087	411	64	of	of	ADP
ejpam-6087	411	65	features	feature	NOUN
ejpam-6087	411	66	45	45	NUM
ejpam-6087	411	67	50	50	NUM
ejpam-6087	411	68	55	55	NUM
ejpam-6087	411	69	60	60	NUM
ejpam-6087	411	70	65	65	NUM
ejpam-6087	411	71	70	70	NUM
ejpam-6087	411	72	75	75	NUM
ejpam-6087	411	73	80	80	NUM
ejpam-6087	411	74	85	85	NUM
ejpam-6087	411	75	90	90	NUM
ejpam-6087	411	76	95	95	NUM
ejpam-6087	411	77	100	100	NUM
ejpam-6087	411	78	p	p	X
ejpam-6087	411	79	re	re	ADP
ejpam-6087	411	80	ci	ci	PROPN
ejpam-6087	411	81	si	si	X
ejpam-6087	411	82	on	on	ADP
ejpam-6087	411	83	(	(	PUNCT
ejpam-6087	411	84	%	%	INTJ
ejpam-6087	411	85	)	)	PUNCT
ejpam-6087	411	86	lr	lr	PROPN
ejpam-6087	411	87	nb	nb	PROPN
ejpam-6087	411	88	knn	knn	PROPN
ejpam-6087	411	89	svm	svm	PROPN
ejpam-6087	411	90	mlp	mlp	PROPN
ejpam-6087	411	91	dt	dt	PROPN
ejpam-6087	411	92	rf	rf	PROPN
ejpam-6087	411	93	b	b	PROPN
ejpam-6087	411	94	precision	precision	NOUN
ejpam-6087	411	95	5	5	NUM
ejpam-6087	411	96	10	10	NUM
ejpam-6087	411	97	15	15	NUM
ejpam-6087	411	98	20	20	NUM
ejpam-6087	411	99	25	25	NUM
ejpam-6087	411	100	30	30	NUM
ejpam-6087	411	101	number	number	NOUN
ejpam-6087	411	102	of	of	ADP
ejpam-6087	411	103	features	feature	NOUN
ejpam-6087	411	104	45	45	NUM
ejpam-6087	411	105	50	50	NUM
ejpam-6087	411	106	55	55	NUM
ejpam-6087	411	107	60	60	NUM
ejpam-6087	411	108	65	65	NUM
ejpam-6087	411	109	70	70	NUM
ejpam-6087	411	110	75	75	NUM
ejpam-6087	411	111	80	80	NUM
ejpam-6087	411	112	85	85	NUM
ejpam-6087	411	113	90	90	NUM
ejpam-6087	411	114	95	95	NUM
ejpam-6087	411	115	100	100	NUM
ejpam-6087	411	116	r	r	NOUN
ejpam-6087	411	117	ec	ec	PROPN
ejpam-6087	411	118	al	al	PROPN
ejpam-6087	411	119	l	l	PROPN
ejpam-6087	412	1	(	(	PUNCT
ejpam-6087	412	2	%	%	INTJ
ejpam-6087	412	3	)	)	PUNCT
ejpam-6087	412	4	lr	lr	PROPN
ejpam-6087	412	5	nb	nb	PROPN
ejpam-6087	412	6	knn	knn	PROPN
ejpam-6087	412	7	svm	svm	PROPN
ejpam-6087	412	8	mlp	mlp	PROPN
ejpam-6087	412	9	dt	dt	PROPN
ejpam-6087	412	10	rf	rf	PROPN
ejpam-6087	412	11	c	c	NOUN
ejpam-6087	412	12	recall	recall	VERB
ejpam-6087	412	13	5	5	NUM
ejpam-6087	412	14	10	10	NUM
ejpam-6087	412	15	15	15	NUM
ejpam-6087	412	16	20	20	NUM
ejpam-6087	412	17	25	25	NUM
ejpam-6087	412	18	30	30	NUM
ejpam-6087	412	19	number	number	NOUN
ejpam-6087	412	20	of	of	ADP
ejpam-6087	412	21	features	feature	NOUN
ejpam-6087	412	22	30	30	NUM
ejpam-6087	412	23	40	40	NUM
ejpam-6087	412	24	50	50	NUM
ejpam-6087	412	25	60	60	NUM
ejpam-6087	412	26	70	70	NUM
ejpam-6087	412	27	80	80	NUM
ejpam-6087	412	28	90	90	NUM
ejpam-6087	412	29	100	100	NUM
ejpam-6087	412	30	f	f	NOUN
ejpam-6087	412	31	1	1	NUM
ejpam-6087	412	32	s	s	PART
ejpam-6087	412	33	co	co	X
ejpam-6087	412	34	re	re	X
ejpam-6087	412	35	(	(	PUNCT
ejpam-6087	412	36	%	%	INTJ
ejpam-6087	412	37	)	)	PUNCT
ejpam-6087	412	38	lr	lr	PROPN
ejpam-6087	412	39	nb	nb	PROPN
ejpam-6087	412	40	knn	knn	PROPN
ejpam-6087	412	41	svm	svm	PROPN
ejpam-6087	412	42	mlp	mlp	PROPN
ejpam-6087	412	43	dt	dt	PROPN
ejpam-6087	412	44	rf	rf	PROPN
ejpam-6087	412	45	d	d	PROPN
ejpam-6087	412	46	f1	f1	PROPN
ejpam-6087	412	47	score	score	NOUN
ejpam-6087	412	48	5	5	NUM
ejpam-6087	412	49	10	10	NUM
ejpam-6087	412	50	15	15	NUM
ejpam-6087	412	51	20	20	NUM
ejpam-6087	412	52	25	25	NUM
ejpam-6087	412	53	30	30	NUM
ejpam-6087	412	54	number	number	NOUN
ejpam-6087	412	55	of	of	ADP
ejpam-6087	412	56	features	feature	NOUN
ejpam-6087	412	57	50	50	NUM
ejpam-6087	412	58	55	55	NUM
ejpam-6087	412	59	60	60	NUM
ejpam-6087	412	60	65	65	NUM
ejpam-6087	412	61	70	70	NUM
ejpam-6087	412	62	75	75	NUM
ejpam-6087	412	63	80	80	NUM
ejpam-6087	412	64	85	85	NUM
ejpam-6087	412	65	90	90	NUM
ejpam-6087	412	66	95	95	NUM
ejpam-6087	412	67	100	100	NUM
ejpam-6087	412	68	a	a	DET
ejpam-6087	412	69	u	u	NOUN
ejpam-6087	412	70	c	c	NOUN
ejpam-6087	412	71	-r	-r	NOUN
ejpam-6087	412	72	o	o	NOUN
ejpam-6087	412	73	c	c	PROPN
ejpam-6087	412	74	(	(	PUNCT
ejpam-6087	412	75	%	%	NOUN
ejpam-6087	412	76	)	)	PUNCT
ejpam-6087	412	77	lr	lr	PROPN
ejpam-6087	412	78	nb	nb	PROPN
ejpam-6087	412	79	knn	knn	PROPN
ejpam-6087	412	80	svm	svm	PROPN
ejpam-6087	412	81	mlp	mlp	PROPN
ejpam-6087	412	82	dt	dt	PROPN
ejpam-6087	412	83	rf	rf	PROPN
ejpam-6087	412	84	e	e	PROPN
ejpam-6087	412	85	auc	auc	VERB
ejpam-6087	412	86	-	-	PUNCT
ejpam-6087	412	87	roc	roc	NOUN
ejpam-6087	412	88	figure	figure	NOUN
ejpam-6087	412	89	1	1	NUM
ejpam-6087	412	90	:	:	PUNCT
ejpam-6087	412	91	performance	performance	NOUN
ejpam-6087	412	92	metrics	metric	NOUN
ejpam-6087	412	93	of	of	ADP
ejpam-6087	412	94	various	various	ADJ
ejpam-6087	412	95	models	model	NOUN
ejpam-6087	412	96	as	as	SCONJ
ejpam-6087	412	97	feature	feature	NOUN
ejpam-6087	412	98	number	number	NOUN
ejpam-6087	412	99	increases	increase	VERB
ejpam-6087	412	100	under	under	ADP
ejpam-6087	412	101	recursive	recursive	ADJ
ejpam-6087	412	102	feature	feature	NOUN
ejpam-6087	412	103	elimination	elimination	NOUN
ejpam-6087	412	104	.	.	PUNCT
ejpam-6087	413	1	m.	m.	NOUN
ejpam-6087	413	2	al	al	PROPN
ejpam-6087	413	3	-	-	PUNCT
ejpam-6087	413	4	tarawneh	tarawneh	PROPN
ejpam-6087	413	5	et	et	PROPN
ejpam-6087	413	6	al	al	PROPN
ejpam-6087	413	7	.	.	PUNCT
ejpam-6087	413	8	/	/	SYM
ejpam-6087	413	9	eur	eur	PROPN
ejpam-6087	413	10	.	.	PUNCT
ejpam-6087	414	1	j.	j.	PROPN
ejpam-6087	414	2	pure	pure	PROPN
ejpam-6087	414	3	appl	appl	PROPN
ejpam-6087	414	4	.	.	PROPN
ejpam-6087	414	5	math	math	PROPN
ejpam-6087	414	6	,	,	PUNCT
ejpam-6087	414	7	18	18	NUM
ejpam-6087	414	8	(	(	PUNCT
ejpam-6087	414	9	2	2	NUM
ejpam-6087	414	10	)	)	PUNCT
ejpam-6087	414	11	(	(	PUNCT
ejpam-6087	414	12	2025	2025	NUM
ejpam-6087	414	13	)	)	PUNCT
ejpam-6087	414	14	,	,	PUNCT
ejpam-6087	414	15	6087	6087	NUM
ejpam-6087	414	16	23	23	NUM
ejpam-6087	414	17	of	of	ADP
ejpam-6087	414	18	39	39	NUM
ejpam-6087	414	19	the	the	DET
ejpam-6087	414	20	precision	precision	NOUN
ejpam-6087	414	21	results	result	NOUN
ejpam-6087	414	22	for	for	ADP
ejpam-6087	414	23	different	different	ADJ
ejpam-6087	414	24	models	model	NOUN
ejpam-6087	414	25	using	use	VERB
ejpam-6087	414	26	rfe	rfe	NOUN
ejpam-6087	414	27	show	show	VERB
ejpam-6087	414	28	a	a	DET
ejpam-6087	414	29	varied	varied	ADJ
ejpam-6087	414	30	performance	performance	NOUN
ejpam-6087	414	31	in	in	ADP
ejpam-6087	414	32	terms	term	NOUN
ejpam-6087	414	33	of	of	ADP
ejpam-6087	414	34	number	number	NOUN
ejpam-6087	414	35	of	of	ADP
ejpam-6087	414	36	features	feature	NOUN
ejpam-6087	414	37	.	.	PUNCT
ejpam-6087	415	1	for	for	ADP
ejpam-6087	415	2	lr	lr	PROPN
ejpam-6087	415	3	,	,	PUNCT
ejpam-6087	415	4	the	the	DET
ejpam-6087	415	5	precision	precision	NOUN
ejpam-6087	415	6	starts	start	VERB
ejpam-6087	415	7	at	at	ADP
ejpam-6087	415	8	95.80	95.80	NUM
ejpam-6087	415	9	%	%	NOUN
ejpam-6087	415	10	with	with	ADP
ejpam-6087	415	11	5	5	NUM
ejpam-6087	415	12	features	feature	NOUN
ejpam-6087	415	13	and	and	CCONJ
ejpam-6087	415	14	decreases	decrease	VERB
ejpam-6087	415	15	to	to	ADP
ejpam-6087	415	16	60.34	60.34	NUM
ejpam-6087	415	17	%	%	NOUN
ejpam-6087	415	18	as	as	ADP
ejpam-6087	415	19	the	the	DET
ejpam-6087	415	20	number	number	NOUN
ejpam-6087	415	21	of	of	ADP
ejpam-6087	415	22	features	feature	NOUN
ejpam-6087	415	23	increases	increase	NOUN
ejpam-6087	415	24	to	to	ADP
ejpam-6087	415	25	30	30	NUM
ejpam-6087	415	26	.	.	PUNCT
ejpam-6087	416	1	nb	nb	PROPN
ejpam-6087	416	2	shows	show	VERB
ejpam-6087	416	3	a	a	DET
ejpam-6087	416	4	relatively	relatively	ADV
ejpam-6087	416	5	stable	stable	ADJ
ejpam-6087	416	6	precision	precision	NOUN
ejpam-6087	416	7	,	,	PUNCT
ejpam-6087	416	8	with	with	ADP
ejpam-6087	416	9	58.89	58.89	NUM
ejpam-6087	416	10	%	%	NOUN
ejpam-6087	416	11	at	at	ADP
ejpam-6087	416	12	5	5	NUM
ejpam-6087	416	13	features	feature	NOUN
ejpam-6087	416	14	and	and	CCONJ
ejpam-6087	416	15	57.08	57.08	NUM
ejpam-6087	416	16	%	%	NOUN
ejpam-6087	416	17	at	at	ADP
ejpam-6087	416	18	30	30	NUM
ejpam-6087	416	19	features	feature	NOUN
ejpam-6087	416	20	.	.	PUNCT
ejpam-6087	417	1	knn	knn	PROPN
ejpam-6087	417	2	maintains	maintain	VERB
ejpam-6087	417	3	a	a	DET
ejpam-6087	417	4	precision	precision	NOUN
ejpam-6087	417	5	level	level	NOUN
ejpam-6087	417	6	from	from	ADP
ejpam-6087	417	7	72.83	72.83	NUM
ejpam-6087	417	8	%	%	NOUN
ejpam-6087	417	9	at	at	ADP
ejpam-6087	417	10	5	5	NUM
ejpam-6087	417	11	features	feature	NOUN
ejpam-6087	417	12	to	to	ADP
ejpam-6087	417	13	58.72	58.72	NUM
ejpam-6087	417	14	%	%	NOUN
ejpam-6087	417	15	at	at	ADP
ejpam-6087	417	16	30	30	NUM
ejpam-6087	417	17	features	feature	NOUN
ejpam-6087	417	18	,	,	PUNCT
ejpam-6087	417	19	indicating	indicate	VERB
ejpam-6087	417	20	a	a	DET
ejpam-6087	417	21	gradual	gradual	ADJ
ejpam-6087	417	22	decrease	decrease	NOUN
ejpam-6087	417	23	.	.	PUNCT
ejpam-6087	418	1	the	the	DET
ejpam-6087	418	2	svm	svm	PROPN
ejpam-6087	418	3	has	have	VERB
ejpam-6087	418	4	stable	stable	ADJ
ejpam-6087	418	5	but	but	CCONJ
ejpam-6087	418	6	low	low	ADJ
ejpam-6087	418	7	precision	precision	NOUN
ejpam-6087	418	8	,	,	PUNCT
ejpam-6087	418	9	starting	start	VERB
ejpam-6087	418	10	at	at	ADP
ejpam-6087	418	11	53.93	53.93	NUM
ejpam-6087	418	12	%	%	NOUN
ejpam-6087	418	13	and	and	CCONJ
ejpam-6087	418	14	slightly	slightly	ADV
ejpam-6087	418	15	increasing	increase	VERB
ejpam-6087	418	16	to	to	ADP
ejpam-6087	418	17	54.78	54.78	NUM
ejpam-6087	418	18	%	%	NOUN
ejpam-6087	418	19	as	as	SCONJ
ejpam-6087	418	20	more	more	ADJ
ejpam-6087	418	21	features	feature	NOUN
ejpam-6087	418	22	are	be	AUX
ejpam-6087	418	23	used	use	VERB
ejpam-6087	418	24	.	.	PUNCT
ejpam-6087	419	1	mlp	mlp	NOUN
ejpam-6087	419	2	begins	begin	VERB
ejpam-6087	419	3	with	with	ADP
ejpam-6087	419	4	a	a	DET
ejpam-6087	419	5	high	high	ADJ
ejpam-6087	419	6	precision	precision	NOUN
ejpam-6087	419	7	of	of	ADP
ejpam-6087	419	8	95.73	95.73	NUM
ejpam-6087	419	9	%	%	NOUN
ejpam-6087	419	10	at	at	ADP
ejpam-6087	419	11	5	5	NUM
ejpam-6087	419	12	features	feature	NOUN
ejpam-6087	419	13	,	,	PUNCT
ejpam-6087	419	14	which	which	PRON
ejpam-6087	419	15	drops	drop	VERB
ejpam-6087	419	16	to	to	ADP
ejpam-6087	419	17	81.86	81.86	NUM
ejpam-6087	419	18	%	%	NOUN
ejpam-6087	419	19	at	at	ADP
ejpam-6087	419	20	10	10	NUM
ejpam-6087	419	21	features	feature	NOUN
ejpam-6087	419	22	,	,	PUNCT
ejpam-6087	419	23	and	and	CCONJ
ejpam-6087	419	24	then	then	ADV
ejpam-6087	419	25	fluctuates	fluctuate	VERB
ejpam-6087	419	26	before	before	ADP
ejpam-6087	419	27	increasing	increase	VERB
ejpam-6087	419	28	to	to	ADP
ejpam-6087	419	29	88.14	88.14	NUM
ejpam-6087	419	30	%	%	NOUN
ejpam-6087	419	31	at	at	ADP
ejpam-6087	419	32	30	30	NUM
ejpam-6087	419	33	features	feature	NOUN
ejpam-6087	419	34	.	.	PUNCT
ejpam-6087	420	1	dt	dt	PROPN
ejpam-6087	420	2	shows	show	VERB
ejpam-6087	420	3	high	high	ADJ
ejpam-6087	420	4	and	and	CCONJ
ejpam-6087	420	5	consistent	consistent	ADJ
ejpam-6087	420	6	precision	precision	NOUN
ejpam-6087	420	7	throughout	throughout	ADP
ejpam-6087	420	8	,	,	PUNCT
ejpam-6087	420	9	with	with	ADP
ejpam-6087	420	10	values	value	NOUN
ejpam-6087	420	11	from	from	ADP
ejpam-6087	420	12	91.37	91.37	NUM
ejpam-6087	420	13	%	%	NOUN
ejpam-6087	420	14	to	to	ADP
ejpam-6087	420	15	90.80	90.80	NUM
ejpam-6087	420	16	%	%	NOUN
ejpam-6087	420	17	throughout	throughout	ADP
ejpam-6087	420	18	the	the	DET
ejpam-6087	420	19	range	range	NOUN
ejpam-6087	420	20	of	of	ADP
ejpam-6087	420	21	features	feature	NOUN
ejpam-6087	420	22	.	.	PUNCT
ejpam-6087	421	1	rf	rf	NOUN
ejpam-6087	421	2	exhibits	exhibit	VERB
ejpam-6087	421	3	high	high	ADJ
ejpam-6087	421	4	and	and	CCONJ
ejpam-6087	421	5	steady	steady	ADJ
ejpam-6087	421	6	precision	precision	NOUN
ejpam-6087	421	7	across	across	ADP
ejpam-6087	421	8	all	all	DET
ejpam-6087	421	9	set	set	NOUN
ejpam-6087	421	10	of	of	ADP
ejpam-6087	421	11	features	feature	NOUN
ejpam-6087	421	12	,	,	PUNCT
ejpam-6087	421	13	starting	start	VERB
ejpam-6087	421	14	at	at	ADP
ejpam-6087	421	15	95.70	95.70	NUM
ejpam-6087	421	16	%	%	NOUN
ejpam-6087	421	17	and	and	CCONJ
ejpam-6087	421	18	remaining	remain	VERB
ejpam-6087	421	19	at	at	ADP
ejpam-6087	421	20	95.80	95.80	NUM
ejpam-6087	421	21	%	%	NOUN
ejpam-6087	421	22	.	.	PUNCT
ejpam-6087	422	1	the	the	DET
ejpam-6087	422	2	recall	recall	NOUN
ejpam-6087	422	3	results	result	NOUN
ejpam-6087	422	4	for	for	ADP
ejpam-6087	422	5	different	different	ADJ
ejpam-6087	422	6	models	model	NOUN
ejpam-6087	422	7	using	use	VERB
ejpam-6087	422	8	rfe	rfe	NOUN
ejpam-6087	422	9	display	display	NOUN
ejpam-6087	422	10	diverse	diverse	ADJ
ejpam-6087	422	11	trends	trend	NOUN
ejpam-6087	422	12	as	as	ADP
ejpam-6087	422	13	the	the	DET
ejpam-6087	422	14	number	number	NOUN
ejpam-6087	422	15	of	of	ADP
ejpam-6087	422	16	features	feature	NOUN
ejpam-6087	422	17	increases	increase	NOUN
ejpam-6087	422	18	as	as	SCONJ
ejpam-6087	422	19	shown	show	VERB
ejpam-6087	422	20	in	in	ADP
ejpam-6087	422	21	figure	figure	NOUN
ejpam-6087	422	22	1c	1c	NOUN
ejpam-6087	422	23	.	.	PUNCT
ejpam-6087	423	1	lr	lr	PROPN
ejpam-6087	423	2	achieves	achieve	VERB
ejpam-6087	423	3	a	a	DET
ejpam-6087	423	4	high	high	ADJ
ejpam-6087	423	5	recall	recall	NOUN
ejpam-6087	423	6	of	of	ADP
ejpam-6087	423	7	95.80	95.80	NUM
ejpam-6087	423	8	%	%	NOUN
ejpam-6087	423	9	in	in	ADP
ejpam-6087	423	10	both	both	DET
ejpam-6087	423	11	5	5	NUM
ejpam-6087	423	12	and	and	CCONJ
ejpam-6087	423	13	10	10	NUM
ejpam-6087	423	14	features	feature	NOUN
ejpam-6087	423	15	,	,	PUNCT
ejpam-6087	423	16	but	but	CCONJ
ejpam-6087	423	17	experiences	experience	VERB
ejpam-6087	423	18	a	a	DET
ejpam-6087	423	19	significant	significant	ADJ
ejpam-6087	423	20	decline	decline	NOUN
ejpam-6087	423	21	to	to	ADP
ejpam-6087	423	22	60	60	NUM
ejpam-6087	423	23	.	.	PUNCT
ejpam-6087	424	1	19	19	NUM
ejpam-6087	424	2	%	%	NOUN
ejpam-6087	424	3	in	in	ADP
ejpam-6087	424	4	30	30	NUM
ejpam-6087	424	5	features	feature	NOUN
ejpam-6087	424	6	.	.	PUNCT
ejpam-6087	425	1	nb	nb	PROPN
ejpam-6087	425	2	shows	show	VERB
ejpam-6087	425	3	a	a	DET
ejpam-6087	425	4	consistently	consistently	ADV
ejpam-6087	425	5	low	low	ADJ
ejpam-6087	425	6	recall	recall	NOUN
ejpam-6087	425	7	,	,	PUNCT
ejpam-6087	425	8	with	with	ADP
ejpam-6087	425	9	values	value	NOUN
ejpam-6087	425	10	of	of	ADP
ejpam-6087	425	11	52.28	52.28	NUM
ejpam-6087	425	12	%	%	NOUN
ejpam-6087	425	13	at	at	ADP
ejpam-6087	425	14	five	five	NUM
ejpam-6087	425	15	features	feature	NOUN
ejpam-6087	425	16	and	and	CCONJ
ejpam-6087	425	17	remaining	remain	VERB
ejpam-6087	425	18	around	around	ADV
ejpam-6087	425	19	49.88	49.88	NUM
ejpam-6087	425	20	%	%	NOUN
ejpam-6087	425	21	at	at	ADP
ejpam-6087	425	22	30	30	NUM
ejpam-6087	425	23	characteristics	characteristic	NOUN
ejpam-6087	425	24	.	.	PUNCT
ejpam-6087	426	1	knn	knn	PROPN
ejpam-6087	426	2	begins	begin	VERB
ejpam-6087	426	3	with	with	ADP
ejpam-6087	426	4	a	a	DET
ejpam-6087	426	5	high	high	ADJ
ejpam-6087	426	6	recall	recall	NOUN
ejpam-6087	426	7	of	of	ADP
ejpam-6087	426	8	72.84	72.84	NUM
ejpam-6087	426	9	%	%	NOUN
ejpam-6087	426	10	for	for	ADP
ejpam-6087	426	11	five	five	NUM
ejpam-6087	426	12	features	feature	NOUN
ejpam-6087	426	13	and	and	CCONJ
ejpam-6087	426	14	stabilizes	stabilize	VERB
ejpam-6087	426	15	at	at	ADP
ejpam-6087	426	16	58.73	58.73	NUM
ejpam-6087	426	17	%	%	NOUN
ejpam-6087	426	18	for	for	ADP
ejpam-6087	426	19	feature	feature	NOUN
ejpam-6087	426	20	counts	count	NOUN
ejpam-6087	426	21	of	of	ADP
ejpam-6087	426	22	10	10	NUM
ejpam-6087	426	23	and	and	CCONJ
ejpam-6087	426	24	above	above	ADV
ejpam-6087	426	25	.	.	PUNCT
ejpam-6087	427	1	svm	svm	PROPN
ejpam-6087	427	2	maintains	maintain	VERB
ejpam-6087	427	3	a	a	DET
ejpam-6087	427	4	constant	constant	ADJ
ejpam-6087	427	5	recall	recall	NOUN
ejpam-6087	427	6	of	of	ADP
ejpam-6087	427	7	53.49	53.49	NUM
ejpam-6087	427	8	%	%	NOUN
ejpam-6087	427	9	across	across	ADP
ejpam-6087	427	10	all	all	DET
ejpam-6087	427	11	feature	feature	NOUN
ejpam-6087	427	12	sets	set	NOUN
ejpam-6087	427	13	,	,	PUNCT
ejpam-6087	427	14	indicating	indicate	VERB
ejpam-6087	427	15	that	that	SCONJ
ejpam-6087	427	16	it	it	PRON
ejpam-6087	427	17	is	be	AUX
ejpam-6087	427	18	relatively	relatively	ADV
ejpam-6087	427	19	unaffected	unaffected	ADJ
ejpam-6087	427	20	by	by	ADP
ejpam-6087	427	21	the	the	DET
ejpam-6087	427	22	number	number	NOUN
ejpam-6087	427	23	of	of	ADP
ejpam-6087	427	24	features	feature	NOUN
ejpam-6087	427	25	.	.	PUNCT
ejpam-6087	428	1	mlp	mlp	NOUN
ejpam-6087	428	2	starts	start	VERB
ejpam-6087	428	3	with	with	ADP
ejpam-6087	428	4	a	a	DET
ejpam-6087	428	5	high	high	ADJ
ejpam-6087	428	6	recall	recall	NOUN
ejpam-6087	428	7	of	of	ADP
ejpam-6087	428	8	95.73	95.73	NUM
ejpam-6087	428	9	%	%	NOUN
ejpam-6087	428	10	at	at	ADP
ejpam-6087	428	11	5	5	NUM
ejpam-6087	428	12	features	feature	NOUN
ejpam-6087	428	13	,	,	PUNCT
ejpam-6087	428	14	decreases	decrease	VERB
ejpam-6087	428	15	to	to	ADP
ejpam-6087	428	16	79.09	79.09	NUM
ejpam-6087	428	17	%	%	NOUN
ejpam-6087	428	18	at	at	ADP
ejpam-6087	428	19	10	10	NUM
ejpam-6087	428	20	features	feature	NOUN
ejpam-6087	428	21	,	,	PUNCT
ejpam-6087	428	22	and	and	CCONJ
ejpam-6087	428	23	then	then	ADV
ejpam-6087	428	24	fluctuates	fluctuate	VERB
ejpam-6087	428	25	before	before	ADP
ejpam-6087	428	26	rising	rise	VERB
ejpam-6087	428	27	to	to	ADP
ejpam-6087	428	28	86.53	86.53	NUM
ejpam-6087	428	29	%	%	NOUN
ejpam-6087	428	30	at	at	ADP
ejpam-6087	428	31	30	30	NUM
ejpam-6087	428	32	features	feature	NOUN
ejpam-6087	428	33	,	,	PUNCT
ejpam-6087	428	34	showing	show	VERB
ejpam-6087	428	35	an	an	DET
ejpam-6087	428	36	overall	overall	ADJ
ejpam-6087	428	37	trend	trend	NOUN
ejpam-6087	428	38	of	of	ADP
ejpam-6087	428	39	improvement	improvement	NOUN
ejpam-6087	428	40	with	with	ADP
ejpam-6087	428	41	more	more	ADJ
ejpam-6087	428	42	features	feature	NOUN
ejpam-6087	428	43	.	.	PUNCT
ejpam-6087	429	1	dt	dt	PROPN
ejpam-6087	429	2	shows	show	VERB
ejpam-6087	429	3	a	a	DET
ejpam-6087	429	4	high	high	ADJ
ejpam-6087	429	5	and	and	CCONJ
ejpam-6087	429	6	stable	stable	ADJ
ejpam-6087	429	7	recall	recall	NOUN
ejpam-6087	429	8	from	from	ADP
ejpam-6087	429	9	91.36	91.36	NUM
ejpam-6087	429	10	%	%	NOUN
ejpam-6087	429	11	at	at	ADP
ejpam-6087	429	12	5	5	NUM
ejpam-6087	429	13	features	feature	NOUN
ejpam-6087	429	14	to	to	ADP
ejpam-6087	429	15	90.80	90.80	NUM
ejpam-6087	429	16	%	%	NOUN
ejpam-6087	429	17	at	at	ADP
ejpam-6087	429	18	30	30	NUM
ejpam-6087	429	19	features	feature	NOUN
ejpam-6087	429	20	,	,	PUNCT
ejpam-6087	429	21	demonstrating	demonstrate	VERB
ejpam-6087	429	22	robust	robust	ADJ
ejpam-6087	429	23	performance	performance	NOUN
ejpam-6087	429	24	regardless	regardless	ADV
ejpam-6087	429	25	of	of	ADP
ejpam-6087	429	26	the	the	DET
ejpam-6087	429	27	number	number	NOUN
ejpam-6087	429	28	of	of	ADP
ejpam-6087	429	29	features	feature	NOUN
ejpam-6087	429	30	.	.	PUNCT
ejpam-6087	430	1	lastly	lastly	ADV
ejpam-6087	430	2	,	,	PUNCT
ejpam-6087	430	3	rf	rf	NOUN
ejpam-6087	430	4	also	also	ADV
ejpam-6087	430	5	shows	show	VERB
ejpam-6087	430	6	high	high	ADJ
ejpam-6087	430	7	and	and	CCONJ
ejpam-6087	430	8	consistent	consistent	ADJ
ejpam-6087	430	9	recall	recall	NOUN
ejpam-6087	430	10	from	from	ADP
ejpam-6087	430	11	95.70	95.70	NUM
ejpam-6087	430	12	%	%	NOUN
ejpam-6087	430	13	with	with	ADP
ejpam-6087	430	14	5	5	NUM
ejpam-6087	430	15	features	feature	NOUN
ejpam-6087	430	16	to	to	ADP
ejpam-6087	430	17	95.80	95.80	NUM
ejpam-6087	430	18	%	%	NOUN
ejpam-6087	430	19	with	with	ADP
ejpam-6087	430	20	30	30	NUM
ejpam-6087	430	21	features	feature	NOUN
ejpam-6087	430	22	,	,	PUNCT
ejpam-6087	430	23	representing	represent	VERB
ejpam-6087	430	24	the	the	DET
ejpam-6087	430	25	most	most	ADV
ejpam-6087	430	26	reliable	reliable	ADJ
ejpam-6087	430	27	and	and	CCONJ
ejpam-6087	430	28	effective	effective	ADJ
ejpam-6087	430	29	model	model	NOUN
ejpam-6087	430	30	in	in	ADP
ejpam-6087	430	31	terms	term	NOUN
ejpam-6087	430	32	of	of	ADP
ejpam-6087	430	33	recall	recall	NOUN
ejpam-6087	430	34	.	.	PUNCT
ejpam-6087	431	1	the	the	DET
ejpam-6087	431	2	results	result	NOUN
ejpam-6087	431	3	of	of	ADP
ejpam-6087	431	4	the	the	DET
ejpam-6087	431	5	f1	f1	ADJ
ejpam-6087	431	6	score	score	NOUN
ejpam-6087	431	7	for	for	ADP
ejpam-6087	431	8	various	various	ADJ
ejpam-6087	431	9	models	model	NOUN
ejpam-6087	431	10	using	use	VERB
ejpam-6087	431	11	rfe	rfe	NOUN
ejpam-6087	431	12	reveal	reveal	VERB
ejpam-6087	431	13	different	different	ADJ
ejpam-6087	431	14	trends	trend	NOUN
ejpam-6087	431	15	as	as	ADP
ejpam-6087	431	16	the	the	DET
ejpam-6087	431	17	number	number	NOUN
ejpam-6087	431	18	of	of	ADP
ejpam-6087	431	19	features	feature	NOUN
ejpam-6087	431	20	increases	increase	NOUN
ejpam-6087	431	21	as	as	SCONJ
ejpam-6087	431	22	depicted	depict	VERB
ejpam-6087	431	23	in	in	ADP
ejpam-6087	431	24	figure	figure	NOUN
ejpam-6087	431	25	1d	1d	NUM
ejpam-6087	431	26	.	.	PUNCT
ejpam-6087	432	1	lr	lr	PROPN
ejpam-6087	432	2	maintains	maintain	VERB
ejpam-6087	432	3	a	a	DET
ejpam-6087	432	4	high	high	ADJ
ejpam-6087	432	5	f1	f1	NOUN
ejpam-6087	432	6	score	score	NOUN
ejpam-6087	432	7	of	of	ADP
ejpam-6087	432	8	95.80	95.80	NUM
ejpam-6087	432	9	%	%	NOUN
ejpam-6087	432	10	for	for	ADP
ejpam-6087	432	11	5	5	NUM
ejpam-6087	432	12	and	and	CCONJ
ejpam-6087	432	13	10	10	NUM
ejpam-6087	432	14	features	feature	NOUN
ejpam-6087	432	15	,	,	PUNCT
ejpam-6087	432	16	but	but	CCONJ
ejpam-6087	432	17	shows	show	VERB
ejpam-6087	432	18	a	a	DET
ejpam-6087	432	19	notable	notable	ADJ
ejpam-6087	432	20	decrease	decrease	NOUN
ejpam-6087	432	21	to	to	ADP
ejpam-6087	432	22	59.75	59.75	NUM
ejpam-6087	432	23	%	%	NOUN
ejpam-6087	432	24	for	for	ADP
ejpam-6087	432	25	30	30	NUM
ejpam-6087	432	26	features	feature	NOUN
ejpam-6087	432	27	,	,	PUNCT
ejpam-6087	432	28	indicating	indicate	VERB
ejpam-6087	432	29	a	a	DET
ejpam-6087	432	30	decrease	decrease	NOUN
ejpam-6087	432	31	in	in	ADP
ejpam-6087	432	32	performance	performance	NOUN
ejpam-6087	432	33	as	as	SCONJ
ejpam-6087	432	34	more	more	ADJ
ejpam-6087	432	35	features	feature	NOUN
ejpam-6087	432	36	are	be	AUX
ejpam-6087	432	37	added	add	VERB
ejpam-6087	432	38	.	.	PUNCT
ejpam-6087	433	1	nb	nb	PROPN
ejpam-6087	433	2	exhibits	exhibit	VERB
ejpam-6087	433	3	low	low	ADJ
ejpam-6087	433	4	f1	f1	PROPN
ejpam-6087	433	5	scores	score	NOUN
ejpam-6087	433	6	throughout	throughout	ADV
ejpam-6087	433	7	,	,	PUNCT
ejpam-6087	433	8	starting	start	VERB
ejpam-6087	433	9	at	at	ADP
ejpam-6087	433	10	43.73	43.73	NUM
ejpam-6087	433	11	%	%	NOUN
ejpam-6087	433	12	with	with	ADP
ejpam-6087	433	13	5	5	NUM
ejpam-6087	433	14	features	feature	NOUN
ejpam-6087	433	15	and	and	CCONJ
ejpam-6087	433	16	stabilizing	stabilize	VERB
ejpam-6087	433	17	around	around	ADV
ejpam-6087	433	18	36.80	36.80	NUM
ejpam-6087	433	19	%	%	NOUN
ejpam-6087	433	20	from	from	ADP
ejpam-6087	433	21	10	10	NUM
ejpam-6087	433	22	to	to	PART
ejpam-6087	433	23	30	30	NUM
ejpam-6087	433	24	features	feature	NOUN
ejpam-6087	433	25	,	,	PUNCT
ejpam-6087	433	26	reflecting	reflect	VERB
ejpam-6087	433	27	poor	poor	ADJ
ejpam-6087	433	28	performance	performance	NOUN
ejpam-6087	433	29	with	with	ADP
ejpam-6087	433	30	increasing	increase	VERB
ejpam-6087	433	31	features	feature	NOUN
ejpam-6087	433	32	.	.	PUNCT
ejpam-6087	434	1	knn	knn	PROPN
ejpam-6087	434	2	achieves	achieve	VERB
ejpam-6087	434	3	a	a	DET
ejpam-6087	434	4	high	high	ADJ
ejpam-6087	434	5	f1	f1	NOUN
ejpam-6087	434	6	score	score	NOUN
ejpam-6087	434	7	of	of	ADP
ejpam-6087	434	8	72.82	72.82	NUM
ejpam-6087	434	9	%	%	NOUN
ejpam-6087	434	10	for	for	ADP
ejpam-6087	434	11	five	five	NUM
ejpam-6087	434	12	features	feature	NOUN
ejpam-6087	434	13	,	,	PUNCT
ejpam-6087	434	14	but	but	CCONJ
ejpam-6087	434	15	then	then	ADV
ejpam-6087	434	16	shows	show	VERB
ejpam-6087	434	17	a	a	DET
ejpam-6087	434	18	decrease	decrease	NOUN
ejpam-6087	434	19	to	to	ADP
ejpam-6087	434	20	about	about	ADP
ejpam-6087	434	21	58.73	58.73	NUM
ejpam-6087	434	22	%	%	NOUN
ejpam-6087	434	23	for	for	ADP
ejpam-6087	434	24	feature	feature	NOUN
ejpam-6087	434	25	counts	count	NOUN
ejpam-6087	434	26	of	of	ADP
ejpam-6087	434	27	10	10	NUM
ejpam-6087	434	28	and	and	CCONJ
ejpam-6087	434	29	above	above	ADV
ejpam-6087	434	30	.	.	PUNCT
ejpam-6087	435	1	svm	svm	PROPN
ejpam-6087	435	2	demonstrates	demonstrate	VERB
ejpam-6087	435	3	a	a	DET
ejpam-6087	435	4	relatively	relatively	ADV
ejpam-6087	435	5	consistent	consistent	ADJ
ejpam-6087	435	6	f1	f1	NOUN
ejpam-6087	435	7	score	score	NOUN
ejpam-6087	435	8	of	of	ADP
ejpam-6087	435	9	51.72	51.72	NUM
ejpam-6087	435	10	%	%	NOUN
ejpam-6087	435	11	in	in	ADP
ejpam-6087	435	12	most	most	ADJ
ejpam-6087	435	13	feature	feature	NOUN
ejpam-6087	435	14	sets	set	NOUN
ejpam-6087	435	15	,	,	PUNCT
ejpam-6087	435	16	with	with	ADP
ejpam-6087	435	17	a	a	DET
ejpam-6087	435	18	slight	slight	ADJ
ejpam-6087	435	19	peak	peak	NOUN
ejpam-6087	435	20	at	at	ADP
ejpam-6087	435	21	53.94	53.94	NUM
ejpam-6087	435	22	%	%	NOUN
ejpam-6087	435	23	with	with	ADP
ejpam-6087	435	24	5	5	NUM
ejpam-6087	435	25	features	feature	NOUN
ejpam-6087	435	26	.	.	PUNCT
ejpam-6087	436	1	mlp	mlp	NOUN
ejpam-6087	436	2	starts	start	VERB
ejpam-6087	436	3	with	with	ADP
ejpam-6087	436	4	a	a	DET
ejpam-6087	436	5	high	high	ADJ
ejpam-6087	436	6	f1	f1	NOUN
ejpam-6087	436	7	score	score	NOUN
ejpam-6087	436	8	of	of	ADP
ejpam-6087	436	9	95.73	95.73	NUM
ejpam-6087	436	10	%	%	NOUN
ejpam-6087	436	11	at	at	ADP
ejpam-6087	436	12	5	5	NUM
ejpam-6087	436	13	features	feature	NOUN
ejpam-6087	436	14	,	,	PUNCT
ejpam-6087	436	15	drops	drop	VERB
ejpam-6087	436	16	to	to	ADP
ejpam-6087	436	17	78.74	78.74	NUM
ejpam-6087	436	18	%	%	NOUN
ejpam-6087	436	19	at	at	ADP
ejpam-6087	436	20	10	10	NUM
ejpam-6087	436	21	features	feature	NOUN
ejpam-6087	436	22	,	,	PUNCT
ejpam-6087	436	23	and	and	CCONJ
ejpam-6087	436	24	fluctuates	fluctuate	NOUN
ejpam-6087	436	25	before	before	ADP
ejpam-6087	436	26	reaching	reach	VERB
ejpam-6087	436	27	86.43	86.43	NUM
ejpam-6087	436	28	%	%	NOUN
ejpam-6087	436	29	at	at	ADP
ejpam-6087	436	30	30	30	NUM
ejpam-6087	436	31	features	feature	NOUN
ejpam-6087	436	32	,	,	PUNCT
ejpam-6087	436	33	suggesting	suggest	VERB
ejpam-6087	436	34	improved	improved	ADJ
ejpam-6087	436	35	performance	performance	NOUN
ejpam-6087	436	36	with	with	ADP
ejpam-6087	436	37	more	more	ADJ
ejpam-6087	436	38	features	feature	NOUN
ejpam-6087	436	39	.	.	PUNCT
ejpam-6087	437	1	dt	dt	PROPN
ejpam-6087	437	2	shows	show	VERB
ejpam-6087	437	3	high	high	ADJ
ejpam-6087	437	4	and	and	CCONJ
ejpam-6087	437	5	stable	stable	ADJ
ejpam-6087	437	6	f1	f1	NOUN
ejpam-6087	437	7	scores	score	NOUN
ejpam-6087	437	8	from	from	ADP
ejpam-6087	437	9	91.36	91.36	NUM
ejpam-6087	437	10	%	%	NOUN
ejpam-6087	437	11	at	at	ADP
ejpam-6087	437	12	5	5	NUM
ejpam-6087	437	13	features	feature	NOUN
ejpam-6087	437	14	to	to	ADP
ejpam-6087	437	15	90.80	90.80	NUM
ejpam-6087	437	16	%	%	NOUN
ejpam-6087	437	17	at	at	ADP
ejpam-6087	437	18	30	30	NUM
ejpam-6087	437	19	features	feature	NOUN
ejpam-6087	437	20	.	.	PUNCT
ejpam-6087	438	1	lastly	lastly	ADV
ejpam-6087	438	2	,	,	PUNCT
ejpam-6087	438	3	rf	rf	NOUN
ejpam-6087	438	4	maintains	maintain	VERB
ejpam-6087	438	5	high	high	ADJ
ejpam-6087	438	6	and	and	CCONJ
ejpam-6087	438	7	consistent	consistent	ADJ
ejpam-6087	438	8	f1	f1	NOUN
ejpam-6087	438	9	scores	score	NOUN
ejpam-6087	438	10	from	from	ADP
ejpam-6087	438	11	95.70	95.70	NUM
ejpam-6087	438	12	%	%	NOUN
ejpam-6087	438	13	at	at	ADP
ejpam-6087	438	14	5	5	NUM
ejpam-6087	438	15	features	feature	NOUN
ejpam-6087	438	16	to	to	ADP
ejpam-6087	438	17	95.80	95.80	NUM
ejpam-6087	438	18	%	%	NOUN
ejpam-6087	438	19	at	at	ADP
ejpam-6087	438	20	30	30	NUM
ejpam-6087	438	21	features	feature	NOUN
ejpam-6087	438	22	,	,	PUNCT
ejpam-6087	438	23	reflecting	reflect	VERB
ejpam-6087	438	24	the	the	DET
ejpam-6087	438	25	most	most	ADV
ejpam-6087	438	26	effective	effective	ADJ
ejpam-6087	438	27	and	and	CCONJ
ejpam-6087	438	28	reliable	reliable	ADJ
ejpam-6087	438	29	performance	performance	NOUN
ejpam-6087	438	30	in	in	ADP
ejpam-6087	438	31	terms	term	NOUN
ejpam-6087	438	32	of	of	ADP
ejpam-6087	438	33	the	the	DET
ejpam-6087	438	34	f1	f1	NOUN
ejpam-6087	438	35	score	score	NOUN
ejpam-6087	438	36	.	.	PUNCT
ejpam-6087	439	1	the	the	DET
ejpam-6087	439	2	auc	auc	PROPN
ejpam-6087	439	3	-	-	PUNCT
ejpam-6087	439	4	roc	roc	NOUN
ejpam-6087	439	5	results	result	NOUN
ejpam-6087	439	6	for	for	ADP
ejpam-6087	439	7	various	various	ADJ
ejpam-6087	439	8	models	model	NOUN
ejpam-6087	439	9	using	use	VERB
ejpam-6087	439	10	rfe	rfe	NOUN
ejpam-6087	439	11	illustrate	illustrate	VERB
ejpam-6087	439	12	different	different	ADJ
ejpam-6087	439	13	performance	performance	NOUN
ejpam-6087	439	14	trends	trend	NOUN
ejpam-6087	439	15	as	as	ADP
ejpam-6087	439	16	the	the	DET
ejpam-6087	439	17	number	number	NOUN
ejpam-6087	439	18	of	of	ADP
ejpam-6087	439	19	features	feature	NOUN
ejpam-6087	439	20	increases	increase	NOUN
ejpam-6087	439	21	,	,	PUNCT
ejpam-6087	439	22	as	as	SCONJ
ejpam-6087	439	23	illustrated	illustrate	VERB
ejpam-6087	439	24	in	in	ADP
ejpam-6087	439	25	figure	figure	NOUN
ejpam-6087	439	26	1e	1e	NOUN
ejpam-6087	439	27	.	.	PUNCT
ejpam-6087	440	1	lr	lr	NOUN
ejpam-6087	440	2	and	and	CCONJ
ejpam-6087	440	3	mlp	mlp	PROPN
ejpam-6087	440	4	both	both	PRON
ejpam-6087	440	5	achieve	achieve	VERB
ejpam-6087	440	6	high	high	ADJ
ejpam-6087	440	7	auc	auc	NOUN
ejpam-6087	440	8	-	-	PUNCT
ejpam-6087	440	9	roc	roc	NOUN
ejpam-6087	440	10	scores	score	NOUN
ejpam-6087	440	11	of	of	ADP
ejpam-6087	440	12	95.75	95.75	NUM
ejpam-6087	440	13	%	%	NOUN
ejpam-6087	440	14	at	at	ADP
ejpam-6087	440	15	5	5	NUM
ejpam-6087	440	16	features	feature	NOUN
ejpam-6087	440	17	.	.	PUNCT
ejpam-6087	441	1	however	however	ADV
ejpam-6087	441	2	,	,	PUNCT
ejpam-6087	441	3	while	while	SCONJ
ejpam-6087	441	4	lr	lr	NOUN
ejpam-6087	441	5	exhibits	exhibit	NOUN
ejpam-6087	441	6	m.	m.	PROPN
ejpam-6087	441	7	al	al	PROPN
ejpam-6087	441	8	-	-	PUNCT
ejpam-6087	441	9	tarawneh	tarawneh	PROPN
ejpam-6087	441	10	et	et	PROPN
ejpam-6087	441	11	al	al	PROPN
ejpam-6087	441	12	.	.	PUNCT
ejpam-6087	441	13	/	/	SYM
ejpam-6087	441	14	eur	eur	PROPN
ejpam-6087	441	15	.	.	PUNCT
ejpam-6087	442	1	j.	j.	PROPN
ejpam-6087	442	2	pure	pure	PROPN
ejpam-6087	442	3	appl	appl	PROPN
ejpam-6087	442	4	.	.	PROPN
ejpam-6087	442	5	math	math	PROPN
ejpam-6087	442	6	,	,	PUNCT
ejpam-6087	442	7	18	18	NUM
ejpam-6087	442	8	(	(	PUNCT
ejpam-6087	442	9	2	2	NUM
ejpam-6087	442	10	)	)	PUNCT
ejpam-6087	442	11	(	(	PUNCT
ejpam-6087	442	12	2025	2025	NUM
ejpam-6087	442	13	)	)	PUNCT
ejpam-6087	442	14	,	,	PUNCT
ejpam-6087	442	15	6087	6087	NUM
ejpam-6087	442	16	24	24	NUM
ejpam-6087	442	17	of	of	ADP
ejpam-6087	442	18	39	39	NUM
ejpam-6087	442	19	a	a	DET
ejpam-6087	442	20	decrease	decrease	NOUN
ejpam-6087	442	21	in	in	ADP
ejpam-6087	442	22	auc	auc	NOUN
ejpam-6087	442	23	-	-	PUNCT
ejpam-6087	442	24	roc	roc	NOUN
ejpam-6087	442	25	to	to	ADP
ejpam-6087	442	26	64.44	64.44	NUM
ejpam-6087	442	27	%	%	NOUN
ejpam-6087	442	28	by	by	ADP
ejpam-6087	442	29	30	30	NUM
ejpam-6087	442	30	features	feature	NOUN
ejpam-6087	443	1	,	,	PUNCT
ejpam-6087	443	2	mlp	mlp	PROPN
ejpam-6087	443	3	maintains	maintain	VERB
ejpam-6087	443	4	a	a	DET
ejpam-6087	443	5	high	high	ADJ
ejpam-6087	443	6	auc	auc	NOUN
ejpam-6087	443	7	-	-	PUNCT
ejpam-6087	443	8	roc	roc	NOUN
ejpam-6087	443	9	with	with	ADP
ejpam-6087	443	10	scores	score	NOUN
ejpam-6087	443	11	of	of	ADP
ejpam-6087	443	12	89.78	89.78	NUM
ejpam-6087	443	13	%	%	NOUN
ejpam-6087	443	14	at	at	ADP
ejpam-6087	443	15	10	10	NUM
ejpam-6087	443	16	features	feature	NOUN
ejpam-6087	443	17	and	and	CCONJ
ejpam-6087	443	18	89.78	89.78	NUM
ejpam-6087	443	19	%	%	NOUN
ejpam-6087	443	20	at	at	ADP
ejpam-6087	443	21	30	30	NUM
ejpam-6087	443	22	features	feature	NOUN
ejpam-6087	443	23	.	.	PUNCT
ejpam-6087	444	1	nb	nb	PROPN
ejpam-6087	444	2	shows	show	VERB
ejpam-6087	444	3	a	a	DET
ejpam-6087	444	4	significant	significant	ADJ
ejpam-6087	444	5	drop	drop	NOUN
ejpam-6087	444	6	in	in	ADP
ejpam-6087	444	7	auc	auc	NOUN
ejpam-6087	444	8	-	-	PUNCT
ejpam-6087	444	9	roc	roc	NOUN
ejpam-6087	444	10	from	from	ADP
ejpam-6087	444	11	87.87	87.87	NUM
ejpam-6087	444	12	%	%	NOUN
ejpam-6087	444	13	at	at	ADP
ejpam-6087	444	14	5	5	NUM
ejpam-6087	444	15	features	feature	NOUN
ejpam-6087	444	16	to	to	ADP
ejpam-6087	444	17	58.24	58.24	NUM
ejpam-6087	444	18	%	%	NOUN
ejpam-6087	444	19	at	at	ADP
ejpam-6087	444	20	30	30	NUM
ejpam-6087	444	21	features	feature	NOUN
ejpam-6087	444	22	.	.	PUNCT
ejpam-6087	445	1	knn	knn	PROPN
ejpam-6087	445	2	achieves	achieve	VERB
ejpam-6087	445	3	a	a	DET
ejpam-6087	445	4	maximum	maximum	ADJ
ejpam-6087	445	5	auc	auc	NOUN
ejpam-6087	445	6	-	-	PUNCT
ejpam-6087	445	7	roc	roc	NOUN
ejpam-6087	445	8	of	of	ADP
ejpam-6087	445	9	79.51	79.51	NUM
ejpam-6087	445	10	%	%	NOUN
ejpam-6087	445	11	at	at	ADP
ejpam-6087	445	12	5	5	NUM
ejpam-6087	445	13	features	feature	NOUN
ejpam-6087	445	14	but	but	CCONJ
ejpam-6087	445	15	stabilizes	stabilize	VERB
ejpam-6087	445	16	around	around	ADV
ejpam-6087	445	17	62.61	62.61	NUM
ejpam-6087	445	18	%	%	NOUN
ejpam-6087	445	19	from	from	ADP
ejpam-6087	445	20	10	10	NUM
ejpam-6087	445	21	features	feature	NOUN
ejpam-6087	445	22	onward	onward	ADV
ejpam-6087	445	23	.	.	PUNCT
ejpam-6087	446	1	svm	svm	PROPN
ejpam-6087	446	2	consistently	consistently	ADV
ejpam-6087	446	3	shows	show	VERB
ejpam-6087	446	4	a	a	DET
ejpam-6087	446	5	low	low	ADJ
ejpam-6087	446	6	auc	auc	NOUN
ejpam-6087	446	7	-	-	PUNCT
ejpam-6087	446	8	roc	roc	NOUN
ejpam-6087	446	9	score	score	NOUN
ejpam-6087	446	10	of	of	ADP
ejpam-6087	446	11	55.73	55.73	NUM
ejpam-6087	446	12	%	%	NOUN
ejpam-6087	446	13	in	in	ADP
ejpam-6087	446	14	all	all	DET
ejpam-6087	446	15	feature	feature	NOUN
ejpam-6087	446	16	sets	set	NOUN
ejpam-6087	446	17	.	.	PUNCT
ejpam-6087	447	1	dt	dt	PUNCT
ejpam-6087	447	2	demonstrates	demonstrate	VERB
ejpam-6087	447	3	stable	stable	ADJ
ejpam-6087	447	4	auc	auc	NOUN
ejpam-6087	447	5	-	-	PUNCT
ejpam-6087	447	6	roc	roc	NOUN
ejpam-6087	447	7	scores	score	NOUN
ejpam-6087	447	8	of	of	ADP
ejpam-6087	447	9	approximately	approximately	ADV
ejpam-6087	447	10	90.80	90.80	NUM
ejpam-6087	447	11	%	%	NOUN
ejpam-6087	447	12	on	on	ADP
ejpam-6087	447	13	all	all	DET
ejpam-6087	447	14	feature	feature	NOUN
ejpam-6087	447	15	counts	count	NOUN
ejpam-6087	447	16	.	.	PUNCT
ejpam-6087	448	1	finally	finally	ADV
ejpam-6087	448	2	,	,	PUNCT
ejpam-6087	448	3	rf	rf	NOUN
ejpam-6087	448	4	maintains	maintain	VERB
ejpam-6087	448	5	the	the	DET
ejpam-6087	448	6	highest	high	ADJ
ejpam-6087	448	7	and	and	CCONJ
ejpam-6087	448	8	most	most	ADV
ejpam-6087	448	9	consistent	consistent	ADJ
ejpam-6087	448	10	auc	auc	ADJ
ejpam-6087	448	11	-	-	PUNCT
ejpam-6087	448	12	roc	roc	NOUN
ejpam-6087	448	13	scores	score	NOUN
ejpam-6087	448	14	,	,	PUNCT
ejpam-6087	448	15	ranging	range	VERB
ejpam-6087	448	16	from	from	ADP
ejpam-6087	448	17	95.81	95.81	NUM
ejpam-6087	448	18	%	%	NOUN
ejpam-6087	448	19	at	at	ADP
ejpam-6087	448	20	5	5	NUM
ejpam-6087	448	21	features	feature	NOUN
ejpam-6087	448	22	to	to	ADP
ejpam-6087	448	23	95.85	95.85	NUM
ejpam-6087	448	24	%	%	NOUN
ejpam-6087	448	25	at	at	ADP
ejpam-6087	448	26	30	30	NUM
ejpam-6087	448	27	features	feature	NOUN
ejpam-6087	448	28	.	.	PUNCT
ejpam-6087	449	1	4.3.2	4.3.2	X
ejpam-6087	449	2	.	.	PUNCT
ejpam-6087	450	1	selectkbest	selectkb	ADJ
ejpam-6087	450	2	restlts	restlt	NOUN
ejpam-6087	450	3	figure	figure	VERB
ejpam-6087	450	4	2	2	NUM
ejpam-6087	450	5	illustrates	illustrate	VERB
ejpam-6087	450	6	the	the	DET
ejpam-6087	450	7	impact	impact	NOUN
ejpam-6087	450	8	of	of	ADP
ejpam-6087	450	9	feature	feature	NOUN
ejpam-6087	450	10	selection	selection	NOUN
ejpam-6087	450	11	on	on	ADP
ejpam-6087	450	12	the	the	DET
ejpam-6087	450	13	performance	performance	NOUN
ejpam-6087	450	14	of	of	ADP
ejpam-6087	450	15	various	various	ADJ
ejpam-6087	450	16	ml	ml	NOUN
ejpam-6087	450	17	algorithms	algorithm	NOUN
ejpam-6087	450	18	as	as	ADP
ejpam-6087	450	19	the	the	DET
ejpam-6087	450	20	number	number	NOUN
ejpam-6087	450	21	of	of	ADP
ejpam-6087	450	22	selected	select	VERB
ejpam-6087	450	23	features	feature	NOUN
ejpam-6087	450	24	increases	increase	NOUN
ejpam-6087	450	25	from	from	ADP
ejpam-6087	450	26	5	5	NUM
ejpam-6087	450	27	to	to	PART
ejpam-6087	450	28	30	30	NUM
ejpam-6087	450	29	.	.	PUNCT
ejpam-6087	451	1	as	as	SCONJ
ejpam-6087	451	2	shown	show	VERB
ejpam-6087	451	3	in	in	ADP
ejpam-6087	451	4	the	the	DET
ejpam-6087	451	5	figure	figure	NOUN
ejpam-6087	451	6	2a	2a	NUM
ejpam-6087	451	7	,	,	PUNCT
ejpam-6087	451	8	the	the	DET
ejpam-6087	451	9	accuracy	accuracy	NOUN
ejpam-6087	451	10	of	of	ADP
ejpam-6087	451	11	various	various	ADJ
ejpam-6087	451	12	ml	ml	NOUN
ejpam-6087	451	13	models	model	NOUN
ejpam-6087	451	14	is	be	AUX
ejpam-6087	451	15	affected	affect	VERB
ejpam-6087	451	16	by	by	ADP
ejpam-6087	451	17	increasing	increase	VERB
ejpam-6087	451	18	the	the	DET
ejpam-6087	451	19	number	number	NOUN
ejpam-6087	451	20	of	of	ADP
ejpam-6087	451	21	features	feature	NOUN
ejpam-6087	451	22	selected	select	VERB
ejpam-6087	451	23	using	use	VERB
ejpam-6087	451	24	the	the	DET
ejpam-6087	451	25	selectkbest	selectkb	ADJ
ejpam-6087	451	26	method	method	NOUN
ejpam-6087	451	27	from	from	ADP
ejpam-6087	451	28	5	5	NUM
ejpam-6087	451	29	to	to	PART
ejpam-6087	451	30	30	30	NUM
ejpam-6087	451	31	.	.	PUNCT
ejpam-6087	452	1	for	for	ADP
ejpam-6087	452	2	lr	lr	PROPN
ejpam-6087	452	3	,	,	PUNCT
ejpam-6087	452	4	dt	dt	X
ejpam-6087	452	5	,	,	PUNCT
ejpam-6087	452	6	and	and	CCONJ
ejpam-6087	452	7	rf	rf	VERB
ejpam-6087	452	8	,	,	PUNCT
ejpam-6087	452	9	the	the	DET
ejpam-6087	452	10	accuracy	accuracy	NOUN
ejpam-6087	452	11	remains	remain	VERB
ejpam-6087	452	12	relatively	relatively	ADV
ejpam-6087	452	13	stable	stable	ADJ
ejpam-6087	452	14	across	across	ADP
ejpam-6087	452	15	the	the	DET
ejpam-6087	452	16	range	range	NOUN
ejpam-6087	452	17	of	of	ADP
ejpam-6087	452	18	features	feature	NOUN
ejpam-6087	452	19	.	.	PUNCT
ejpam-6087	453	1	specifically	specifically	ADV
ejpam-6087	453	2	,	,	PUNCT
ejpam-6087	453	3	lr	lr	PROPN
ejpam-6087	453	4	maintains	maintain	VERB
ejpam-6087	453	5	a	a	DET
ejpam-6087	453	6	high	high	ADJ
ejpam-6087	453	7	accuracy	accuracy	NOUN
ejpam-6087	453	8	of	of	ADP
ejpam-6087	453	9	approximately	approximately	ADV
ejpam-6087	453	10	95.80	95.80	NUM
ejpam-6087	453	11	%	%	NOUN
ejpam-6087	453	12	throughout	throughout	ADP
ejpam-6087	453	13	the	the	DET
ejpam-6087	453	14	feature	feature	NOUN
ejpam-6087	453	15	selection	selection	NOUN
ejpam-6087	453	16	process	process	NOUN
ejpam-6087	453	17	.	.	PUNCT
ejpam-6087	454	1	svm	svm	PROPN
ejpam-6087	454	2	exhibits	exhibit	VERB
ejpam-6087	454	3	a	a	DET
ejpam-6087	454	4	significant	significant	ADJ
ejpam-6087	454	5	decrease	decrease	NOUN
ejpam-6087	454	6	in	in	ADP
ejpam-6087	454	7	accuracy	accuracy	NOUN
ejpam-6087	454	8	,	,	PUNCT
ejpam-6087	454	9	dropping	drop	VERB
ejpam-6087	454	10	from	from	ADP
ejpam-6087	454	11	95.80	95.80	NUM
ejpam-6087	454	12	%	%	NOUN
ejpam-6087	454	13	at	at	ADP
ejpam-6087	454	14	5	5	NUM
ejpam-6087	454	15	features	feature	NOUN
ejpam-6087	454	16	to	to	ADP
ejpam-6087	454	17	53.99	53.99	NUM
ejpam-6087	454	18	%	%	NOUN
ejpam-6087	454	19	at	at	ADP
ejpam-6087	454	20	30	30	NUM
ejpam-6087	454	21	features	feature	NOUN
ejpam-6087	454	22	,	,	PUNCT
ejpam-6087	454	23	indicating	indicate	VERB
ejpam-6087	454	24	a	a	DET
ejpam-6087	454	25	considerable	considerable	ADJ
ejpam-6087	454	26	decline	decline	NOUN
ejpam-6087	454	27	in	in	ADP
ejpam-6087	454	28	performance	performance	NOUN
ejpam-6087	454	29	as	as	ADP
ejpam-6087	454	30	the	the	DET
ejpam-6087	454	31	number	number	NOUN
ejpam-6087	454	32	of	of	ADP
ejpam-6087	454	33	features	feature	NOUN
ejpam-6087	454	34	increases	increase	NOUN
ejpam-6087	454	35	.	.	PUNCT
ejpam-6087	455	1	in	in	ADP
ejpam-6087	455	2	contrast	contrast	NOUN
ejpam-6087	455	3	,	,	PUNCT
ejpam-6087	455	4	nb	nb	PROPN
ejpam-6087	455	5	and	and	CCONJ
ejpam-6087	455	6	knn	knn	PROPN
ejpam-6087	455	7	also	also	ADV
ejpam-6087	455	8	show	show	VERB
ejpam-6087	455	9	a	a	DET
ejpam-6087	455	10	decrease	decrease	NOUN
ejpam-6087	455	11	in	in	ADP
ejpam-6087	455	12	accuracy	accuracy	NOUN
ejpam-6087	455	13	,	,	PUNCT
ejpam-6087	455	14	but	but	CCONJ
ejpam-6087	455	15	not	not	PART
ejpam-6087	455	16	as	as	ADV
ejpam-6087	455	17	drastic	drastic	ADJ
ejpam-6087	455	18	as	as	ADP
ejpam-6087	455	19	svm	svm	ADJ
ejpam-6087	455	20	,	,	PUNCT
ejpam-6087	455	21	with	with	ADP
ejpam-6087	455	22	nb	nb	INTJ
ejpam-6087	455	23	falling	fall	VERB
ejpam-6087	455	24	from	from	ADP
ejpam-6087	455	25	95.80	95.80	NUM
ejpam-6087	455	26	%	%	NOUN
ejpam-6087	455	27	at	at	ADP
ejpam-6087	455	28	5	5	NUM
ejpam-6087	455	29	features	feature	NOUN
ejpam-6087	455	30	to	to	ADP
ejpam-6087	455	31	around	around	ADP
ejpam-6087	455	32	61.20	61.20	NUM
ejpam-6087	455	33	%	%	NOUN
ejpam-6087	455	34	at	at	ADP
ejpam-6087	455	35	30	30	NUM
ejpam-6087	455	36	features	feature	NOUN
ejpam-6087	455	37	and	and	CCONJ
ejpam-6087	455	38	knn	knn	AUX
ejpam-6087	455	39	decreasing	decrease	VERB
ejpam-6087	455	40	from	from	ADP
ejpam-6087	455	41	95.69	95.69	NUM
ejpam-6087	455	42	%	%	NOUN
ejpam-6087	455	43	at	at	ADP
ejpam-6087	455	44	5	5	NUM
ejpam-6087	455	45	features	feature	NOUN
ejpam-6087	455	46	to	to	ADP
ejpam-6087	455	47	57.34	57.34	NUM
ejpam-6087	455	48	%	%	NOUN
ejpam-6087	455	49	at	at	ADP
ejpam-6087	455	50	30	30	NUM
ejpam-6087	455	51	features	feature	NOUN
ejpam-6087	455	52	.	.	PUNCT
ejpam-6087	456	1	mlp	mlp	NOUN
ejpam-6087	456	2	demonstrates	demonstrate	VERB
ejpam-6087	456	3	variable	variable	ADJ
ejpam-6087	456	4	performance	performance	NOUN
ejpam-6087	456	5	,	,	PUNCT
ejpam-6087	456	6	with	with	ADP
ejpam-6087	456	7	a	a	DET
ejpam-6087	456	8	notable	notable	ADJ
ejpam-6087	456	9	drop	drop	NOUN
ejpam-6087	456	10	in	in	ADP
ejpam-6087	456	11	15	15	NUM
ejpam-6087	456	12	features	feature	NOUN
ejpam-6087	456	13	but	but	CCONJ
ejpam-6087	456	14	improved	improve	VERB
ejpam-6087	456	15	to	to	ADP
ejpam-6087	456	16	around	around	ADP
ejpam-6087	456	17	90.64	90.64	NUM
ejpam-6087	456	18	%	%	NOUN
ejpam-6087	456	19	at	at	ADP
ejpam-6087	456	20	30	30	NUM
ejpam-6087	456	21	features	feature	NOUN
ejpam-6087	456	22	.	.	PUNCT
ejpam-6087	457	1	overall	overall	ADV
ejpam-6087	457	2	,	,	PUNCT
ejpam-6087	457	3	most	most	ADJ
ejpam-6087	457	4	models	model	NOUN
ejpam-6087	457	5	achieve	achieve	VERB
ejpam-6087	457	6	optimal	optimal	ADJ
ejpam-6087	457	7	or	or	CCONJ
ejpam-6087	457	8	near	near	ADV
ejpam-6087	457	9	-	-	PUNCT
ejpam-6087	457	10	optimal	optimal	ADJ
ejpam-6087	457	11	accuracy	accuracy	NOUN
ejpam-6087	457	12	with	with	ADP
ejpam-6087	457	13	a	a	DET
ejpam-6087	457	14	higher	high	ADJ
ejpam-6087	457	15	number	number	NOUN
ejpam-6087	457	16	of	of	ADP
ejpam-6087	457	17	features	feature	NOUN
ejpam-6087	457	18	,	,	PUNCT
ejpam-6087	457	19	except	except	SCONJ
ejpam-6087	457	20	for	for	ADP
ejpam-6087	457	21	svm	svm	PROPN
ejpam-6087	457	22	,	,	PUNCT
ejpam-6087	457	23	which	which	PRON
ejpam-6087	457	24	experiences	experience	VERB
ejpam-6087	457	25	a	a	DET
ejpam-6087	457	26	significant	significant	ADJ
ejpam-6087	457	27	drop	drop	NOUN
ejpam-6087	457	28	in	in	ADP
ejpam-6087	457	29	performance	performance	NOUN
ejpam-6087	457	30	.	.	PUNCT
ejpam-6087	458	1	figure	figure	NOUN
ejpam-6087	458	2	2b	2b	NOUN
ejpam-6087	458	3	illustrates	illustrate	VERB
ejpam-6087	458	4	how	how	SCONJ
ejpam-6087	458	5	the	the	DET
ejpam-6087	458	6	precision	precision	NOUN
ejpam-6087	458	7	of	of	ADP
ejpam-6087	458	8	various	various	ADJ
ejpam-6087	458	9	ml	ml	NOUN
ejpam-6087	458	10	models	model	NOUN
ejpam-6087	458	11	changes	change	NOUN
ejpam-6087	458	12	as	as	ADP
ejpam-6087	458	13	the	the	DET
ejpam-6087	458	14	number	number	NOUN
ejpam-6087	458	15	of	of	ADP
ejpam-6087	458	16	features	feature	NOUN
ejpam-6087	458	17	selected	select	VERB
ejpam-6087	458	18	using	use	VERB
ejpam-6087	458	19	the	the	DET
ejpam-6087	458	20	selectkbest	selectkb	ADJ
ejpam-6087	458	21	method	method	NOUN
ejpam-6087	458	22	increases	increase	NOUN
ejpam-6087	458	23	from	from	ADP
ejpam-6087	458	24	5	5	NUM
ejpam-6087	458	25	to	to	PART
ejpam-6087	458	26	30	30	NUM
ejpam-6087	458	27	.	.	PUNCT
ejpam-6087	459	1	for	for	ADP
ejpam-6087	459	2	lr	lr	PROPN
ejpam-6087	459	3	,	,	PUNCT
ejpam-6087	459	4	dt	dt	X
ejpam-6087	459	5	,	,	PUNCT
ejpam-6087	459	6	and	and	CCONJ
ejpam-6087	459	7	rf	rf	VERB
ejpam-6087	459	8	,	,	PUNCT
ejpam-6087	459	9	the	the	DET
ejpam-6087	459	10	precision	precision	NOUN
ejpam-6087	459	11	remains	remain	VERB
ejpam-6087	459	12	relatively	relatively	ADV
ejpam-6087	459	13	high	high	ADJ
ejpam-6087	459	14	and	and	CCONJ
ejpam-6087	459	15	stable	stable	ADJ
ejpam-6087	459	16	throughout	throughout	ADP
ejpam-6087	459	17	the	the	DET
ejpam-6087	459	18	range	range	NOUN
ejpam-6087	459	19	of	of	ADP
ejpam-6087	459	20	feature	feature	NOUN
ejpam-6087	459	21	selection	selection	NOUN
ejpam-6087	459	22	.	.	PUNCT
ejpam-6087	460	1	lr	lr	PROPN
ejpam-6087	460	2	maintains	maintain	VERB
ejpam-6087	460	3	a	a	DET
ejpam-6087	460	4	consistent	consistent	ADJ
ejpam-6087	460	5	precision	precision	NOUN
ejpam-6087	460	6	of	of	ADP
ejpam-6087	460	7	approximately	approximately	ADV
ejpam-6087	460	8	95.80	95.80	NUM
ejpam-6087	460	9	%	%	NOUN
ejpam-6087	460	10	,	,	PUNCT
ejpam-6087	460	11	and	and	CCONJ
ejpam-6087	460	12	rf	rf	NOUN
ejpam-6087	460	13	also	also	ADV
ejpam-6087	460	14	shows	show	VERB
ejpam-6087	460	15	high	high	ADJ
ejpam-6087	460	16	precision	precision	NOUN
ejpam-6087	460	17	around	around	ADP
ejpam-6087	460	18	95.80	95.80	NUM
ejpam-6087	460	19	%	%	NOUN
ejpam-6087	460	20	throughout	throughout	ADP
ejpam-6087	460	21	the	the	DET
ejpam-6087	460	22	feature	feature	NOUN
ejpam-6087	460	23	selection	selection	NOUN
ejpam-6087	460	24	process	process	NOUN
ejpam-6087	460	25	.	.	PUNCT
ejpam-6087	461	1	in	in	ADP
ejpam-6087	461	2	contrast	contrast	NOUN
ejpam-6087	461	3	,	,	PUNCT
ejpam-6087	461	4	nb	nb	INTJ
ejpam-6087	461	5	exhibits	exhibit	VERB
ejpam-6087	461	6	a	a	DET
ejpam-6087	461	7	significant	significant	ADJ
ejpam-6087	461	8	decline	decline	NOUN
ejpam-6087	461	9	in	in	ADP
ejpam-6087	461	10	precision	precision	NOUN
ejpam-6087	461	11	from	from	ADP
ejpam-6087	461	12	95.80	95.80	NUM
ejpam-6087	461	13	%	%	NOUN
ejpam-6087	461	14	at	at	ADP
ejpam-6087	461	15	5	5	NUM
ejpam-6087	461	16	features	feature	NOUN
ejpam-6087	461	17	to	to	ADP
ejpam-6087	461	18	approximately	approximately	ADV
ejpam-6087	461	19	61.22	61.22	NUM
ejpam-6087	461	20	%	%	NOUN
ejpam-6087	461	21	at	at	ADP
ejpam-6087	461	22	30	30	NUM
ejpam-6087	461	23	features	feature	NOUN
ejpam-6087	461	24	,	,	PUNCT
ejpam-6087	461	25	with	with	ADP
ejpam-6087	461	26	only	only	ADJ
ejpam-6087	461	27	minor	minor	ADJ
ejpam-6087	461	28	fluctuations	fluctuation	NOUN
ejpam-6087	461	29	at	at	ADP
ejpam-6087	461	30	intermediate	intermediate	ADJ
ejpam-6087	461	31	feature	feature	NOUN
ejpam-6087	461	32	counts	count	NOUN
ejpam-6087	461	33	.	.	PUNCT
ejpam-6087	462	1	knn	knn	PROPN
ejpam-6087	462	2	also	also	ADV
ejpam-6087	462	3	experiences	experience	VERB
ejpam-6087	462	4	a	a	DET
ejpam-6087	462	5	decrease	decrease	NOUN
ejpam-6087	462	6	in	in	ADP
ejpam-6087	462	7	precision	precision	NOUN
ejpam-6087	462	8	from	from	ADP
ejpam-6087	462	9	95.69	95.69	NUM
ejpam-6087	462	10	%	%	NOUN
ejpam-6087	462	11	with	with	ADP
ejpam-6087	462	12	5	5	NUM
ejpam-6087	462	13	features	feature	NOUN
ejpam-6087	462	14	to	to	ADP
ejpam-6087	462	15	57.32	57.32	NUM
ejpam-6087	462	16	%	%	NOUN
ejpam-6087	462	17	with	with	ADP
ejpam-6087	462	18	30	30	NUM
ejpam-6087	462	19	features	feature	NOUN
ejpam-6087	462	20	.	.	PUNCT
ejpam-6087	463	1	the	the	DET
ejpam-6087	463	2	mlp	mlp	NOUN
ejpam-6087	463	3	shows	show	VERB
ejpam-6087	463	4	a	a	DET
ejpam-6087	463	5	fluctuating	fluctuate	VERB
ejpam-6087	463	6	precision	precision	NOUN
ejpam-6087	463	7	,	,	PUNCT
ejpam-6087	463	8	dropping	drop	VERB
ejpam-6087	463	9	to	to	ADP
ejpam-6087	463	10	78.46	78.46	NUM
ejpam-6087	463	11	%	%	NOUN
ejpam-6087	463	12	at	at	ADP
ejpam-6087	463	13	15	15	NUM
ejpam-6087	463	14	features	feature	NOUN
ejpam-6087	463	15	but	but	CCONJ
ejpam-6087	463	16	recovering	recover	VERB
ejpam-6087	463	17	to	to	ADP
ejpam-6087	463	18	around	around	ADP
ejpam-6087	463	19	90.84	90.84	NUM
ejpam-6087	463	20	%	%	NOUN
ejpam-6087	463	21	at	at	ADP
ejpam-6087	463	22	30	30	NUM
ejpam-6087	463	23	features	feature	NOUN
ejpam-6087	463	24	.	.	PUNCT
ejpam-6087	464	1	the	the	DET
ejpam-6087	464	2	dt	dt	PROPN
ejpam-6087	464	3	maintains	maintain	VERB
ejpam-6087	464	4	relatively	relatively	ADV
ejpam-6087	464	5	constant	constant	ADJ
ejpam-6087	464	6	precision	precision	NOUN
ejpam-6087	464	7	around	around	ADP
ejpam-6087	464	8	90.63	90.63	NUM
ejpam-6087	464	9	%	%	NOUN
ejpam-6087	464	10	to	to	ADP
ejpam-6087	464	11	90.69	90.69	NUM
ejpam-6087	464	12	%	%	NOUN
ejpam-6087	464	13	throughout	throughout	ADP
ejpam-6087	464	14	the	the	DET
ejpam-6087	464	15	feature	feature	NOUN
ejpam-6087	464	16	range	range	NOUN
ejpam-6087	464	17	.	.	PUNCT
ejpam-6087	465	1	overall	overall	ADV
ejpam-6087	465	2	,	,	PUNCT
ejpam-6087	465	3	most	most	ADJ
ejpam-6087	465	4	models	model	NOUN
ejpam-6087	465	5	see	see	VERB
ejpam-6087	465	6	stable	stable	ADJ
ejpam-6087	465	7	or	or	CCONJ
ejpam-6087	465	8	improved	improved	ADJ
ejpam-6087	465	9	precision	precision	NOUN
ejpam-6087	465	10	as	as	ADP
ejpam-6087	465	11	the	the	DET
ejpam-6087	465	12	number	number	NOUN
ejpam-6087	465	13	of	of	ADP
ejpam-6087	465	14	features	feature	NOUN
ejpam-6087	465	15	increases	increase	NOUN
ejpam-6087	465	16	,	,	PUNCT
ejpam-6087	465	17	with	with	ADP
ejpam-6087	465	18	notable	notable	ADJ
ejpam-6087	465	19	exceptions	exception	NOUN
ejpam-6087	465	20	for	for	ADP
ejpam-6087	465	21	nb	nb	NOUN
ejpam-6087	465	22	and	and	CCONJ
ejpam-6087	465	23	knn	knn	PROPN
ejpam-6087	465	24	,	,	PUNCT
ejpam-6087	465	25	which	which	PRON
ejpam-6087	465	26	exhibit	exhibit	VERB
ejpam-6087	465	27	considerable	considerable	ADJ
ejpam-6087	465	28	variability	variability	NOUN
ejpam-6087	465	29	.	.	PUNCT
ejpam-6087	466	1	m.	m.	NOUN
ejpam-6087	466	2	al	al	PROPN
ejpam-6087	466	3	-	-	PUNCT
ejpam-6087	466	4	tarawneh	tarawneh	PROPN
ejpam-6087	466	5	et	et	PROPN
ejpam-6087	466	6	al	al	PROPN
ejpam-6087	466	7	.	.	PUNCT
ejpam-6087	466	8	/	/	SYM
ejpam-6087	466	9	eur	eur	PROPN
ejpam-6087	466	10	.	.	PUNCT
ejpam-6087	467	1	j.	j.	PROPN
ejpam-6087	467	2	pure	pure	PROPN
ejpam-6087	467	3	appl	appl	PROPN
ejpam-6087	467	4	.	.	PROPN
ejpam-6087	467	5	math	math	PROPN
ejpam-6087	467	6	,	,	PUNCT
ejpam-6087	467	7	18	18	NUM
ejpam-6087	467	8	(	(	PUNCT
ejpam-6087	467	9	2	2	NUM
ejpam-6087	467	10	)	)	PUNCT
ejpam-6087	467	11	(	(	PUNCT
ejpam-6087	467	12	2025	2025	NUM
ejpam-6087	467	13	)	)	PUNCT
ejpam-6087	467	14	,	,	PUNCT
ejpam-6087	467	15	6087	6087	NUM
ejpam-6087	467	16	25	25	NUM
ejpam-6087	467	17	of	of	ADP
ejpam-6087	467	18	39	39	NUM
ejpam-6087	467	19	5	5	NUM
ejpam-6087	467	20	10	10	NUM
ejpam-6087	467	21	15	15	NUM
ejpam-6087	467	22	20	20	NUM
ejpam-6087	467	23	25	25	NUM
ejpam-6087	467	24	30	30	NUM
ejpam-6087	467	25	number	number	NOUN
ejpam-6087	467	26	of	of	ADP
ejpam-6087	467	27	features	feature	NOUN
ejpam-6087	467	28	50	50	NUM
ejpam-6087	467	29	60	60	NUM
ejpam-6087	467	30	70	70	NUM
ejpam-6087	467	31	80	80	NUM
ejpam-6087	467	32	90	90	NUM
ejpam-6087	467	33	100	100	NUM
ejpam-6087	467	34	a	a	PRON
ejpam-6087	467	35	cc	cc	ADP
ejpam-6087	467	36	ur	ur	INTJ
ejpam-6087	467	37	ac	ac	PROPN
ejpam-6087	467	38	y	y	PROPN
ejpam-6087	467	39	(	(	PUNCT
ejpam-6087	467	40	%	%	NOUN
ejpam-6087	467	41	)	)	PUNCT
ejpam-6087	467	42	lr	lr	PROPN
ejpam-6087	467	43	nb	nb	PROPN
ejpam-6087	467	44	knn	knn	PROPN
ejpam-6087	467	45	svm	svm	PROPN
ejpam-6087	467	46	mlp	mlp	PROPN
ejpam-6087	467	47	dt	dt	PROPN
ejpam-6087	467	48	rf	rf	ADP
ejpam-6087	467	49	a	a	DET
ejpam-6087	467	50	accuracy	accuracy	NOUN
ejpam-6087	467	51	5	5	NUM
ejpam-6087	467	52	10	10	NUM
ejpam-6087	467	53	15	15	NUM
ejpam-6087	467	54	20	20	NUM
ejpam-6087	467	55	25	25	NUM
ejpam-6087	467	56	30	30	NUM
ejpam-6087	467	57	number	number	NOUN
ejpam-6087	467	58	of	of	ADP
ejpam-6087	467	59	features	feature	NOUN
ejpam-6087	467	60	50	50	NUM
ejpam-6087	467	61	60	60	NUM
ejpam-6087	467	62	70	70	NUM
ejpam-6087	467	63	80	80	NUM
ejpam-6087	467	64	90	90	NUM
ejpam-6087	467	65	100	100	NUM
ejpam-6087	467	66	p	p	NOUN
ejpam-6087	467	67	re	re	ADP
ejpam-6087	467	68	ci	ci	PROPN
ejpam-6087	467	69	si	si	X
ejpam-6087	467	70	on	on	ADP
ejpam-6087	467	71	(	(	PUNCT
ejpam-6087	467	72	%	%	INTJ
ejpam-6087	467	73	)	)	PUNCT
ejpam-6087	467	74	lr	lr	PROPN
ejpam-6087	467	75	nb	nb	PROPN
ejpam-6087	467	76	knn	knn	PROPN
ejpam-6087	467	77	svm	svm	PROPN
ejpam-6087	467	78	mlp	mlp	PROPN
ejpam-6087	467	79	dt	dt	PROPN
ejpam-6087	467	80	rf	rf	PROPN
ejpam-6087	467	81	b	b	PROPN
ejpam-6087	467	82	precision	precision	NOUN
ejpam-6087	467	83	5	5	NUM
ejpam-6087	467	84	10	10	NUM
ejpam-6087	467	85	15	15	NUM
ejpam-6087	467	86	20	20	NUM
ejpam-6087	467	87	25	25	NUM
ejpam-6087	467	88	30	30	NUM
ejpam-6087	467	89	number	number	NOUN
ejpam-6087	467	90	of	of	ADP
ejpam-6087	467	91	features	feature	NOUN
ejpam-6087	467	92	50	50	NUM
ejpam-6087	467	93	60	60	NUM
ejpam-6087	467	94	70	70	NUM
ejpam-6087	467	95	80	80	NUM
ejpam-6087	467	96	90	90	NUM
ejpam-6087	467	97	100	100	NUM
ejpam-6087	467	98	r	r	NOUN
ejpam-6087	467	99	ec	ec	PROPN
ejpam-6087	467	100	al	al	PROPN
ejpam-6087	467	101	l	l	PROPN
ejpam-6087	467	102	(	(	PUNCT
ejpam-6087	467	103	%	%	INTJ
ejpam-6087	467	104	)	)	PUNCT
ejpam-6087	467	105	lr	lr	PROPN
ejpam-6087	467	106	nb	nb	PROPN
ejpam-6087	467	107	knn	knn	PROPN
ejpam-6087	467	108	svm	svm	PROPN
ejpam-6087	467	109	mlp	mlp	PROPN
ejpam-6087	467	110	dt	dt	PROPN
ejpam-6087	467	111	rf	rf	PROPN
ejpam-6087	467	112	c	c	NOUN
ejpam-6087	467	113	recall	recall	VERB
ejpam-6087	467	114	5	5	NUM
ejpam-6087	467	115	10	10	NUM
ejpam-6087	467	116	15	15	NUM
ejpam-6087	467	117	20	20	NUM
ejpam-6087	467	118	25	25	NUM
ejpam-6087	467	119	30	30	NUM
ejpam-6087	467	120	number	number	NOUN
ejpam-6087	467	121	of	of	ADP
ejpam-6087	467	122	features	feature	NOUN
ejpam-6087	468	1	50	50	NUM
ejpam-6087	468	2	60	60	NUM
ejpam-6087	468	3	70	70	NUM
ejpam-6087	468	4	80	80	NUM
ejpam-6087	468	5	90	90	NUM
ejpam-6087	468	6	100	100	NUM
ejpam-6087	468	7	f	f	NOUN
ejpam-6087	468	8	-1	-1	PROPN
ejpam-6087	468	9	s	s	X
ejpam-6087	468	10	co	co	VERB
ejpam-6087	468	11	re	re	X
ejpam-6087	468	12	(	(	PUNCT
ejpam-6087	468	13	%	%	INTJ
ejpam-6087	468	14	)	)	PUNCT
ejpam-6087	468	15	lr	lr	PROPN
ejpam-6087	468	16	nb	nb	PROPN
ejpam-6087	468	17	knn	knn	PROPN
ejpam-6087	468	18	svm	svm	PROPN
ejpam-6087	468	19	mlp	mlp	PROPN
ejpam-6087	468	20	dt	dt	PROPN
ejpam-6087	468	21	rf	rf	PROPN
ejpam-6087	468	22	d	d	PROPN
ejpam-6087	468	23	f1	f1	PROPN
ejpam-6087	468	24	score	score	NOUN
ejpam-6087	468	25	5	5	NUM
ejpam-6087	468	26	10	10	NUM
ejpam-6087	468	27	15	15	NUM
ejpam-6087	468	28	20	20	NUM
ejpam-6087	468	29	25	25	NUM
ejpam-6087	468	30	30	30	NUM
ejpam-6087	468	31	number	number	NOUN
ejpam-6087	468	32	of	of	ADP
ejpam-6087	468	33	features	feature	NOUN
ejpam-6087	468	34	50	50	NUM
ejpam-6087	468	35	60	60	NUM
ejpam-6087	468	36	70	70	NUM
ejpam-6087	468	37	80	80	NUM
ejpam-6087	468	38	90	90	NUM
ejpam-6087	468	39	100	100	NUM
ejpam-6087	468	40	a	a	DET
ejpam-6087	468	41	u	u	NOUN
ejpam-6087	468	42	c	c	NOUN
ejpam-6087	468	43	-r	-r	NOUN
ejpam-6087	468	44	o	o	NOUN
ejpam-6087	468	45	c	c	PROPN
ejpam-6087	468	46	(	(	PUNCT
ejpam-6087	468	47	%	%	NOUN
ejpam-6087	468	48	)	)	PUNCT
ejpam-6087	468	49	lr	lr	PROPN
ejpam-6087	468	50	nb	nb	PROPN
ejpam-6087	468	51	knn	knn	PROPN
ejpam-6087	468	52	svm	svm	PROPN
ejpam-6087	468	53	mlp	mlp	PROPN
ejpam-6087	468	54	dt	dt	PROPN
ejpam-6087	468	55	rf	rf	PROPN
ejpam-6087	468	56	e	e	PROPN
ejpam-6087	468	57	auc	auc	VERB
ejpam-6087	468	58	-	-	PUNCT
ejpam-6087	468	59	roc	roc	NOUN
ejpam-6087	468	60	figure	figure	NOUN
ejpam-6087	468	61	2	2	NUM
ejpam-6087	468	62	:	:	PUNCT
ejpam-6087	468	63	performance	performance	NOUN
ejpam-6087	468	64	metrics	metric	NOUN
ejpam-6087	468	65	of	of	ADP
ejpam-6087	468	66	various	various	ADJ
ejpam-6087	468	67	models	model	NOUN
ejpam-6087	468	68	as	as	SCONJ
ejpam-6087	468	69	feature	feature	NOUN
ejpam-6087	468	70	number	number	NOUN
ejpam-6087	468	71	increases	increase	VERB
ejpam-6087	468	72	under	under	ADP
ejpam-6087	468	73	selectkbest	selectkb	ADJ
ejpam-6087	468	74	.	.	PUNCT
ejpam-6087	469	1	m.	m.	NOUN
ejpam-6087	469	2	al	al	PROPN
ejpam-6087	469	3	-	-	PUNCT
ejpam-6087	469	4	tarawneh	tarawneh	PROPN
ejpam-6087	469	5	et	et	PROPN
ejpam-6087	469	6	al	al	PROPN
ejpam-6087	469	7	.	.	PUNCT
ejpam-6087	469	8	/	/	SYM
ejpam-6087	469	9	eur	eur	PROPN
ejpam-6087	469	10	.	.	PUNCT
ejpam-6087	470	1	j.	j.	PROPN
ejpam-6087	470	2	pure	pure	PROPN
ejpam-6087	470	3	appl	appl	PROPN
ejpam-6087	470	4	.	.	PROPN
ejpam-6087	470	5	math	math	PROPN
ejpam-6087	470	6	,	,	PUNCT
ejpam-6087	470	7	18	18	NUM
ejpam-6087	470	8	(	(	PUNCT
ejpam-6087	470	9	2	2	NUM
ejpam-6087	470	10	)	)	PUNCT
ejpam-6087	470	11	(	(	PUNCT
ejpam-6087	470	12	2025	2025	NUM
ejpam-6087	470	13	)	)	PUNCT
ejpam-6087	470	14	,	,	PUNCT
ejpam-6087	470	15	6087	6087	NUM
ejpam-6087	470	16	26	26	NUM
ejpam-6087	470	17	of	of	ADP
ejpam-6087	470	18	39	39	NUM
ejpam-6087	470	19	figure	figure	NOUN
ejpam-6087	470	20	2c	2c	NOUN
ejpam-6087	470	21	illustrates	illustrate	VERB
ejpam-6087	470	22	how	how	SCONJ
ejpam-6087	470	23	the	the	DET
ejpam-6087	470	24	recall	recall	NOUN
ejpam-6087	470	25	of	of	ADP
ejpam-6087	470	26	various	various	ADJ
ejpam-6087	470	27	ml	ml	NOUN
ejpam-6087	470	28	models	model	NOUN
ejpam-6087	470	29	is	be	AUX
ejpam-6087	470	30	affected	affect	VERB
ejpam-6087	470	31	by	by	ADP
ejpam-6087	470	32	increasing	increase	VERB
ejpam-6087	470	33	the	the	DET
ejpam-6087	470	34	number	number	NOUN
ejpam-6087	470	35	of	of	ADP
ejpam-6087	470	36	features	feature	NOUN
ejpam-6087	470	37	selected	select	VERB
ejpam-6087	470	38	using	use	VERB
ejpam-6087	470	39	the	the	DET
ejpam-6087	470	40	selectkbest	selectkb	ADJ
ejpam-6087	470	41	method	method	NOUN
ejpam-6087	470	42	from	from	ADP
ejpam-6087	470	43	5	5	NUM
ejpam-6087	470	44	to	to	PART
ejpam-6087	470	45	30	30	NUM
ejpam-6087	470	46	.	.	PUNCT
ejpam-6087	471	1	for	for	ADP
ejpam-6087	471	2	lr	lr	PROPN
ejpam-6087	471	3	,	,	PUNCT
ejpam-6087	471	4	mlp	mlp	PROPN
ejpam-6087	471	5	,	,	PUNCT
ejpam-6087	471	6	dt	dt	X
ejpam-6087	471	7	,	,	PUNCT
ejpam-6087	471	8	and	and	CCONJ
ejpam-6087	471	9	rf	rf	NOUN
ejpam-6087	471	10	,	,	PUNCT
ejpam-6087	471	11	recall	recall	VERB
ejpam-6087	471	12	remains	remain	VERB
ejpam-6087	471	13	relatively	relatively	ADV
ejpam-6087	471	14	high	high	ADJ
ejpam-6087	471	15	and	and	CCONJ
ejpam-6087	471	16	stable	stable	ADJ
ejpam-6087	471	17	throughout	throughout	ADP
ejpam-6087	471	18	the	the	DET
ejpam-6087	471	19	range	range	NOUN
ejpam-6087	471	20	of	of	ADP
ejpam-6087	471	21	features	feature	NOUN
ejpam-6087	471	22	.	.	PUNCT
ejpam-6087	472	1	lr	lr	NOUN
ejpam-6087	472	2	and	and	CCONJ
ejpam-6087	472	3	rf	rf	PRON
ejpam-6087	472	4	maintain	maintain	VERB
ejpam-6087	472	5	high	high	ADJ
ejpam-6087	472	6	recall	recall	NOUN
ejpam-6087	472	7	values	value	NOUN
ejpam-6087	472	8	,	,	PUNCT
ejpam-6087	472	9	remaining	remain	VERB
ejpam-6087	472	10	above	above	ADP
ejpam-6087	472	11	95.66	95.66	NUM
ejpam-6087	472	12	%	%	NOUN
ejpam-6087	472	13	for	for	ADP
ejpam-6087	472	14	all	all	DET
ejpam-6087	472	15	feature	feature	NOUN
ejpam-6087	472	16	counts	count	NOUN
ejpam-6087	472	17	.	.	PUNCT
ejpam-6087	473	1	dt	dt	PUNCT
ejpam-6087	473	2	also	also	ADV
ejpam-6087	473	3	shows	show	VERB
ejpam-6087	473	4	stable	stable	ADJ
ejpam-6087	473	5	high	high	ADJ
ejpam-6087	473	6	recall	recall	NOUN
ejpam-6087	473	7	around	around	ADP
ejpam-6087	473	8	90.63	90.63	NUM
ejpam-6087	473	9	%	%	NOUN
ejpam-6087	473	10	to	to	ADP
ejpam-6087	473	11	90.69	90.69	NUM
ejpam-6087	473	12	%	%	NOUN
ejpam-6087	473	13	.	.	PUNCT
ejpam-6087	474	1	in	in	ADP
ejpam-6087	474	2	contrast	contrast	NOUN
ejpam-6087	474	3	,	,	PUNCT
ejpam-6087	474	4	nb	nb	PROPN
ejpam-6087	474	5	and	and	CCONJ
ejpam-6087	474	6	knn	knn	PROPN
ejpam-6087	474	7	exhibit	exhibit	VERB
ejpam-6087	474	8	significant	significant	ADJ
ejpam-6087	474	9	variability	variability	NOUN
ejpam-6087	474	10	in	in	ADP
ejpam-6087	474	11	recall	recall	NOUN
ejpam-6087	474	12	.	.	PUNCT
ejpam-6087	475	1	nb	nb	PROPN
ejpam-6087	475	2	starts	start	VERB
ejpam-6087	475	3	with	with	ADP
ejpam-6087	475	4	a	a	DET
ejpam-6087	475	5	high	high	ADJ
ejpam-6087	475	6	recall	recall	NOUN
ejpam-6087	475	7	of	of	ADP
ejpam-6087	475	8	95.80	95.80	NUM
ejpam-6087	475	9	%	%	NOUN
ejpam-6087	475	10	at	at	ADP
ejpam-6087	475	11	5	5	NUM
ejpam-6087	475	12	features	feature	NOUN
ejpam-6087	475	13	but	but	CCONJ
ejpam-6087	475	14	drops	drop	NOUN
ejpam-6087	475	15	to	to	ADP
ejpam-6087	475	16	61.20	61.20	NUM
ejpam-6087	475	17	%	%	NOUN
ejpam-6087	475	18	at	at	ADP
ejpam-6087	475	19	30	30	NUM
ejpam-6087	475	20	features	feature	NOUN
ejpam-6087	475	21	,	,	PUNCT
ejpam-6087	475	22	with	with	ADP
ejpam-6087	475	23	a	a	DET
ejpam-6087	475	24	slight	slight	ADJ
ejpam-6087	475	25	increase	increase	NOUN
ejpam-6087	475	26	at	at	ADP
ejpam-6087	475	27	30	30	NUM
ejpam-6087	475	28	features	feature	NOUN
ejpam-6087	475	29	compared	compare	VERB
ejpam-6087	475	30	to	to	ADP
ejpam-6087	475	31	15	15	NUM
ejpam-6087	475	32	and	and	CCONJ
ejpam-6087	475	33	25	25	NUM
ejpam-6087	475	34	features	feature	NOUN
ejpam-6087	475	35	.	.	PUNCT
ejpam-6087	476	1	knn	knn	PROPN
ejpam-6087	476	2	also	also	ADV
ejpam-6087	476	3	shows	show	VERB
ejpam-6087	476	4	a	a	DET
ejpam-6087	476	5	decrease	decrease	NOUN
ejpam-6087	476	6	in	in	ADP
ejpam-6087	476	7	recall	recall	NOUN
ejpam-6087	476	8	from	from	ADP
ejpam-6087	476	9	95.69	95.69	NUM
ejpam-6087	476	10	%	%	NOUN
ejpam-6087	476	11	at	at	ADP
ejpam-6087	476	12	5	5	NUM
ejpam-6087	476	13	features	feature	NOUN
ejpam-6087	476	14	to	to	ADP
ejpam-6087	476	15	57.34	57.34	NUM
ejpam-6087	476	16	%	%	NOUN
ejpam-6087	476	17	at	at	ADP
ejpam-6087	476	18	30	30	NUM
ejpam-6087	476	19	features	feature	NOUN
ejpam-6087	476	20	,	,	PUNCT
ejpam-6087	476	21	reflecting	reflect	VERB
ejpam-6087	476	22	a	a	DET
ejpam-6087	476	23	steady	steady	ADJ
ejpam-6087	476	24	decline	decline	NOUN
ejpam-6087	476	25	in	in	ADP
ejpam-6087	476	26	performance	performance	NOUN
ejpam-6087	476	27	.	.	PUNCT
ejpam-6087	477	1	svm	svm	PROPN
ejpam-6087	477	2	shows	show	VERB
ejpam-6087	477	3	a	a	DET
ejpam-6087	477	4	generally	generally	ADV
ejpam-6087	477	5	low	low	ADJ
ejpam-6087	477	6	recall	recall	NOUN
ejpam-6087	477	7	across	across	ADP
ejpam-6087	477	8	all	all	DET
ejpam-6087	477	9	feature	feature	NOUN
ejpam-6087	477	10	counts	count	NOUN
ejpam-6087	477	11	,	,	PUNCT
ejpam-6087	477	12	starting	start	VERB
ejpam-6087	477	13	at	at	ADP
ejpam-6087	477	14	95.80	95.80	NUM
ejpam-6087	477	15	%	%	NOUN
ejpam-6087	477	16	at	at	ADP
ejpam-6087	477	17	5	5	NUM
ejpam-6087	477	18	features	feature	NOUN
ejpam-6087	477	19	and	and	CCONJ
ejpam-6087	477	20	decreasing	decrease	VERB
ejpam-6087	477	21	to	to	ADP
ejpam-6087	477	22	53.99	53.99	NUM
ejpam-6087	477	23	%	%	NOUN
ejpam-6087	477	24	at	at	ADP
ejpam-6087	477	25	30	30	NUM
ejpam-6087	477	26	features	feature	NOUN
ejpam-6087	477	27	.	.	PUNCT
ejpam-6087	478	1	mlp	mlp	NOUN
ejpam-6087	478	2	exhibits	exhibit	VERB
ejpam-6087	478	3	a	a	DET
ejpam-6087	478	4	notable	notable	ADJ
ejpam-6087	478	5	increase	increase	NOUN
ejpam-6087	478	6	in	in	ADP
ejpam-6087	478	7	recall	recall	NOUN
ejpam-6087	478	8	from	from	ADP
ejpam-6087	478	9	68.71	68.71	NUM
ejpam-6087	478	10	%	%	NOUN
ejpam-6087	478	11	at	at	ADP
ejpam-6087	478	12	15	15	NUM
ejpam-6087	478	13	features	feature	NOUN
ejpam-6087	478	14	to	to	ADP
ejpam-6087	478	15	90.64	90.64	NUM
ejpam-6087	478	16	%	%	NOUN
ejpam-6087	478	17	at	at	ADP
ejpam-6087	478	18	30	30	NUM
ejpam-6087	478	19	features	feature	NOUN
ejpam-6087	478	20	.	.	PUNCT
ejpam-6087	479	1	in	in	ADP
ejpam-6087	479	2	general	general	ADJ
ejpam-6087	479	3	,	,	PUNCT
ejpam-6087	479	4	most	most	ADJ
ejpam-6087	479	5	models	model	NOUN
ejpam-6087	479	6	maintain	maintain	VERB
ejpam-6087	479	7	or	or	CCONJ
ejpam-6087	479	8	improve	improve	VERB
ejpam-6087	479	9	recall	recall	NOUN
ejpam-6087	479	10	with	with	ADP
ejpam-6087	479	11	more	more	ADJ
ejpam-6087	479	12	features	feature	NOUN
ejpam-6087	479	13	,	,	PUNCT
ejpam-6087	479	14	except	except	SCONJ
ejpam-6087	479	15	for	for	ADP
ejpam-6087	479	16	nb	nb	PROPN
ejpam-6087	479	17	,	,	PUNCT
ejpam-6087	479	18	knn	knn	PROPN
ejpam-6087	479	19	,	,	PUNCT
ejpam-6087	479	20	and	and	CCONJ
ejpam-6087	479	21	svm	svm	PROPN
ejpam-6087	479	22	,	,	PUNCT
ejpam-6087	479	23	which	which	PRON
ejpam-6087	479	24	show	show	VERB
ejpam-6087	479	25	significant	significant	ADJ
ejpam-6087	479	26	declines	decline	NOUN
ejpam-6087	479	27	in	in	ADP
ejpam-6087	479	28	recall	recall	NOUN
ejpam-6087	479	29	performance	performance	NOUN
ejpam-6087	479	30	.	.	PUNCT
ejpam-6087	480	1	figure	figure	NOUN
ejpam-6087	480	2	2d	2d	PROPN
ejpam-6087	480	3	shows	show	VERB
ejpam-6087	480	4	the	the	DET
ejpam-6087	480	5	f-1	f-1	NOUN
ejpam-6087	480	6	scores	score	NOUN
ejpam-6087	480	7	of	of	ADP
ejpam-6087	480	8	various	various	ADJ
ejpam-6087	480	9	ml	ml	NOUN
ejpam-6087	480	10	models	model	NOUN
ejpam-6087	480	11	as	as	ADP
ejpam-6087	480	12	the	the	DET
ejpam-6087	480	13	number	number	NOUN
ejpam-6087	480	14	of	of	ADP
ejpam-6087	480	15	features	feature	NOUN
ejpam-6087	480	16	selected	select	VERB
ejpam-6087	480	17	using	use	VERB
ejpam-6087	480	18	the	the	DET
ejpam-6087	480	19	selectkbest	selectkb	ADJ
ejpam-6087	480	20	method	method	NOUN
ejpam-6087	480	21	increases	increase	NOUN
ejpam-6087	480	22	from	from	ADP
ejpam-6087	480	23	5	5	NUM
ejpam-6087	480	24	to	to	PART
ejpam-6087	480	25	30	30	NUM
ejpam-6087	480	26	.	.	PUNCT
ejpam-6087	481	1	lr	lr	PROPN
ejpam-6087	481	2	maintains	maintain	VERB
ejpam-6087	481	3	a	a	DET
ejpam-6087	481	4	consistently	consistently	ADV
ejpam-6087	481	5	high	high	ADJ
ejpam-6087	481	6	f-1	f-1	NOUN
ejpam-6087	481	7	score	score	NOUN
ejpam-6087	481	8	close	close	ADV
ejpam-6087	481	9	to	to	ADP
ejpam-6087	481	10	95.80	95.80	NUM
ejpam-6087	481	11	%	%	NOUN
ejpam-6087	481	12	under	under	ADP
ejpam-6087	481	13	all	all	DET
ejpam-6087	481	14	feature	feature	NOUN
ejpam-6087	481	15	counts	count	NOUN
ejpam-6087	481	16	.	.	PUNCT
ejpam-6087	482	1	rf	rf	PRON
ejpam-6087	482	2	also	also	ADV
ejpam-6087	482	3	shows	show	VERB
ejpam-6087	482	4	a	a	DET
ejpam-6087	482	5	stable	stable	ADJ
ejpam-6087	482	6	high	high	ADJ
ejpam-6087	482	7	f-1	f-1	NOUN
ejpam-6087	482	8	score	score	VERB
ejpam-6087	482	9	around	around	ADP
ejpam-6087	482	10	95.80	95.80	NUM
ejpam-6087	482	11	%	%	NOUN
ejpam-6087	482	12	,	,	PUNCT
ejpam-6087	482	13	demonstrating	demonstrate	VERB
ejpam-6087	482	14	consistent	consistent	ADJ
ejpam-6087	482	15	performance	performance	NOUN
ejpam-6087	482	16	regardless	regardless	ADV
ejpam-6087	482	17	of	of	ADP
ejpam-6087	482	18	the	the	DET
ejpam-6087	482	19	number	number	NOUN
ejpam-6087	482	20	of	of	ADP
ejpam-6087	482	21	features	feature	NOUN
ejpam-6087	482	22	.	.	PUNCT
ejpam-6087	483	1	nb	nb	INTJ
ejpam-6087	483	2	and	and	CCONJ
ejpam-6087	483	3	knn	knn	PROPN
ejpam-6087	483	4	exhibit	exhibit	VERB
ejpam-6087	483	5	notable	notable	ADJ
ejpam-6087	483	6	decreases	decrease	NOUN
ejpam-6087	483	7	in	in	ADP
ejpam-6087	483	8	f-1	f-1	NOUN
ejpam-6087	483	9	score	score	NOUN
ejpam-6087	483	10	as	as	ADP
ejpam-6087	483	11	the	the	DET
ejpam-6087	483	12	number	number	NOUN
ejpam-6087	483	13	of	of	ADP
ejpam-6087	483	14	features	feature	NOUN
ejpam-6087	483	15	increases	increase	NOUN
ejpam-6087	483	16	.	.	PUNCT
ejpam-6087	484	1	nb	nb	PROPN
ejpam-6087	484	2	starts	start	VERB
ejpam-6087	484	3	with	with	ADP
ejpam-6087	484	4	a	a	DET
ejpam-6087	484	5	high	high	ADJ
ejpam-6087	484	6	f-1	f-1	NOUN
ejpam-6087	484	7	score	score	NOUN
ejpam-6087	484	8	of	of	ADP
ejpam-6087	484	9	95.80	95.80	NUM
ejpam-6087	484	10	%	%	NOUN
ejpam-6087	484	11	at	at	ADP
ejpam-6087	484	12	five	five	NUM
ejpam-6087	484	13	features	feature	NOUN
ejpam-6087	484	14	but	but	CCONJ
ejpam-6087	484	15	decreases	decrease	NOUN
ejpam-6087	484	16	to	to	ADP
ejpam-6087	484	17	around	around	ADP
ejpam-6087	484	18	61.02	61.02	NUM
ejpam-6087	484	19	%	%	NOUN
ejpam-6087	484	20	at	at	ADP
ejpam-6087	484	21	30	30	NUM
ejpam-6087	484	22	features	feature	NOUN
ejpam-6087	484	23	.	.	PUNCT
ejpam-6087	485	1	similarly	similarly	ADV
ejpam-6087	485	2	,	,	PUNCT
ejpam-6087	485	3	knn	knn	PROPN
ejpam-6087	485	4	shows	show	VERB
ejpam-6087	485	5	a	a	DET
ejpam-6087	485	6	drop	drop	NOUN
ejpam-6087	485	7	in	in	ADP
ejpam-6087	485	8	the	the	DET
ejpam-6087	485	9	f-1	f-1	NOUN
ejpam-6087	485	10	score	score	NOUN
ejpam-6087	485	11	from	from	ADP
ejpam-6087	485	12	95.69	95.69	NUM
ejpam-6087	485	13	%	%	NOUN
ejpam-6087	485	14	at	at	ADP
ejpam-6087	485	15	5	5	NUM
ejpam-6087	485	16	features	feature	NOUN
ejpam-6087	485	17	to	to	ADP
ejpam-6087	485	18	57.32	57.32	NUM
ejpam-6087	485	19	%	%	NOUN
ejpam-6087	485	20	at	at	ADP
ejpam-6087	485	21	30	30	NUM
ejpam-6087	485	22	features	feature	NOUN
ejpam-6087	485	23	,	,	PUNCT
ejpam-6087	485	24	reflecting	reflect	VERB
ejpam-6087	485	25	a	a	DET
ejpam-6087	485	26	steady	steady	ADJ
ejpam-6087	485	27	decline	decline	NOUN
ejpam-6087	485	28	in	in	ADP
ejpam-6087	485	29	performance	performance	NOUN
ejpam-6087	485	30	.	.	PUNCT
ejpam-6087	486	1	svm	svm	PROPN
ejpam-6087	486	2	maintains	maintain	VERB
ejpam-6087	486	3	a	a	DET
ejpam-6087	486	4	low	low	ADJ
ejpam-6087	486	5	and	and	CCONJ
ejpam-6087	486	6	relatively	relatively	ADV
ejpam-6087	486	7	stable	stable	ADJ
ejpam-6087	486	8	f-1	f-1	NOUN
ejpam-6087	486	9	score	score	NOUN
ejpam-6087	486	10	throughout	throughout	ADP
ejpam-6087	486	11	the	the	DET
ejpam-6087	486	12	feature	feature	NOUN
ejpam-6087	486	13	range	range	NOUN
ejpam-6087	486	14	,	,	PUNCT
ejpam-6087	486	15	starting	start	VERB
ejpam-6087	486	16	at	at	ADP
ejpam-6087	486	17	95.80	95.80	NUM
ejpam-6087	486	18	%	%	NOUN
ejpam-6087	486	19	at	at	ADP
ejpam-6087	486	20	5	5	NUM
ejpam-6087	486	21	features	feature	NOUN
ejpam-6087	486	22	and	and	CCONJ
ejpam-6087	486	23	dropping	drop	VERB
ejpam-6087	486	24	to	to	ADP
ejpam-6087	486	25	54.00	54.00	NUM
ejpam-6087	486	26	%	%	NOUN
ejpam-6087	486	27	at	at	ADP
ejpam-6087	486	28	30	30	NUM
ejpam-6087	486	29	features	feature	NOUN
ejpam-6087	486	30	,	,	PUNCT
ejpam-6087	486	31	showing	show	VERB
ejpam-6087	486	32	a	a	DET
ejpam-6087	486	33	significant	significant	ADJ
ejpam-6087	486	34	decline	decline	NOUN
ejpam-6087	486	35	in	in	ADP
ejpam-6087	486	36	performance	performance	NOUN
ejpam-6087	486	37	as	as	ADP
ejpam-6087	486	38	the	the	DET
ejpam-6087	486	39	number	number	NOUN
ejpam-6087	486	40	of	of	ADP
ejpam-6087	486	41	features	feature	NOUN
ejpam-6087	486	42	increases	increase	NOUN
ejpam-6087	486	43	.	.	PUNCT
ejpam-6087	487	1	mlp	mlp	NOUN
ejpam-6087	487	2	exhibits	exhibit	VERB
ejpam-6087	487	3	fluctuations	fluctuation	NOUN
ejpam-6087	487	4	in	in	ADP
ejpam-6087	487	5	the	the	DET
ejpam-6087	487	6	f-1	f-1	NOUN
ejpam-6087	487	7	score	score	NOUN
ejpam-6087	487	8	,	,	PUNCT
ejpam-6087	487	9	with	with	ADP
ejpam-6087	487	10	a	a	DET
ejpam-6087	487	11	significant	significant	ADJ
ejpam-6087	487	12	drop	drop	NOUN
ejpam-6087	487	13	to	to	ADP
ejpam-6087	487	14	65.38	65.38	NUM
ejpam-6087	487	15	%	%	NOUN
ejpam-6087	487	16	at	at	ADP
ejpam-6087	487	17	15	15	NUM
ejpam-6087	487	18	features	feature	NOUN
ejpam-6087	487	19	but	but	CCONJ
ejpam-6087	487	20	recovered	recover	VERB
ejpam-6087	487	21	to	to	ADP
ejpam-6087	487	22	approximately	approximately	ADV
ejpam-6087	487	23	90.62	90.62	NUM
ejpam-6087	487	24	%	%	NOUN
ejpam-6087	487	25	at	at	ADP
ejpam-6087	487	26	30	30	NUM
ejpam-6087	487	27	features	feature	NOUN
ejpam-6087	487	28	.	.	PUNCT
ejpam-6087	488	1	dt	dt	PROPN
ejpam-6087	488	2	maintains	maintain	VERB
ejpam-6087	488	3	a	a	DET
ejpam-6087	488	4	relatively	relatively	ADV
ejpam-6087	488	5	stable	stable	ADJ
ejpam-6087	488	6	f-1	f-1	NOUN
ejpam-6087	488	7	score	score	NOUN
ejpam-6087	488	8	of	of	ADP
ejpam-6087	488	9	around	around	ADP
ejpam-6087	488	10	90.63	90.63	NUM
ejpam-6087	488	11	%	%	NOUN
ejpam-6087	488	12	to	to	ADP
ejpam-6087	488	13	90.69	90.69	NUM
ejpam-6087	488	14	%	%	NOUN
ejpam-6087	488	15	throughout	throughout	ADP
ejpam-6087	488	16	the	the	DET
ejpam-6087	488	17	feature	feature	NOUN
ejpam-6087	488	18	selection	selection	NOUN
ejpam-6087	488	19	process	process	NOUN
ejpam-6087	488	20	.	.	PUNCT
ejpam-6087	489	1	in	in	ADP
ejpam-6087	489	2	general	general	ADJ
ejpam-6087	489	3	,	,	PUNCT
ejpam-6087	489	4	most	most	ADJ
ejpam-6087	489	5	models	model	NOUN
ejpam-6087	489	6	,	,	PUNCT
ejpam-6087	489	7	except	except	SCONJ
ejpam-6087	489	8	nb	nb	PROPN
ejpam-6087	489	9	,	,	PUNCT
ejpam-6087	489	10	knn	knn	PROPN
ejpam-6087	489	11	and	and	CCONJ
ejpam-6087	489	12	svm	svm	PROPN
ejpam-6087	489	13	,	,	PUNCT
ejpam-6087	489	14	maintain	maintain	VERB
ejpam-6087	489	15	or	or	CCONJ
ejpam-6087	489	16	improve	improve	VERB
ejpam-6087	489	17	their	their	PRON
ejpam-6087	489	18	f-1	f-1	NOUN
ejpam-6087	489	19	scores	score	NOUN
ejpam-6087	489	20	as	as	ADP
ejpam-6087	489	21	the	the	DET
ejpam-6087	489	22	number	number	NOUN
ejpam-6087	489	23	of	of	ADP
ejpam-6087	489	24	features	feature	NOUN
ejpam-6087	489	25	increases	increase	NOUN
ejpam-6087	489	26	.	.	PUNCT
ejpam-6087	490	1	nb	nb	INTJ
ejpam-6087	490	2	and	and	CCONJ
ejpam-6087	490	3	knn	knn	PROPN
ejpam-6087	490	4	show	show	VERB
ejpam-6087	490	5	significant	significant	ADJ
ejpam-6087	490	6	decreases	decrease	NOUN
ejpam-6087	490	7	,	,	PUNCT
ejpam-6087	490	8	while	while	SCONJ
ejpam-6087	490	9	svm	svm	PROPN
ejpam-6087	490	10	’s	’s	PART
ejpam-6087	490	11	performance	performance	NOUN
ejpam-6087	490	12	declines	decline	VERB
ejpam-6087	490	13	substantially	substantially	ADV
ejpam-6087	490	14	.	.	PUNCT
ejpam-6087	491	1	figure	figure	VERB
ejpam-6087	491	2	2e	2e	PROPN
ejpam-6087	491	3	presents	present	VERB
ejpam-6087	491	4	the	the	DET
ejpam-6087	491	5	auc	auc	NOUN
ejpam-6087	491	6	-	-	PUNCT
ejpam-6087	491	7	roc	roc	NOUN
ejpam-6087	491	8	scores	score	NOUN
ejpam-6087	491	9	for	for	ADP
ejpam-6087	491	10	different	different	ADJ
ejpam-6087	491	11	ml	ml	NOUN
ejpam-6087	491	12	models	model	NOUN
ejpam-6087	491	13	as	as	ADP
ejpam-6087	491	14	the	the	DET
ejpam-6087	491	15	number	number	NOUN
ejpam-6087	491	16	of	of	ADP
ejpam-6087	491	17	features	feature	NOUN
ejpam-6087	491	18	selected	select	VERB
ejpam-6087	491	19	using	use	VERB
ejpam-6087	491	20	the	the	DET
ejpam-6087	491	21	selectkbest	selectkb	ADJ
ejpam-6087	491	22	method	method	NOUN
ejpam-6087	491	23	increases	increase	NOUN
ejpam-6087	491	24	from	from	ADP
ejpam-6087	491	25	5	5	NUM
ejpam-6087	491	26	to	to	PART
ejpam-6087	491	27	30	30	NUM
ejpam-6087	491	28	.	.	PUNCT
ejpam-6087	492	1	lr	lr	PROPN
ejpam-6087	492	2	maintains	maintain	VERB
ejpam-6087	492	3	a	a	DET
ejpam-6087	492	4	high	high	ADJ
ejpam-6087	492	5	and	and	CCONJ
ejpam-6087	492	6	relatively	relatively	ADV
ejpam-6087	492	7	stable	stable	ADJ
ejpam-6087	492	8	auc	auc	NOUN
ejpam-6087	492	9	-	-	PUNCT
ejpam-6087	492	10	roc	roc	NOUN
ejpam-6087	492	11	score	score	NOUN
ejpam-6087	492	12	,	,	PUNCT
ejpam-6087	492	13	starting	start	VERB
ejpam-6087	492	14	at	at	ADP
ejpam-6087	492	15	95.83	95.83	NUM
ejpam-6087	492	16	%	%	NOUN
ejpam-6087	492	17	with	with	ADP
ejpam-6087	492	18	5	5	NUM
ejpam-6087	492	19	features	feature	NOUN
ejpam-6087	492	20	and	and	CCONJ
ejpam-6087	492	21	slightly	slightly	ADV
ejpam-6087	492	22	decreasing	decrease	VERB
ejpam-6087	492	23	to	to	ADP
ejpam-6087	492	24	95.65	95.65	NUM
ejpam-6087	492	25	%	%	NOUN
ejpam-6087	492	26	with	with	ADP
ejpam-6087	492	27	30	30	NUM
ejpam-6087	492	28	features	feature	NOUN
ejpam-6087	492	29	.	.	PUNCT
ejpam-6087	493	1	similarly	similarly	ADV
ejpam-6087	493	2	,	,	PUNCT
ejpam-6087	493	3	rf	rf	NOUN
ejpam-6087	493	4	exhibits	exhibit	VERB
ejpam-6087	493	5	a	a	DET
ejpam-6087	493	6	high	high	ADJ
ejpam-6087	493	7	and	and	CCONJ
ejpam-6087	493	8	stable	stable	ADJ
ejpam-6087	493	9	aucroc	aucroc	NOUN
ejpam-6087	493	10	score	score	NOUN
ejpam-6087	493	11	,	,	PUNCT
ejpam-6087	493	12	remaining	remain	VERB
ejpam-6087	493	13	between	between	ADP
ejpam-6087	493	14	95.58	95.58	NUM
ejpam-6087	493	15	%	%	NOUN
ejpam-6087	493	16	and	and	CCONJ
ejpam-6087	493	17	95.74	95.74	NUM
ejpam-6087	493	18	%	%	NOUN
ejpam-6087	493	19	in	in	ADP
ejpam-6087	493	20	the	the	DET
ejpam-6087	493	21	range	range	NOUN
ejpam-6087	493	22	of	of	ADP
ejpam-6087	493	23	features	feature	NOUN
ejpam-6087	493	24	.	.	PUNCT
ejpam-6087	494	1	nb	nb	INTJ
ejpam-6087	494	2	and	and	CCONJ
ejpam-6087	494	3	knn	knn	PROPN
ejpam-6087	494	4	demonstrate	demonstrate	VERB
ejpam-6087	494	5	a	a	DET
ejpam-6087	494	6	consistent	consistent	ADJ
ejpam-6087	494	7	decline	decline	NOUN
ejpam-6087	494	8	in	in	ADP
ejpam-6087	494	9	auc	auc	NOUN
ejpam-6087	494	10	-	-	PUNCT
ejpam-6087	494	11	roc	roc	NOUN
ejpam-6087	494	12	as	as	SCONJ
ejpam-6087	494	13	more	more	ADJ
ejpam-6087	494	14	features	feature	NOUN
ejpam-6087	494	15	are	be	AUX
ejpam-6087	494	16	added	add	VERB
ejpam-6087	494	17	.	.	PUNCT
ejpam-6087	495	1	nb	nb	PROPN
ejpam-6087	495	2	has	have	VERB
ejpam-6087	495	3	an	an	DET
ejpam-6087	495	4	initial	initial	ADJ
ejpam-6087	495	5	auc	auc	NOUN
ejpam-6087	495	6	-	-	PUNCT
ejpam-6087	495	7	roc	roc	NOUN
ejpam-6087	495	8	of	of	ADP
ejpam-6087	495	9	95.82	95.82	NUM
ejpam-6087	495	10	%	%	NOUN
ejpam-6087	495	11	at	at	ADP
ejpam-6087	495	12	5	5	NUM
ejpam-6087	495	13	features	feature	NOUN
ejpam-6087	495	14	,	,	PUNCT
ejpam-6087	495	15	but	but	CCONJ
ejpam-6087	495	16	this	this	DET
ejpam-6087	495	17	score	score	NOUN
ejpam-6087	495	18	drops	drop	VERB
ejpam-6087	495	19	to	to	ADP
ejpam-6087	495	20	65.82	65.82	NUM
ejpam-6087	495	21	%	%	NOUN
ejpam-6087	495	22	at	at	ADP
ejpam-6087	495	23	30	30	NUM
ejpam-6087	495	24	features	feature	NOUN
ejpam-6087	495	25	.	.	PUNCT
ejpam-6087	496	1	knn	knn	PROPN
ejpam-6087	496	2	also	also	ADV
ejpam-6087	496	3	shows	show	VERB
ejpam-6087	496	4	a	a	DET
ejpam-6087	496	5	decrease	decrease	NOUN
ejpam-6087	496	6	from	from	ADP
ejpam-6087	496	7	95.77	95.77	NUM
ejpam-6087	496	8	%	%	NOUN
ejpam-6087	496	9	at	at	ADP
ejpam-6087	496	10	5	5	NUM
ejpam-6087	496	11	features	feature	NOUN
ejpam-6087	496	12	to	to	ADP
ejpam-6087	496	13	60.99	60.99	NUM
ejpam-6087	496	14	%	%	NOUN
ejpam-6087	496	15	at	at	ADP
ejpam-6087	496	16	30	30	NUM
ejpam-6087	496	17	features	feature	NOUN
ejpam-6087	496	18	,	,	PUNCT
ejpam-6087	496	19	reflecting	reflect	VERB
ejpam-6087	496	20	a	a	DET
ejpam-6087	496	21	clear	clear	ADJ
ejpam-6087	496	22	downward	downward	ADJ
ejpam-6087	496	23	trend	trend	NOUN
ejpam-6087	496	24	in	in	ADP
ejpam-6087	496	25	performance	performance	NOUN
ejpam-6087	496	26	.	.	PUNCT
ejpam-6087	497	1	svm	svm	PROPN
ejpam-6087	497	2	maintains	maintain	VERB
ejpam-6087	497	3	a	a	DET
ejpam-6087	497	4	lower	low	ADJ
ejpam-6087	497	5	and	and	CCONJ
ejpam-6087	497	6	more	more	ADV
ejpam-6087	497	7	stable	stable	ADJ
ejpam-6087	497	8	auc	auc	NOUN
ejpam-6087	497	9	-	-	PUNCT
ejpam-6087	497	10	roc	roc	NOUN
ejpam-6087	497	11	score	score	NOUN
ejpam-6087	497	12	across	across	ADP
ejpam-6087	497	13	all	all	DET
ejpam-6087	497	14	feature	feature	NOUN
ejpam-6087	497	15	counts	count	NOUN
ejpam-6087	497	16	,	,	PUNCT
ejpam-6087	497	17	starting	start	VERB
ejpam-6087	497	18	at	at	ADP
ejpam-6087	497	19	95.78	95.78	NUM
ejpam-6087	497	20	%	%	NOUN
ejpam-6087	497	21	at	at	ADP
ejpam-6087	497	22	5	5	NUM
ejpam-6087	497	23	features	feature	NOUN
ejpam-6087	497	24	and	and	CCONJ
ejpam-6087	497	25	decreasing	decrease	VERB
ejpam-6087	497	26	marginally	marginally	ADV
ejpam-6087	497	27	to	to	ADP
ejpam-6087	497	28	55.56	55.56	NUM
ejpam-6087	497	29	%	%	NOUN
ejpam-6087	497	30	at	at	ADP
ejpam-6087	497	31	30	30	NUM
ejpam-6087	497	32	features	feature	NOUN
ejpam-6087	497	33	.	.	PUNCT
ejpam-6087	498	1	this	this	PRON
ejpam-6087	498	2	indicates	indicate	VERB
ejpam-6087	498	3	a	a	DET
ejpam-6087	498	4	steady	steady	ADJ
ejpam-6087	498	5	decline	decline	NOUN
ejpam-6087	498	6	in	in	ADP
ejpam-6087	498	7	svm	svm	PROPN
ejpam-6087	498	8	’s	’s	PART
ejpam-6087	498	9	performance	performance	NOUN
ejpam-6087	498	10	as	as	SCONJ
ejpam-6087	498	11	more	more	ADJ
ejpam-6087	498	12	features	feature	NOUN
ejpam-6087	498	13	are	be	AUX
ejpam-6087	498	14	considered	consider	VERB
ejpam-6087	498	15	.	.	PUNCT
ejpam-6087	499	1	mlp	mlp	NOUN
ejpam-6087	499	2	shows	show	VERB
ejpam-6087	499	3	significant	significant	ADJ
ejpam-6087	499	4	variability	variability	NOUN
ejpam-6087	499	5	in	in	ADP
ejpam-6087	499	6	auc	auc	NOUN
ejpam-6087	499	7	-	-	PUNCT
ejpam-6087	499	8	roc	roc	NOUN
ejpam-6087	499	9	,	,	PUNCT
ejpam-6087	499	10	with	with	ADP
ejpam-6087	499	11	a	a	DET
ejpam-6087	499	12	marked	mark	VERB
ejpam-6087	499	13	drop	drop	NOUN
ejpam-6087	499	14	to	to	ADP
ejpam-6087	499	15	77.41	77.41	NUM
ejpam-6087	499	16	%	%	NOUN
ejpam-6087	499	17	at	at	ADP
ejpam-6087	499	18	15	15	NUM
ejpam-6087	499	19	features	feature	NOUN
ejpam-6087	499	20	but	but	CCONJ
ejpam-6087	499	21	recovered	recover	VERB
ejpam-6087	499	22	to	to	ADP
ejpam-6087	499	23	92.15	92.15	NUM
ejpam-6087	499	24	%	%	NOUN
ejpam-6087	499	25	at	at	ADP
ejpam-6087	499	26	30	30	NUM
ejpam-6087	499	27	features	feature	NOUN
ejpam-6087	499	28	,	,	PUNCT
ejpam-6087	499	29	showing	show	VERB
ejpam-6087	499	30	that	that	SCONJ
ejpam-6087	499	31	the	the	DET
ejpam-6087	499	32	performance	performance	NOUN
ejpam-6087	499	33	of	of	ADP
ejpam-6087	499	34	this	this	DET
ejpam-6087	499	35	model	model	NOUN
ejpam-6087	499	36	is	be	AUX
ejpam-6087	499	37	sensitive	sensitive	ADJ
ejpam-6087	499	38	to	to	ADP
ejpam-6087	499	39	the	the	DET
ejpam-6087	499	40	number	number	NOUN
ejpam-6087	499	41	of	of	ADP
ejpam-6087	499	42	features	feature	NOUN
ejpam-6087	499	43	selected	select	VERB
ejpam-6087	499	44	.	.	PUNCT
ejpam-6087	500	1	dt	dt	PROPN
ejpam-6087	500	2	maintains	maintain	VERB
ejpam-6087	500	3	m.	m.	PROPN
ejpam-6087	500	4	al	al	PROPN
ejpam-6087	500	5	-	-	PUNCT
ejpam-6087	500	6	tarawneh	tarawneh	PROPN
ejpam-6087	500	7	et	et	PROPN
ejpam-6087	500	8	al	al	PROPN
ejpam-6087	500	9	.	.	PUNCT
ejpam-6087	500	10	/	/	SYM
ejpam-6087	500	11	eur	eur	PROPN
ejpam-6087	500	12	.	.	PUNCT
ejpam-6087	501	1	j.	j.	PROPN
ejpam-6087	501	2	pure	pure	PROPN
ejpam-6087	501	3	appl	appl	PROPN
ejpam-6087	501	4	.	.	PROPN
ejpam-6087	501	5	math	math	PROPN
ejpam-6087	501	6	,	,	PUNCT
ejpam-6087	501	7	18	18	NUM
ejpam-6087	501	8	(	(	PUNCT
ejpam-6087	501	9	2	2	NUM
ejpam-6087	501	10	)	)	PUNCT
ejpam-6087	501	11	(	(	PUNCT
ejpam-6087	501	12	2025	2025	NUM
ejpam-6087	501	13	)	)	PUNCT
ejpam-6087	501	14	,	,	PUNCT
ejpam-6087	501	15	6087	6087	NUM
ejpam-6087	501	16	27	27	NUM
ejpam-6087	501	17	of	of	ADP
ejpam-6087	501	18	39	39	NUM
ejpam-6087	501	19	a	a	DET
ejpam-6087	501	20	relatively	relatively	ADV
ejpam-6087	501	21	stable	stable	ADJ
ejpam-6087	501	22	auc	auc	NOUN
ejpam-6087	501	23	-	-	PUNCT
ejpam-6087	501	24	roc	roc	NOUN
ejpam-6087	501	25	score	score	NOUN
ejpam-6087	501	26	between	between	ADP
ejpam-6087	501	27	90.62	90.62	NUM
ejpam-6087	501	28	%	%	NOUN
ejpam-6087	501	29	and	and	CCONJ
ejpam-6087	501	30	90.93	90.93	NUM
ejpam-6087	501	31	%	%	NOUN
ejpam-6087	501	32	throughout	throughout	ADP
ejpam-6087	501	33	the	the	DET
ejpam-6087	501	34	feature	feature	NOUN
ejpam-6087	501	35	selection	selection	NOUN
ejpam-6087	501	36	process	process	NOUN
ejpam-6087	501	37	,	,	PUNCT
ejpam-6087	501	38	indicating	indicate	VERB
ejpam-6087	501	39	consistent	consistent	ADJ
ejpam-6087	501	40	performance	performance	NOUN
ejpam-6087	501	41	with	with	ADP
ejpam-6087	501	42	the	the	DET
ejpam-6087	501	43	increasing	increase	VERB
ejpam-6087	501	44	number	number	NOUN
ejpam-6087	501	45	of	of	ADP
ejpam-6087	501	46	features	feature	NOUN
ejpam-6087	501	47	.	.	PUNCT
ejpam-6087	502	1	most	most	ADJ
ejpam-6087	502	2	models	model	NOUN
ejpam-6087	502	3	other	other	ADJ
ejpam-6087	502	4	than	than	SCONJ
ejpam-6087	502	5	nb	nb	ADV
ejpam-6087	502	6	and	and	CCONJ
ejpam-6087	502	7	knn	knn	PROPN
ejpam-6087	502	8	exhibit	exhibit	VERB
ejpam-6087	502	9	stable	stable	ADJ
ejpam-6087	502	10	or	or	CCONJ
ejpam-6087	502	11	improved	improve	VERB
ejpam-6087	502	12	auc	auc	NOUN
ejpam-6087	502	13	-	-	PUNCT
ejpam-6087	502	14	roc	roc	NOUN
ejpam-6087	502	15	scores	score	NOUN
ejpam-6087	502	16	as	as	ADP
ejpam-6087	502	17	the	the	DET
ejpam-6087	502	18	number	number	NOUN
ejpam-6087	502	19	of	of	ADP
ejpam-6087	502	20	features	feature	NOUN
ejpam-6087	502	21	increases	increase	NOUN
ejpam-6087	502	22	.	.	PUNCT
ejpam-6087	503	1	the	the	DET
ejpam-6087	503	2	performance	performance	NOUN
ejpam-6087	503	3	of	of	ADP
ejpam-6087	503	4	nb	nb	NOUN
ejpam-6087	503	5	and	and	CCONJ
ejpam-6087	503	6	knn	knn	PROPN
ejpam-6087	503	7	deteriorates	deteriorate	VERB
ejpam-6087	503	8	significantly	significantly	ADV
ejpam-6087	503	9	,	,	PUNCT
ejpam-6087	503	10	while	while	SCONJ
ejpam-6087	503	11	svm	svm	PROPN
ejpam-6087	503	12	shows	show	VERB
ejpam-6087	503	13	a	a	DET
ejpam-6087	503	14	consistent	consistent	ADJ
ejpam-6087	503	15	decline	decline	NOUN
ejpam-6087	503	16	,	,	PUNCT
ejpam-6087	503	17	and	and	CCONJ
ejpam-6087	503	18	mlp	mlp	NOUN
ejpam-6087	503	19	exhibits	exhibit	VERB
ejpam-6087	503	20	fluctuating	fluctuate	VERB
ejpam-6087	503	21	performance	performance	NOUN
ejpam-6087	503	22	.	.	PUNCT
ejpam-6087	504	1	evaluation	evaluation	NOUN
ejpam-6087	504	2	of	of	ADP
ejpam-6087	504	3	ml	ml	NOUN
ejpam-6087	504	4	models	model	NOUN
ejpam-6087	504	5	using	use	VERB
ejpam-6087	504	6	four	four	NUM
ejpam-6087	504	7	different	different	ADJ
ejpam-6087	504	8	performance	performance	NOUN
ejpam-6087	504	9	metrics	metric	NOUN
ejpam-6087	504	10	:	:	PUNCT
ejpam-6087	504	11	accuracy	accuracy	NOUN
ejpam-6087	504	12	,	,	PUNCT
ejpam-6087	504	13	precision	precision	NOUN
ejpam-6087	504	14	,	,	PUNCT
ejpam-6087	504	15	recall	recall	NOUN
ejpam-6087	504	16	,	,	PUNCT
ejpam-6087	504	17	and	and	CCONJ
ejpam-6087	504	18	auc	auc	NOUN
ejpam-6087	504	19	-	-	PUNCT
ejpam-6087	504	20	roc	roc	NOUN
ejpam-6087	504	21	reveals	reveal	VERB
ejpam-6087	504	22	distinct	distinct	ADJ
ejpam-6087	504	23	patterns	pattern	NOUN
ejpam-6087	504	24	in	in	ADP
ejpam-6087	504	25	how	how	SCONJ
ejpam-6087	504	26	these	these	DET
ejpam-6087	504	27	models	model	NOUN
ejpam-6087	504	28	perform	perform	VERB
ejpam-6087	504	29	as	as	ADP
ejpam-6087	504	30	the	the	DET
ejpam-6087	504	31	number	number	NOUN
ejpam-6087	504	32	of	of	ADP
ejpam-6087	504	33	features	feature	NOUN
ejpam-6087	504	34	increases	increase	NOUN
ejpam-6087	504	35	from	from	ADP
ejpam-6087	504	36	5	5	NUM
ejpam-6087	504	37	to	to	PART
ejpam-6087	504	38	30	30	NUM
ejpam-6087	504	39	using	use	VERB
ejpam-6087	504	40	the	the	DET
ejpam-6087	504	41	selectkbest	selectkb	ADJ
ejpam-6087	504	42	method	method	NOUN
ejpam-6087	504	43	.	.	PUNCT
ejpam-6087	505	1	lr	lr	NOUN
ejpam-6087	505	2	and	and	CCONJ
ejpam-6087	505	3	rf	rf	PROPN
ejpam-6087	505	4	exhibit	exhibit	VERB
ejpam-6087	505	5	robust	robust	ADJ
ejpam-6087	505	6	and	and	CCONJ
ejpam-6087	505	7	consistent	consistent	ADJ
ejpam-6087	505	8	performance	performance	NOUN
ejpam-6087	505	9	across	across	ADP
ejpam-6087	505	10	all	all	DET
ejpam-6087	505	11	metrics	metric	NOUN
ejpam-6087	505	12	.	.	PUNCT
ejpam-6087	506	1	for	for	ADP
ejpam-6087	506	2	accuracy	accuracy	NOUN
ejpam-6087	506	3	and	and	CCONJ
ejpam-6087	506	4	auc	auc	NOUN
ejpam-6087	506	5	-	-	PUNCT
ejpam-6087	506	6	roc	roc	NOUN
ejpam-6087	506	7	,	,	PUNCT
ejpam-6087	506	8	lr	lr	X
ejpam-6087	506	9	remains	remain	VERB
ejpam-6087	506	10	stable	stable	ADJ
ejpam-6087	506	11	at	at	ADP
ejpam-6087	506	12	approximately	approximately	ADV
ejpam-6087	506	13	95.80	95.80	NUM
ejpam-6087	506	14	%	%	NOUN
ejpam-6087	506	15	and	and	CCONJ
ejpam-6087	506	16	95.83	95.83	NUM
ejpam-6087	506	17	%	%	NOUN
ejpam-6087	506	18	respectively	respectively	ADV
ejpam-6087	506	19	,	,	PUNCT
ejpam-6087	506	20	while	while	SCONJ
ejpam-6087	506	21	rf	rf	PRON
ejpam-6087	506	22	also	also	ADV
ejpam-6087	506	23	maintains	maintain	VERB
ejpam-6087	506	24	high	high	ADJ
ejpam-6087	506	25	performance	performance	NOUN
ejpam-6087	506	26	with	with	ADP
ejpam-6087	506	27	accuracy	accuracy	NOUN
ejpam-6087	506	28	around	around	ADP
ejpam-6087	506	29	95.80	95.80	NUM
ejpam-6087	506	30	%	%	NOUN
ejpam-6087	506	31	and	and	CCONJ
ejpam-6087	506	32	auc	auc	NOUN
ejpam-6087	506	33	-	-	PUNCT
ejpam-6087	506	34	roc	roc	NOUN
ejpam-6087	506	35	between	between	ADP
ejpam-6087	506	36	95.58	95.58	NUM
ejpam-6087	506	37	%	%	NOUN
ejpam-6087	506	38	and	and	CCONJ
ejpam-6087	506	39	95.74	95.74	NUM
ejpam-6087	506	40	%	%	NOUN
ejpam-6087	506	41	.	.	PUNCT
ejpam-6087	507	1	mlp	mlp	NOUN
ejpam-6087	507	2	shows	show	VERB
ejpam-6087	507	3	variability	variability	NOUN
ejpam-6087	507	4	in	in	ADP
ejpam-6087	507	5	performance	performance	NOUN
ejpam-6087	507	6	,	,	PUNCT
ejpam-6087	507	7	with	with	ADP
ejpam-6087	507	8	fluctuating	fluctuate	VERB
ejpam-6087	507	9	precision	precision	NOUN
ejpam-6087	507	10	,	,	PUNCT
ejpam-6087	507	11	recall	recall	NOUN
ejpam-6087	507	12	,	,	PUNCT
ejpam-6087	507	13	and	and	CCONJ
ejpam-6087	507	14	auc	auc	NOUN
ejpam-6087	507	15	-	-	PUNCT
ejpam-6087	507	16	roc	roc	NOUN
ejpam-6087	507	17	,	,	PUNCT
ejpam-6087	507	18	reflecting	reflect	VERB
ejpam-6087	507	19	the	the	DET
ejpam-6087	507	20	sensitivity	sensitivity	NOUN
ejpam-6087	507	21	to	to	ADP
ejpam-6087	507	22	the	the	DET
ejpam-6087	507	23	number	number	NOUN
ejpam-6087	507	24	of	of	ADP
ejpam-6087	507	25	features	feature	NOUN
ejpam-6087	507	26	selected	select	VERB
ejpam-6087	507	27	.	.	PUNCT
ejpam-6087	508	1	in	in	ADP
ejpam-6087	508	2	contrast	contrast	NOUN
ejpam-6087	508	3	,	,	PUNCT
ejpam-6087	508	4	nb	nb	PROPN
ejpam-6087	508	5	and	and	CCONJ
ejpam-6087	508	6	knn	knn	PROPN
ejpam-6087	508	7	show	show	VERB
ejpam-6087	508	8	significant	significant	ADJ
ejpam-6087	508	9	declines	decline	NOUN
ejpam-6087	508	10	in	in	ADP
ejpam-6087	508	11	performance	performance	NOUN
ejpam-6087	508	12	metrics	metric	NOUN
ejpam-6087	508	13	as	as	ADP
ejpam-6087	508	14	the	the	DET
ejpam-6087	508	15	number	number	NOUN
ejpam-6087	508	16	of	of	ADP
ejpam-6087	508	17	features	feature	NOUN
ejpam-6087	508	18	increases	increase	NOUN
ejpam-6087	508	19	.	.	PUNCT
ejpam-6087	509	1	nb	nb	NOUN
ejpam-6087	509	2	experiences	experience	VERB
ejpam-6087	509	3	substantial	substantial	ADJ
ejpam-6087	509	4	drops	drop	NOUN
ejpam-6087	509	5	in	in	ADP
ejpam-6087	509	6	precision	precision	NOUN
ejpam-6087	509	7	,	,	PUNCT
ejpam-6087	509	8	recall	recall	NOUN
ejpam-6087	509	9	,	,	PUNCT
ejpam-6087	509	10	and	and	CCONJ
ejpam-6087	509	11	auc	auc	NOUN
ejpam-6087	509	12	-	-	PUNCT
ejpam-6087	509	13	roc	roc	NOUN
ejpam-6087	509	14	from	from	ADP
ejpam-6087	509	15	high	high	ADJ
ejpam-6087	509	16	initial	initial	ADJ
ejpam-6087	509	17	values	value	NOUN
ejpam-6087	509	18	to	to	ADP
ejpam-6087	509	19	around	around	ADP
ejpam-6087	509	20	61.02	61.02	NUM
ejpam-6087	509	21	%	%	NOUN
ejpam-6087	509	22	,	,	PUNCT
ejpam-6087	509	23	65.82	65.82	NUM
ejpam-6087	509	24	%	%	NOUN
ejpam-6087	509	25	,	,	PUNCT
ejpam-6087	509	26	and	and	CCONJ
ejpam-6087	509	27	65.82	65.82	NUM
ejpam-6087	509	28	%	%	NOUN
ejpam-6087	509	29	respectively	respectively	ADV
ejpam-6087	509	30	,	,	PUNCT
ejpam-6087	509	31	at	at	ADP
ejpam-6087	509	32	30	30	NUM
ejpam-6087	509	33	features	feature	NOUN
ejpam-6087	509	34	.	.	PUNCT
ejpam-6087	510	1	similarly	similarly	ADV
ejpam-6087	510	2	,	,	PUNCT
ejpam-6087	510	3	knn	knn	PROPN
ejpam-6087	510	4	demonstrates	demonstrate	VERB
ejpam-6087	510	5	a	a	DET
ejpam-6087	510	6	decrease	decrease	NOUN
ejpam-6087	510	7	in	in	ADP
ejpam-6087	510	8	accuracy	accuracy	NOUN
ejpam-6087	510	9	from	from	ADP
ejpam-6087	510	10	95.69	95.69	NUM
ejpam-6087	510	11	%	%	NOUN
ejpam-6087	510	12	to	to	ADP
ejpam-6087	510	13	57.32	57.32	NUM
ejpam-6087	510	14	%	%	NOUN
ejpam-6087	510	15	,	,	PUNCT
ejpam-6087	510	16	precision	precision	NOUN
ejpam-6087	510	17	from	from	ADP
ejpam-6087	510	18	95.69	95.69	NUM
ejpam-6087	510	19	%	%	NOUN
ejpam-6087	510	20	to	to	ADP
ejpam-6087	510	21	57.32	57.32	NUM
ejpam-6087	510	22	%	%	NOUN
ejpam-6087	510	23	,	,	PUNCT
ejpam-6087	510	24	and	and	CCONJ
ejpam-6087	510	25	auc	auc	NOUN
ejpam-6087	510	26	-	-	PUNCT
ejpam-6087	510	27	roc	roc	NOUN
ejpam-6087	510	28	from	from	ADP
ejpam-6087	510	29	95.77	95.77	NUM
ejpam-6087	510	30	%	%	NOUN
ejpam-6087	510	31	to	to	ADP
ejpam-6087	510	32	60.99	60.99	NUM
ejpam-6087	510	33	%	%	NOUN
ejpam-6087	510	34	over	over	ADP
ejpam-6087	510	35	the	the	DET
ejpam-6087	510	36	same	same	ADJ
ejpam-6087	510	37	range	range	NOUN
ejpam-6087	510	38	.	.	PUNCT
ejpam-6087	511	1	svm	svm	PROPN
ejpam-6087	511	2	consistently	consistently	ADV
ejpam-6087	511	3	shows	show	VERB
ejpam-6087	511	4	lower	low	ADJ
ejpam-6087	511	5	performance	performance	NOUN
ejpam-6087	511	6	across	across	ADP
ejpam-6087	511	7	all	all	DET
ejpam-6087	511	8	metrics	metric	NOUN
ejpam-6087	511	9	,	,	PUNCT
ejpam-6087	511	10	with	with	ADP
ejpam-6087	511	11	a	a	DET
ejpam-6087	511	12	substantial	substantial	ADJ
ejpam-6087	511	13	decrease	decrease	NOUN
ejpam-6087	511	14	in	in	ADP
ejpam-6087	511	15	recall	recall	NOUN
ejpam-6087	511	16	and	and	CCONJ
ejpam-6087	511	17	auc	auc	NOUN
ejpam-6087	511	18	-	-	PUNCT
ejpam-6087	511	19	roc	roc	NOUN
ejpam-6087	511	20	as	as	ADP
ejpam-6087	511	21	the	the	DET
ejpam-6087	511	22	number	number	NOUN
ejpam-6087	511	23	of	of	ADP
ejpam-6087	511	24	features	feature	NOUN
ejpam-6087	511	25	increases	increase	NOUN
ejpam-6087	511	26	,	,	PUNCT
ejpam-6087	511	27	and	and	CCONJ
ejpam-6087	511	28	a	a	DET
ejpam-6087	511	29	lower	low	ADJ
ejpam-6087	511	30	aucroc	aucroc	NOUN
ejpam-6087	511	31	value	value	NOUN
ejpam-6087	511	32	from	from	ADP
ejpam-6087	511	33	95.78	95.78	NUM
ejpam-6087	511	34	%	%	NOUN
ejpam-6087	511	35	to	to	ADP
ejpam-6087	511	36	55.56	55.56	NUM
ejpam-6087	511	37	%	%	NOUN
ejpam-6087	511	38	.	.	PUNCT
ejpam-6087	512	1	finally	finally	ADV
ejpam-6087	512	2	,	,	PUNCT
ejpam-6087	512	3	dt	dt	PUNCT
ejpam-6087	512	4	exhibits	exhibit	VERB
ejpam-6087	512	5	stable	stable	ADJ
ejpam-6087	512	6	performance	performance	NOUN
ejpam-6087	512	7	in	in	ADP
ejpam-6087	512	8	accuracy	accuracy	NOUN
ejpam-6087	512	9	,	,	PUNCT
ejpam-6087	512	10	precision	precision	NOUN
ejpam-6087	512	11	and	and	CCONJ
ejpam-6087	512	12	recall	recall	NOUN
ejpam-6087	512	13	,	,	PUNCT
ejpam-6087	512	14	maintaining	maintain	VERB
ejpam-6087	512	15	a	a	DET
ejpam-6087	512	16	consistent	consistent	ADJ
ejpam-6087	512	17	auc	auc	NOUN
ejpam-6087	512	18	-	-	PUNCT
ejpam-6087	512	19	roc	roc	NOUN
ejpam-6087	512	20	around	around	ADP
ejpam-6087	512	21	90.62	90.62	NUM
ejpam-6087	512	22	%	%	NOUN
ejpam-6087	512	23	to	to	ADP
ejpam-6087	512	24	90.93	90.93	NUM
ejpam-6087	512	25	%	%	NOUN
ejpam-6087	512	26	.	.	PUNCT
ejpam-6087	513	1	4.3.3	4.3.3	X
ejpam-6087	513	2	.	.	PUNCT
ejpam-6087	513	3	principal	principal	ADJ
ejpam-6087	513	4	component	component	NOUN
ejpam-6087	513	5	analysis	analysis	NOUN
ejpam-6087	513	6	results	result	NOUN
ejpam-6087	513	7	this	this	DET
ejpam-6087	513	8	section	section	NOUN
ejpam-6087	513	9	shows	show	VERB
ejpam-6087	513	10	the	the	DET
ejpam-6087	513	11	results	result	NOUN
ejpam-6087	513	12	of	of	ADP
ejpam-6087	513	13	applying	apply	VERB
ejpam-6087	513	14	pca	pca	PROPN
ejpam-6087	513	15	to	to	PART
ejpam-6087	513	16	reduce	reduce	VERB
ejpam-6087	513	17	dataset	dataset	ADJ
ejpam-6087	513	18	dimensionality	dimensionality	NOUN
ejpam-6087	513	19	and	and	CCONJ
ejpam-6087	513	20	its	its	PRON
ejpam-6087	513	21	effects	effect	NOUN
ejpam-6087	513	22	on	on	ADP
ejpam-6087	513	23	the	the	DET
ejpam-6087	513	24	performance	performance	NOUN
ejpam-6087	513	25	of	of	ADP
ejpam-6087	513	26	various	various	ADJ
ejpam-6087	513	27	ml	ml	NOUN
ejpam-6087	513	28	algorithms	algorithm	NOUN
ejpam-6087	513	29	.	.	PUNCT
ejpam-6087	514	1	examines	examine	VERB
ejpam-6087	514	2	how	how	SCONJ
ejpam-6087	514	3	increasing	increase	VERB
ejpam-6087	514	4	the	the	DET
ejpam-6087	514	5	number	number	NOUN
ejpam-6087	514	6	of	of	ADP
ejpam-6087	514	7	principal	principal	ADJ
ejpam-6087	514	8	components	component	NOUN
ejpam-6087	514	9	from	from	ADP
ejpam-6087	514	10	5	5	NUM
ejpam-6087	514	11	to	to	PART
ejpam-6087	514	12	30	30	NUM
ejpam-6087	514	13	influences	influence	NOUN
ejpam-6087	514	14	the	the	DET
ejpam-6087	514	15	accuracy	accuracy	NOUN
ejpam-6087	514	16	,	,	PUNCT
ejpam-6087	514	17	precision	precision	NOUN
ejpam-6087	514	18	,	,	PUNCT
ejpam-6087	514	19	recall	recall	NOUN
ejpam-6087	514	20	,	,	PUNCT
ejpam-6087	514	21	and	and	CCONJ
ejpam-6087	514	22	auc	auc	NOUN
ejpam-6087	514	23	-	-	PUNCT
ejpam-6087	514	24	roc	roc	NOUN
ejpam-6087	514	25	of	of	ADP
ejpam-6087	514	26	the	the	DET
ejpam-6087	514	27	models	model	NOUN
ejpam-6087	514	28	.	.	PUNCT
ejpam-6087	515	1	figure	figure	NOUN
ejpam-6087	515	2	3a	3a	PROPN
ejpam-6087	515	3	illustrates	illustrate	VERB
ejpam-6087	515	4	the	the	DET
ejpam-6087	515	5	accuracy	accuracy	NOUN
ejpam-6087	515	6	of	of	ADP
ejpam-6087	515	7	various	various	ADJ
ejpam-6087	515	8	ml	ml	NOUN
ejpam-6087	515	9	models	model	NOUN
ejpam-6087	515	10	as	as	ADP
ejpam-6087	515	11	the	the	DET
ejpam-6087	515	12	number	number	NOUN
ejpam-6087	515	13	of	of	ADP
ejpam-6087	515	14	principal	principal	ADJ
ejpam-6087	515	15	components	component	NOUN
ejpam-6087	515	16	increases	increase	NOUN
ejpam-6087	515	17	from	from	ADP
ejpam-6087	515	18	5	5	NUM
ejpam-6087	515	19	to	to	PART
ejpam-6087	515	20	30	30	NUM
ejpam-6087	515	21	.	.	PUNCT
ejpam-6087	516	1	the	the	DET
ejpam-6087	516	2	accuracy	accuracy	NOUN
ejpam-6087	516	3	results	result	NOUN
ejpam-6087	516	4	obtained	obtain	VERB
ejpam-6087	516	5	through	through	ADP
ejpam-6087	516	6	pca	pca	PROPN
ejpam-6087	516	7	are	be	AUX
ejpam-6087	516	8	generally	generally	ADV
ejpam-6087	516	9	lower	low	ADJ
ejpam-6087	516	10	compared	compare	VERB
ejpam-6087	516	11	to	to	ADP
ejpam-6087	516	12	those	those	PRON
ejpam-6087	516	13	of	of	ADP
ejpam-6087	516	14	other	other	ADJ
ejpam-6087	516	15	feature	feature	NOUN
ejpam-6087	516	16	selection	selection	NOUN
ejpam-6087	516	17	methods	method	NOUN
ejpam-6087	516	18	.	.	PUNCT
ejpam-6087	517	1	rf	rf	PRON
ejpam-6087	517	2	achieves	achieve	VERB
ejpam-6087	517	3	the	the	DET
ejpam-6087	517	4	highest	high	ADJ
ejpam-6087	517	5	accuracy	accuracy	NOUN
ejpam-6087	517	6	,	,	PUNCT
ejpam-6087	517	7	with	with	ADP
ejpam-6087	517	8	an	an	DET
ejpam-6087	517	9	increase	increase	NOUN
ejpam-6087	517	10	from	from	ADP
ejpam-6087	517	11	61.34	61.34	NUM
ejpam-6087	517	12	%	%	NOUN
ejpam-6087	517	13	at	at	ADP
ejpam-6087	517	14	5	5	NUM
ejpam-6087	517	15	components	component	NOUN
ejpam-6087	517	16	to	to	ADP
ejpam-6087	517	17	64.59	64.59	NUM
ejpam-6087	517	18	%	%	NOUN
ejpam-6087	517	19	at	at	ADP
ejpam-6087	517	20	30	30	NUM
ejpam-6087	517	21	components	component	NOUN
ejpam-6087	517	22	,	,	PUNCT
ejpam-6087	517	23	demonstrating	demonstrate	VERB
ejpam-6087	517	24	the	the	DET
ejpam-6087	517	25	best	good	ADJ
ejpam-6087	517	26	performance	performance	NOUN
ejpam-6087	517	27	among	among	ADP
ejpam-6087	517	28	the	the	DET
ejpam-6087	517	29	models	model	NOUN
ejpam-6087	517	30	evaluated	evaluate	VERB
ejpam-6087	517	31	.	.	PUNCT
ejpam-6087	518	1	in	in	ADP
ejpam-6087	518	2	contrast	contrast	NOUN
ejpam-6087	518	3	,	,	PUNCT
ejpam-6087	518	4	mlp	mlp	NOUN
ejpam-6087	518	5	exhibits	exhibit	VERB
ejpam-6087	518	6	the	the	DET
ejpam-6087	518	7	lowest	low	ADJ
ejpam-6087	518	8	accuracy	accuracy	NOUN
ejpam-6087	518	9	,	,	PUNCT
ejpam-6087	518	10	starting	start	VERB
ejpam-6087	518	11	at	at	ADP
ejpam-6087	518	12	49.23	49.23	NUM
ejpam-6087	518	13	%	%	NOUN
ejpam-6087	518	14	at	at	ADP
ejpam-6087	518	15	5	5	NUM
ejpam-6087	518	16	components	component	NOUN
ejpam-6087	518	17	and	and	CCONJ
ejpam-6087	518	18	slightly	slightly	ADV
ejpam-6087	518	19	improving	improve	VERB
ejpam-6087	518	20	to	to	ADP
ejpam-6087	518	21	52.40	52.40	NUM
ejpam-6087	518	22	%	%	NOUN
ejpam-6087	518	23	at	at	ADP
ejpam-6087	518	24	30	30	NUM
ejpam-6087	518	25	components	component	NOUN
ejpam-6087	518	26	.	.	PUNCT
ejpam-6087	519	1	knn	knn	PROPN
ejpam-6087	519	2	and	and	CCONJ
ejpam-6087	519	3	svm	svm	PROPN
ejpam-6087	519	4	maintain	maintain	VERB
ejpam-6087	519	5	relatively	relatively	ADV
ejpam-6087	519	6	stable	stable	ADJ
ejpam-6087	519	7	accuracies	accuracy	NOUN
ejpam-6087	519	8	around	around	ADP
ejpam-6087	519	9	58.67	58.67	NUM
ejpam-6087	519	10	%	%	NOUN
ejpam-6087	519	11	and	and	CCONJ
ejpam-6087	519	12	53.50	53.50	NUM
ejpam-6087	519	13	%	%	NOUN
ejpam-6087	519	14	,	,	PUNCT
ejpam-6087	519	15	respectively	respectively	ADV
ejpam-6087	519	16	,	,	PUNCT
ejpam-6087	519	17	while	while	SCONJ
ejpam-6087	519	18	lr	lr	NOUN
ejpam-6087	519	19	shows	show	VERB
ejpam-6087	519	20	a	a	DET
ejpam-6087	519	21	gradual	gradual	ADJ
ejpam-6087	519	22	increase	increase	NOUN
ejpam-6087	519	23	from	from	ADP
ejpam-6087	519	24	54.17	54.17	NUM
ejpam-6087	519	25	%	%	NOUN
ejpam-6087	519	26	to	to	ADP
ejpam-6087	519	27	58.30	58.30	NUM
ejpam-6087	519	28	%	%	NOUN
ejpam-6087	519	29	.	.	PUNCT
ejpam-6087	520	1	dt	dt	PUNCT
ejpam-6087	520	2	demonstrates	demonstrate	VERB
ejpam-6087	520	3	relatively	relatively	ADV
ejpam-6087	520	4	lower	low	ADJ
ejpam-6087	520	5	performance	performance	NOUN
ejpam-6087	520	6	,	,	PUNCT
ejpam-6087	520	7	with	with	ADP
ejpam-6087	520	8	accuracy	accuracy	NOUN
ejpam-6087	520	9	values	value	NOUN
ejpam-6087	520	10	ranging	range	VERB
ejpam-6087	520	11	from	from	ADP
ejpam-6087	520	12	58.58	58.58	NUM
ejpam-6087	520	13	%	%	NOUN
ejpam-6087	520	14	to	to	ADP
ejpam-6087	520	15	56.07	56.07	NUM
ejpam-6087	520	16	%	%	NOUN
ejpam-6087	520	17	.	.	PUNCT
ejpam-6087	521	1	overall	overall	ADV
ejpam-6087	521	2	,	,	PUNCT
ejpam-6087	521	3	the	the	DET
ejpam-6087	521	4	use	use	NOUN
ejpam-6087	521	5	of	of	ADP
ejpam-6087	521	6	pca	pca	NOUN
ejpam-6087	521	7	for	for	ADP
ejpam-6087	521	8	dimensionality	dimensionality	NOUN
ejpam-6087	521	9	reduction	reduction	NOUN
ejpam-6087	521	10	results	result	NOUN
ejpam-6087	521	11	in	in	ADP
ejpam-6087	521	12	lower	low	ADJ
ejpam-6087	521	13	accuracy	accuracy	NOUN
ejpam-6087	521	14	compared	compare	VERB
ejpam-6087	521	15	to	to	ADP
ejpam-6087	521	16	other	other	ADJ
ejpam-6087	521	17	feature	feature	NOUN
ejpam-6087	521	18	selection	selection	NOUN
ejpam-6087	521	19	methods	method	NOUN
ejpam-6087	521	20	,	,	PUNCT
ejpam-6087	521	21	indicating	indicate	VERB
ejpam-6087	521	22	that	that	SCONJ
ejpam-6087	521	23	while	while	SCONJ
ejpam-6087	521	24	pca	pca	PROPN
ejpam-6087	521	25	may	may	AUX
ejpam-6087	521	26	reduce	reduce	VERB
ejpam-6087	521	27	dimensionality	dimensionality	NOUN
ejpam-6087	521	28	,	,	PUNCT
ejpam-6087	521	29	it	it	PRON
ejpam-6087	521	30	does	do	AUX
ejpam-6087	521	31	not	not	PART
ejpam-6087	521	32	necessarily	necessarily	ADV
ejpam-6087	521	33	enhance	enhance	VERB
ejpam-6087	521	34	model	model	NOUN
ejpam-6087	521	35	performance	performance	NOUN
ejpam-6087	521	36	as	as	ADV
ejpam-6087	521	37	effectively	effectively	ADV
ejpam-6087	521	38	as	as	ADP
ejpam-6087	521	39	methods	method	NOUN
ejpam-6087	521	40	like	like	ADP
ejpam-6087	521	41	rfe	rfe	NOUN
ejpam-6087	521	42	or	or	CCONJ
ejpam-6087	521	43	selectkbest	selectkb	ADJ
ejpam-6087	521	44	.	.	PUNCT
ejpam-6087	522	1	m.	m.	NOUN
ejpam-6087	522	2	al	al	PROPN
ejpam-6087	522	3	-	-	PUNCT
ejpam-6087	522	4	tarawneh	tarawneh	PROPN
ejpam-6087	522	5	et	et	PROPN
ejpam-6087	522	6	al	al	PROPN
ejpam-6087	522	7	.	.	PUNCT
ejpam-6087	522	8	/	/	SYM
ejpam-6087	522	9	eur	eur	PROPN
ejpam-6087	522	10	.	.	PUNCT
ejpam-6087	523	1	j.	j.	PROPN
ejpam-6087	523	2	pure	pure	PROPN
ejpam-6087	523	3	appl	appl	PROPN
ejpam-6087	523	4	.	.	PROPN
ejpam-6087	523	5	math	math	PROPN
ejpam-6087	523	6	,	,	PUNCT
ejpam-6087	523	7	18	18	NUM
ejpam-6087	523	8	(	(	PUNCT
ejpam-6087	523	9	2	2	NUM
ejpam-6087	523	10	)	)	PUNCT
ejpam-6087	523	11	(	(	PUNCT
ejpam-6087	523	12	2025	2025	NUM
ejpam-6087	523	13	)	)	PUNCT
ejpam-6087	523	14	,	,	PUNCT
ejpam-6087	523	15	6087	6087	NUM
ejpam-6087	523	16	28	28	NUM
ejpam-6087	523	17	of	of	ADP
ejpam-6087	523	18	39	39	NUM
ejpam-6087	523	19	5	5	NUM
ejpam-6087	523	20	10	10	NUM
ejpam-6087	523	21	15	15	NUM
ejpam-6087	523	22	20	20	NUM
ejpam-6087	523	23	25	25	NUM
ejpam-6087	523	24	30	30	NUM
ejpam-6087	523	25	number	number	NOUN
ejpam-6087	523	26	of	of	ADP
ejpam-6087	523	27	features	feature	NOUN
ejpam-6087	523	28	50	50	NUM
ejpam-6087	523	29	55	55	NUM
ejpam-6087	523	30	60	60	NUM
ejpam-6087	523	31	65	65	NUM
ejpam-6087	524	1	a	a	DET
ejpam-6087	524	2	cc	cc	X
ejpam-6087	524	3	ur	ur	INTJ
ejpam-6087	524	4	ac	ac	PROPN
ejpam-6087	524	5	y	y	PROPN
ejpam-6087	524	6	(	(	PUNCT
ejpam-6087	524	7	%	%	NOUN
ejpam-6087	524	8	)	)	PUNCT
ejpam-6087	524	9	lr	lr	PROPN
ejpam-6087	524	10	nb	nb	PROPN
ejpam-6087	524	11	knn	knn	PROPN
ejpam-6087	524	12	svm	svm	PROPN
ejpam-6087	524	13	mlp	mlp	PROPN
ejpam-6087	524	14	dt	dt	PROPN
ejpam-6087	524	15	rf	rf	ADP
ejpam-6087	524	16	a	a	DET
ejpam-6087	524	17	accuracy	accuracy	NOUN
ejpam-6087	524	18	5	5	NUM
ejpam-6087	524	19	10	10	NUM
ejpam-6087	524	20	15	15	NUM
ejpam-6087	524	21	20	20	NUM
ejpam-6087	524	22	25	25	NUM
ejpam-6087	524	23	30	30	NUM
ejpam-6087	524	24	number	number	NOUN
ejpam-6087	524	25	of	of	ADP
ejpam-6087	524	26	features	feature	NOUN
ejpam-6087	524	27	45	45	NUM
ejpam-6087	524	28	50	50	NUM
ejpam-6087	524	29	55	55	NUM
ejpam-6087	524	30	60	60	NUM
ejpam-6087	524	31	65	65	NUM
ejpam-6087	524	32	p	p	NOUN
ejpam-6087	524	33	re	re	ADP
ejpam-6087	524	34	ci	ci	PROPN
ejpam-6087	524	35	si	si	X
ejpam-6087	524	36	on	on	ADP
ejpam-6087	524	37	(	(	PUNCT
ejpam-6087	524	38	%	%	INTJ
ejpam-6087	524	39	)	)	PUNCT
ejpam-6087	524	40	lr	lr	PROPN
ejpam-6087	524	41	nb	nb	PROPN
ejpam-6087	524	42	knn	knn	PROPN
ejpam-6087	524	43	svm	svm	PROPN
ejpam-6087	524	44	mlp	mlp	PROPN
ejpam-6087	524	45	dt	dt	PROPN
ejpam-6087	524	46	rf	rf	PROPN
ejpam-6087	524	47	b	b	PROPN
ejpam-6087	524	48	precision	precision	NOUN
ejpam-6087	524	49	5	5	NUM
ejpam-6087	524	50	10	10	NUM
ejpam-6087	524	51	15	15	NUM
ejpam-6087	524	52	20	20	NUM
ejpam-6087	524	53	25	25	NUM
ejpam-6087	524	54	30	30	NUM
ejpam-6087	524	55	number	number	NOUN
ejpam-6087	524	56	of	of	ADP
ejpam-6087	524	57	features	feature	NOUN
ejpam-6087	524	58	50	50	NUM
ejpam-6087	524	59	55	55	NUM
ejpam-6087	524	60	60	60	NUM
ejpam-6087	524	61	65	65	NUM
ejpam-6087	524	62	r	r	NOUN
ejpam-6087	524	63	ec	ec	PROPN
ejpam-6087	524	64	al	al	PROPN
ejpam-6087	524	65	l	l	PROPN
ejpam-6087	524	66	(	(	PUNCT
ejpam-6087	524	67	%	%	INTJ
ejpam-6087	524	68	)	)	PUNCT
ejpam-6087	524	69	lr	lr	PROPN
ejpam-6087	524	70	nb	nb	PROPN
ejpam-6087	524	71	knn	knn	PROPN
ejpam-6087	524	72	svm	svm	PROPN
ejpam-6087	524	73	mlp	mlp	PROPN
ejpam-6087	524	74	dt	dt	PROPN
ejpam-6087	524	75	rf	rf	PROPN
ejpam-6087	524	76	c	c	NOUN
ejpam-6087	524	77	recall	recall	VERB
ejpam-6087	524	78	5	5	NUM
ejpam-6087	524	79	10	10	NUM
ejpam-6087	524	80	15	15	NUM
ejpam-6087	524	81	20	20	NUM
ejpam-6087	524	82	25	25	NUM
ejpam-6087	524	83	30	30	NUM
ejpam-6087	524	84	number	number	NOUN
ejpam-6087	524	85	of	of	ADP
ejpam-6087	524	86	features	feature	NOUN
ejpam-6087	525	1	35	35	NUM
ejpam-6087	525	2	40	40	NUM
ejpam-6087	525	3	45	45	NUM
ejpam-6087	525	4	50	50	NUM
ejpam-6087	525	5	55	55	NUM
ejpam-6087	525	6	60	60	NUM
ejpam-6087	525	7	65	65	NUM
ejpam-6087	525	8	f	f	NOUN
ejpam-6087	525	9	-1	-1	PROPN
ejpam-6087	525	10	s	s	X
ejpam-6087	525	11	co	co	VERB
ejpam-6087	525	12	re	re	X
ejpam-6087	525	13	(	(	PUNCT
ejpam-6087	525	14	%	%	INTJ
ejpam-6087	525	15	)	)	PUNCT
ejpam-6087	525	16	lr	lr	PROPN
ejpam-6087	525	17	nb	nb	PROPN
ejpam-6087	525	18	knn	knn	PROPN
ejpam-6087	525	19	svm	svm	PROPN
ejpam-6087	525	20	mlp	mlp	PROPN
ejpam-6087	525	21	dt	dt	PROPN
ejpam-6087	525	22	rf	rf	PROPN
ejpam-6087	525	23	d	d	PROPN
ejpam-6087	525	24	f1	f1	PROPN
ejpam-6087	525	25	score	score	NOUN
ejpam-6087	525	26	5	5	NUM
ejpam-6087	525	27	10	10	NUM
ejpam-6087	525	28	15	15	NUM
ejpam-6087	525	29	20	20	NUM
ejpam-6087	525	30	25	25	NUM
ejpam-6087	525	31	30	30	NUM
ejpam-6087	525	32	number	number	NOUN
ejpam-6087	525	33	of	of	ADP
ejpam-6087	525	34	features	feature	NOUN
ejpam-6087	525	35	45	45	NUM
ejpam-6087	525	36	50	50	NUM
ejpam-6087	525	37	55	55	NUM
ejpam-6087	525	38	60	60	NUM
ejpam-6087	525	39	65	65	NUM
ejpam-6087	525	40	70	70	NUM
ejpam-6087	525	41	75	75	NUM
ejpam-6087	525	42	a	a	DET
ejpam-6087	525	43	u	u	NOUN
ejpam-6087	525	44	c	c	NOUN
ejpam-6087	525	45	-r	-r	NOUN
ejpam-6087	525	46	o	o	NOUN
ejpam-6087	525	47	c	c	PROPN
ejpam-6087	525	48	(	(	PUNCT
ejpam-6087	525	49	%	%	NOUN
ejpam-6087	525	50	)	)	PUNCT
ejpam-6087	525	51	lr	lr	PROPN
ejpam-6087	525	52	nb	nb	PROPN
ejpam-6087	525	53	knn	knn	PROPN
ejpam-6087	525	54	svm	svm	PROPN
ejpam-6087	525	55	mlp	mlp	PROPN
ejpam-6087	525	56	dt	dt	PROPN
ejpam-6087	525	57	rf	rf	PROPN
ejpam-6087	525	58	e	e	PROPN
ejpam-6087	525	59	auc	auc	VERB
ejpam-6087	525	60	-	-	PUNCT
ejpam-6087	525	61	roc	roc	NOUN
ejpam-6087	525	62	figure	figure	NOUN
ejpam-6087	525	63	3	3	NUM
ejpam-6087	525	64	:	:	PUNCT
ejpam-6087	525	65	performance	performance	NOUN
ejpam-6087	525	66	metrics	metric	NOUN
ejpam-6087	525	67	of	of	ADP
ejpam-6087	525	68	various	various	ADJ
ejpam-6087	525	69	models	model	NOUN
ejpam-6087	525	70	under	under	ADP
ejpam-6087	525	71	pca	pca	PROPN
ejpam-6087	525	72	.	.	PUNCT
ejpam-6087	526	1	m.	m.	PROPN
ejpam-6087	526	2	al	al	PROPN
ejpam-6087	526	3	-	-	PUNCT
ejpam-6087	526	4	tarawneh	tarawneh	PROPN
ejpam-6087	526	5	et	et	PROPN
ejpam-6087	526	6	al	al	PROPN
ejpam-6087	526	7	.	.	PUNCT
ejpam-6087	526	8	/	/	SYM
ejpam-6087	526	9	eur	eur	PROPN
ejpam-6087	526	10	.	.	PUNCT
ejpam-6087	527	1	j.	j.	PROPN
ejpam-6087	527	2	pure	pure	PROPN
ejpam-6087	527	3	appl	appl	PROPN
ejpam-6087	527	4	.	.	PROPN
ejpam-6087	527	5	math	math	PROPN
ejpam-6087	527	6	,	,	PUNCT
ejpam-6087	527	7	18	18	NUM
ejpam-6087	527	8	(	(	PUNCT
ejpam-6087	527	9	2	2	NUM
ejpam-6087	527	10	)	)	PUNCT
ejpam-6087	527	11	(	(	PUNCT
ejpam-6087	527	12	2025	2025	NUM
ejpam-6087	527	13	)	)	PUNCT
ejpam-6087	527	14	,	,	PUNCT
ejpam-6087	527	15	6087	6087	NUM
ejpam-6087	527	16	29	29	NUM
ejpam-6087	527	17	of	of	ADP
ejpam-6087	527	18	39	39	NUM
ejpam-6087	527	19	figure	figure	NOUN
ejpam-6087	527	20	3b	3b	NOUN
ejpam-6087	527	21	illustrates	illustrate	VERB
ejpam-6087	527	22	the	the	DET
ejpam-6087	527	23	precision	precision	NOUN
ejpam-6087	527	24	of	of	ADP
ejpam-6087	527	25	various	various	ADJ
ejpam-6087	527	26	ml	ml	NOUN
ejpam-6087	527	27	models	model	NOUN
ejpam-6087	527	28	as	as	ADP
ejpam-6087	527	29	the	the	DET
ejpam-6087	527	30	number	number	NOUN
ejpam-6087	527	31	of	of	ADP
ejpam-6087	527	32	principal	principal	ADJ
ejpam-6087	527	33	components	component	NOUN
ejpam-6087	527	34	increases	increase	NOUN
ejpam-6087	527	35	from	from	ADP
ejpam-6087	527	36	5	5	NUM
ejpam-6087	527	37	to	to	PART
ejpam-6087	527	38	30	30	NUM
ejpam-6087	527	39	.	.	PUNCT
ejpam-6087	528	1	the	the	DET
ejpam-6087	528	2	precision	precision	NOUN
ejpam-6087	528	3	results	result	VERB
ejpam-6087	528	4	for	for	ADP
ejpam-6087	528	5	pca	pca	PROPN
ejpam-6087	528	6	are	be	AUX
ejpam-6087	528	7	generally	generally	ADV
ejpam-6087	528	8	low	low	ADJ
ejpam-6087	528	9	compared	compare	VERB
ejpam-6087	528	10	to	to	ADP
ejpam-6087	528	11	other	other	ADJ
ejpam-6087	528	12	feature	feature	NOUN
ejpam-6087	528	13	selection	selection	NOUN
ejpam-6087	528	14	methods	method	NOUN
ejpam-6087	528	15	,	,	PUNCT
ejpam-6087	528	16	highlighting	highlight	VERB
ejpam-6087	528	17	a	a	DET
ejpam-6087	528	18	notable	notable	ADJ
ejpam-6087	528	19	deficiency	deficiency	NOUN
ejpam-6087	528	20	in	in	ADP
ejpam-6087	528	21	performance	performance	NOUN
ejpam-6087	528	22	for	for	ADP
ejpam-6087	528	23	most	most	ADJ
ejpam-6087	528	24	models	model	NOUN
ejpam-6087	528	25	.	.	PUNCT
ejpam-6087	529	1	rf	rf	PRON
ejpam-6087	529	2	achieves	achieve	VERB
ejpam-6087	529	3	the	the	DET
ejpam-6087	529	4	highest	high	ADJ
ejpam-6087	529	5	precision	precision	NOUN
ejpam-6087	529	6	,	,	PUNCT
ejpam-6087	529	7	increasing	increase	VERB
ejpam-6087	529	8	from	from	ADP
ejpam-6087	529	9	61.34	61.34	NUM
ejpam-6087	529	10	%	%	NOUN
ejpam-6087	529	11	at	at	ADP
ejpam-6087	529	12	5	5	NUM
ejpam-6087	529	13	components	component	NOUN
ejpam-6087	529	14	to	to	ADP
ejpam-6087	529	15	64.60	64.60	NUM
ejpam-6087	529	16	%	%	NOUN
ejpam-6087	529	17	at	at	ADP
ejpam-6087	529	18	30	30	NUM
ejpam-6087	529	19	components	component	NOUN
ejpam-6087	529	20	,	,	PUNCT
ejpam-6087	529	21	indicating	indicate	VERB
ejpam-6087	529	22	that	that	SCONJ
ejpam-6087	529	23	it	it	PRON
ejpam-6087	529	24	is	be	AUX
ejpam-6087	529	25	the	the	DET
ejpam-6087	529	26	most	most	ADV
ejpam-6087	529	27	effective	effective	ADJ
ejpam-6087	529	28	model	model	NOUN
ejpam-6087	529	29	in	in	ADP
ejpam-6087	529	30	terms	term	NOUN
ejpam-6087	529	31	of	of	ADP
ejpam-6087	529	32	precision	precision	NOUN
ejpam-6087	529	33	under	under	ADP
ejpam-6087	529	34	pca	pca	PROPN
ejpam-6087	529	35	.	.	PUNCT
ejpam-6087	530	1	in	in	ADP
ejpam-6087	530	2	contrast	contrast	NOUN
ejpam-6087	530	3	,	,	PUNCT
ejpam-6087	530	4	mlp	mlp	NOUN
ejpam-6087	530	5	exhibits	exhibit	VERB
ejpam-6087	530	6	the	the	DET
ejpam-6087	530	7	lowest	low	ADJ
ejpam-6087	530	8	precision	precision	NOUN
ejpam-6087	530	9	,	,	PUNCT
ejpam-6087	530	10	with	with	ADP
ejpam-6087	530	11	a	a	DET
ejpam-6087	530	12	substantial	substantial	ADJ
ejpam-6087	530	13	drop	drop	NOUN
ejpam-6087	530	14	from	from	ADP
ejpam-6087	530	15	47.77	47.77	NUM
ejpam-6087	530	16	%	%	NOUN
ejpam-6087	530	17	at	at	ADP
ejpam-6087	530	18	5	5	NUM
ejpam-6087	530	19	components	component	NOUN
ejpam-6087	530	20	to	to	ADP
ejpam-6087	530	21	52.93	52.93	NUM
ejpam-6087	530	22	%	%	NOUN
ejpam-6087	530	23	at	at	ADP
ejpam-6087	530	24	30	30	NUM
ejpam-6087	530	25	components	component	NOUN
ejpam-6087	530	26	,	,	PUNCT
ejpam-6087	530	27	which	which	PRON
ejpam-6087	530	28	shows	show	VERB
ejpam-6087	530	29	that	that	SCONJ
ejpam-6087	530	30	this	this	DET
ejpam-6087	530	31	model	model	NOUN
ejpam-6087	530	32	performs	perform	VERB
ejpam-6087	530	33	poorly	poorly	ADV
ejpam-6087	530	34	in	in	ADP
ejpam-6087	530	35	terms	term	NOUN
ejpam-6087	530	36	of	of	ADP
ejpam-6087	530	37	precision	precision	NOUN
ejpam-6087	530	38	.	.	PUNCT
ejpam-6087	531	1	knn	knn	PROPN
ejpam-6087	531	2	and	and	CCONJ
ejpam-6087	531	3	svm	svm	PROPN
ejpam-6087	531	4	maintain	maintain	VERB
ejpam-6087	531	5	relatively	relatively	ADV
ejpam-6087	531	6	stable	stable	ADJ
ejpam-6087	531	7	precision	precision	NOUN
ejpam-6087	531	8	values	value	NOUN
ejpam-6087	531	9	around	around	ADP
ejpam-6087	531	10	54.78	54.78	NUM
ejpam-6087	531	11	%	%	NOUN
ejpam-6087	531	12	and	and	CCONJ
ejpam-6087	531	13	58.73	58.73	NUM
ejpam-6087	531	14	%	%	NOUN
ejpam-6087	531	15	,	,	PUNCT
ejpam-6087	531	16	respectively	respectively	ADV
ejpam-6087	531	17	.	.	PUNCT
ejpam-6087	532	1	lr	lr	NOUN
ejpam-6087	532	2	and	and	CCONJ
ejpam-6087	532	3	nb	nb	PROPN
ejpam-6087	532	4	show	show	VERB
ejpam-6087	532	5	a	a	DET
ejpam-6087	532	6	general	general	ADJ
ejpam-6087	532	7	increase	increase	NOUN
ejpam-6087	532	8	in	in	ADP
ejpam-6087	532	9	precision	precision	NOUN
ejpam-6087	532	10	as	as	SCONJ
ejpam-6087	532	11	the	the	DET
ejpam-6087	532	12	number	number	NOUN
ejpam-6087	532	13	of	of	ADP
ejpam-6087	532	14	components	component	NOUN
ejpam-6087	532	15	grows	grow	VERB
ejpam-6087	532	16	,	,	PUNCT
ejpam-6087	532	17	though	though	SCONJ
ejpam-6087	532	18	their	their	PRON
ejpam-6087	532	19	performance	performance	NOUN
ejpam-6087	532	20	remains	remain	VERB
ejpam-6087	532	21	lower	low	ADJ
ejpam-6087	532	22	than	than	ADP
ejpam-6087	532	23	that	that	PRON
ejpam-6087	532	24	of	of	ADP
ejpam-6087	532	25	rf	rf	NOUN
ejpam-6087	532	26	.	.	PUNCT
ejpam-6087	533	1	specifically	specifically	ADV
ejpam-6087	533	2	,	,	PUNCT
ejpam-6087	533	3	lr	lr	PROPN
ejpam-6087	533	4	improves	improve	VERB
ejpam-6087	533	5	from	from	ADP
ejpam-6087	533	6	54.09	54.09	NUM
ejpam-6087	533	7	%	%	NOUN
ejpam-6087	533	8	to	to	ADP
ejpam-6087	533	9	58.30	58.30	NUM
ejpam-6087	533	10	%	%	NOUN
ejpam-6087	533	11	,	,	PUNCT
ejpam-6087	533	12	and	and	CCONJ
ejpam-6087	533	13	nb	nb	PROPN
ejpam-6087	533	14	shows	show	VERB
ejpam-6087	533	15	a	a	DET
ejpam-6087	533	16	slight	slight	ADJ
ejpam-6087	533	17	increase	increase	NOUN
ejpam-6087	533	18	from	from	ADP
ejpam-6087	533	19	55.24	55.24	NUM
ejpam-6087	533	20	%	%	NOUN
ejpam-6087	533	21	to	to	ADP
ejpam-6087	533	22	56.26	56.26	NUM
ejpam-6087	533	23	%	%	NOUN
ejpam-6087	533	24	.	.	PUNCT
ejpam-6087	534	1	overall	overall	ADV
ejpam-6087	534	2	,	,	PUNCT
ejpam-6087	534	3	the	the	DET
ejpam-6087	534	4	precision	precision	NOUN
ejpam-6087	534	5	results	result	VERB
ejpam-6087	534	6	underscore	underscore	VERB
ejpam-6087	534	7	that	that	PRON
ejpam-6087	534	8	pca	pca	PROPN
ejpam-6087	534	9	’s	’s	PART
ejpam-6087	534	10	dimensionality	dimensionality	NOUN
ejpam-6087	534	11	reduction	reduction	NOUN
ejpam-6087	534	12	approach	approach	NOUN
ejpam-6087	534	13	leads	lead	VERB
ejpam-6087	534	14	to	to	ADP
ejpam-6087	534	15	relatively	relatively	ADV
ejpam-6087	534	16	poor	poor	ADJ
ejpam-6087	534	17	performance	performance	NOUN
ejpam-6087	534	18	across	across	ADP
ejpam-6087	534	19	the	the	DET
ejpam-6087	534	20	models	model	NOUN
ejpam-6087	534	21	,	,	PUNCT
ejpam-6087	534	22	with	with	ADP
ejpam-6087	534	23	rf	rf	PRON
ejpam-6087	534	24	being	be	AUX
ejpam-6087	534	25	the	the	DET
ejpam-6087	534	26	exception	exception	NOUN
ejpam-6087	534	27	and	and	CCONJ
ejpam-6087	534	28	mlp	mlp	NOUN
ejpam-6087	534	29	showing	show	VERB
ejpam-6087	534	30	significant	significant	ADJ
ejpam-6087	534	31	performance	performance	NOUN
ejpam-6087	534	32	issues	issue	NOUN
ejpam-6087	534	33	.	.	PUNCT
ejpam-6087	535	1	figure	figure	VERB
ejpam-6087	535	2	3c	3c	NUM
ejpam-6087	535	3	shows	show	VERB
ejpam-6087	535	4	the	the	DET
ejpam-6087	535	5	recall	recall	NOUN
ejpam-6087	535	6	values	value	NOUN
ejpam-6087	535	7	for	for	ADP
ejpam-6087	535	8	different	different	ADJ
ejpam-6087	535	9	mlp	mlp	NOUN
ejpam-6087	535	10	models	model	NOUN
ejpam-6087	535	11	as	as	ADP
ejpam-6087	535	12	the	the	DET
ejpam-6087	535	13	number	number	NOUN
ejpam-6087	535	14	of	of	ADP
ejpam-6087	535	15	principal	principal	ADJ
ejpam-6087	535	16	components	component	NOUN
ejpam-6087	535	17	increases	increase	NOUN
ejpam-6087	535	18	from	from	ADP
ejpam-6087	535	19	5	5	NUM
ejpam-6087	535	20	to	to	PART
ejpam-6087	535	21	30	30	NUM
ejpam-6087	535	22	.	.	PUNCT
ejpam-6087	536	1	the	the	DET
ejpam-6087	536	2	recall	recall	NOUN
ejpam-6087	536	3	values	value	NOUN
ejpam-6087	536	4	,	,	PUNCT
ejpam-6087	536	5	which	which	PRON
ejpam-6087	536	6	represent	represent	VERB
ejpam-6087	536	7	the	the	DET
ejpam-6087	536	8	proportion	proportion	NOUN
ejpam-6087	536	9	of	of	ADP
ejpam-6087	536	10	true	true	ADJ
ejpam-6087	536	11	positive	positive	ADJ
ejpam-6087	536	12	instances	instance	NOUN
ejpam-6087	536	13	identified	identify	VERB
ejpam-6087	536	14	by	by	ADP
ejpam-6087	536	15	each	each	DET
ejpam-6087	536	16	model	model	NOUN
ejpam-6087	536	17	,	,	PUNCT
ejpam-6087	536	18	reveal	reveal	VERB
ejpam-6087	536	19	a	a	DET
ejpam-6087	536	20	generally	generally	ADV
ejpam-6087	536	21	low	low	ADJ
ejpam-6087	536	22	performance	performance	NOUN
ejpam-6087	536	23	across	across	ADP
ejpam-6087	536	24	the	the	DET
ejpam-6087	536	25	models	model	NOUN
ejpam-6087	536	26	compared	compare	VERB
ejpam-6087	536	27	to	to	ADP
ejpam-6087	536	28	other	other	ADJ
ejpam-6087	536	29	feature	feature	NOUN
ejpam-6087	536	30	selection	selection	NOUN
ejpam-6087	536	31	methods	method	NOUN
ejpam-6087	536	32	.	.	PUNCT
ejpam-6087	537	1	rf	rf	NOUN
ejpam-6087	537	2	exhibits	exhibit	VERB
ejpam-6087	537	3	the	the	DET
ejpam-6087	537	4	highest	high	ADJ
ejpam-6087	537	5	recall	recall	NOUN
ejpam-6087	537	6	,	,	PUNCT
ejpam-6087	537	7	increasing	increase	VERB
ejpam-6087	537	8	from	from	ADP
ejpam-6087	537	9	61.34	61.34	NUM
ejpam-6087	537	10	%	%	NOUN
ejpam-6087	537	11	at	at	ADP
ejpam-6087	537	12	5	5	NUM
ejpam-6087	537	13	components	component	NOUN
ejpam-6087	537	14	to	to	ADP
ejpam-6087	537	15	64.59	64.59	NUM
ejpam-6087	537	16	%	%	NOUN
ejpam-6087	537	17	at	at	ADP
ejpam-6087	537	18	30	30	NUM
ejpam-6087	537	19	components	component	NOUN
ejpam-6087	537	20	.	.	PUNCT
ejpam-6087	538	1	this	this	PRON
ejpam-6087	538	2	indicates	indicate	VERB
ejpam-6087	538	3	that	that	SCONJ
ejpam-6087	538	4	rf	rf	PRON
ejpam-6087	538	5	performs	perform	VERB
ejpam-6087	538	6	the	the	DET
ejpam-6087	538	7	best	good	ADJ
ejpam-6087	538	8	in	in	ADP
ejpam-6087	538	9	identifying	identify	VERB
ejpam-6087	538	10	true	true	ADJ
ejpam-6087	538	11	positive	positive	ADJ
ejpam-6087	538	12	instances	instance	NOUN
ejpam-6087	538	13	as	as	SCONJ
ejpam-6087	538	14	the	the	DET
ejpam-6087	538	15	number	number	NOUN
ejpam-6087	538	16	of	of	ADP
ejpam-6087	538	17	principal	principal	ADJ
ejpam-6087	538	18	components	component	NOUN
ejpam-6087	538	19	grows	grow	VERB
ejpam-6087	538	20	.	.	PUNCT
ejpam-6087	539	1	knn	knn	PROPN
ejpam-6087	539	2	shows	show	VERB
ejpam-6087	539	3	stable	stable	ADJ
ejpam-6087	539	4	recall	recall	NOUN
ejpam-6087	539	5	values	value	NOUN
ejpam-6087	539	6	,	,	PUNCT
ejpam-6087	539	7	remaining	remain	VERB
ejpam-6087	539	8	around	around	ADP
ejpam-6087	539	9	58.67	58.67	NUM
ejpam-6087	539	10	%	%	NOUN
ejpam-6087	539	11	to	to	ADP
ejpam-6087	539	12	58.73	58.73	NUM
ejpam-6087	539	13	%	%	NOUN
ejpam-6087	539	14	throughout	throughout	ADP
ejpam-6087	539	15	the	the	DET
ejpam-6087	539	16	range	range	NOUN
ejpam-6087	539	17	of	of	ADP
ejpam-6087	539	18	principal	principal	ADJ
ejpam-6087	539	19	components	component	NOUN
ejpam-6087	539	20	,	,	PUNCT
ejpam-6087	539	21	reflecting	reflect	VERB
ejpam-6087	539	22	consistent	consistent	ADJ
ejpam-6087	539	23	performance	performance	NOUN
ejpam-6087	539	24	.	.	PUNCT
ejpam-6087	540	1	in	in	ADP
ejpam-6087	540	2	contrast	contrast	NOUN
ejpam-6087	540	3	,	,	PUNCT
ejpam-6087	540	4	mlp	mlp	PROPN
ejpam-6087	540	5	has	have	VERB
ejpam-6087	540	6	relatively	relatively	ADV
ejpam-6087	540	7	poor	poor	ADJ
ejpam-6087	540	8	recall	recall	NOUN
ejpam-6087	540	9	performance	performance	NOUN
ejpam-6087	540	10	,	,	PUNCT
ejpam-6087	540	11	with	with	ADP
ejpam-6087	540	12	values	value	NOUN
ejpam-6087	540	13	starting	start	VERB
ejpam-6087	540	14	at	at	ADP
ejpam-6087	540	15	49.23	49.23	NUM
ejpam-6087	540	16	%	%	NOUN
ejpam-6087	540	17	at	at	ADP
ejpam-6087	540	18	5	5	NUM
ejpam-6087	540	19	components	component	NOUN
ejpam-6087	540	20	and	and	CCONJ
ejpam-6087	540	21	increasing	increase	VERB
ejpam-6087	540	22	to	to	ADP
ejpam-6087	540	23	52.40	52.40	NUM
ejpam-6087	540	24	%	%	NOUN
ejpam-6087	540	25	at	at	ADP
ejpam-6087	540	26	30	30	NUM
ejpam-6087	540	27	components	component	NOUN
ejpam-6087	540	28	.	.	PUNCT
ejpam-6087	541	1	lr	lr	NOUN
ejpam-6087	541	2	and	and	CCONJ
ejpam-6087	541	3	nb	nb	PROPN
ejpam-6087	541	4	demonstrate	demonstrate	VERB
ejpam-6087	541	5	moderate	moderate	ADJ
ejpam-6087	541	6	recall	recall	NOUN
ejpam-6087	541	7	improvements	improvement	NOUN
ejpam-6087	541	8	,	,	PUNCT
ejpam-6087	541	9	with	with	ADP
ejpam-6087	541	10	lr	lr	NOUN
ejpam-6087	541	11	increasing	increase	VERB
ejpam-6087	541	12	from	from	ADP
ejpam-6087	541	13	54.17	54.17	NUM
ejpam-6087	541	14	%	%	NOUN
ejpam-6087	541	15	to	to	ADP
ejpam-6087	541	16	58.30	58.30	NUM
ejpam-6087	541	17	%	%	NOUN
ejpam-6087	541	18	and	and	CCONJ
ejpam-6087	541	19	nb	nb	INTJ
ejpam-6087	541	20	increasing	increase	VERB
ejpam-6087	541	21	from	from	ADP
ejpam-6087	541	22	49.76	49.76	NUM
ejpam-6087	541	23	%	%	NOUN
ejpam-6087	541	24	to	to	ADP
ejpam-6087	541	25	52.01	52.01	NUM
ejpam-6087	541	26	%	%	NOUN
ejpam-6087	541	27	.	.	PUNCT
ejpam-6087	542	1	svm	svm	PROPN
ejpam-6087	542	2	maintains	maintain	VERB
ejpam-6087	542	3	low	low	ADJ
ejpam-6087	542	4	and	and	CCONJ
ejpam-6087	542	5	stable	stable	ADJ
ejpam-6087	542	6	recall	recall	NOUN
ejpam-6087	542	7	values	value	NOUN
ejpam-6087	542	8	,	,	PUNCT
ejpam-6087	542	9	ranging	range	VERB
ejpam-6087	542	10	from	from	ADP
ejpam-6087	542	11	53.49	53.49	NUM
ejpam-6087	542	12	%	%	NOUN
ejpam-6087	542	13	at	at	ADP
ejpam-6087	542	14	5	5	NUM
ejpam-6087	542	15	components	component	NOUN
ejpam-6087	542	16	to	to	ADP
ejpam-6087	542	17	53.49	53.49	NUM
ejpam-6087	542	18	%	%	NOUN
ejpam-6087	542	19	at	at	ADP
ejpam-6087	542	20	30	30	NUM
ejpam-6087	542	21	components	component	NOUN
ejpam-6087	542	22	,	,	PUNCT
ejpam-6087	542	23	showing	show	VERB
ejpam-6087	542	24	minimal	minimal	ADJ
ejpam-6087	542	25	improvement	improvement	NOUN
ejpam-6087	542	26	as	as	ADP
ejpam-6087	542	27	the	the	DET
ejpam-6087	542	28	number	number	NOUN
ejpam-6087	542	29	of	of	ADP
ejpam-6087	542	30	components	component	NOUN
ejpam-6087	542	31	increases	increase	NOUN
ejpam-6087	542	32	.	.	PUNCT
ejpam-6087	543	1	dt	dt	PROPN
ejpam-6087	543	2	shows	show	VERB
ejpam-6087	543	3	a	a	DET
ejpam-6087	543	4	gradual	gradual	ADJ
ejpam-6087	543	5	increase	increase	NOUN
ejpam-6087	543	6	in	in	ADP
ejpam-6087	543	7	recall	recall	NOUN
ejpam-6087	543	8	from	from	ADP
ejpam-6087	543	9	58.58	58.58	NUM
ejpam-6087	543	10	%	%	NOUN
ejpam-6087	543	11	to	to	ADP
ejpam-6087	543	12	56.07	56.07	NUM
ejpam-6087	543	13	%	%	NOUN
ejpam-6087	543	14	,	,	PUNCT
ejpam-6087	543	15	reflecting	reflect	VERB
ejpam-6087	543	16	its	its	PRON
ejpam-6087	543	17	generally	generally	ADV
ejpam-6087	543	18	strong	strong	ADJ
ejpam-6087	543	19	performance	performance	NOUN
ejpam-6087	543	20	compared	compare	VERB
ejpam-6087	543	21	to	to	ADP
ejpam-6087	543	22	most	most	ADJ
ejpam-6087	543	23	models	model	NOUN
ejpam-6087	543	24	.	.	PUNCT
ejpam-6087	544	1	in	in	ADP
ejpam-6087	544	2	general	general	ADJ
ejpam-6087	544	3	,	,	PUNCT
ejpam-6087	544	4	the	the	DET
ejpam-6087	544	5	recall	recall	NOUN
ejpam-6087	544	6	values	value	NOUN
ejpam-6087	544	7	are	be	AUX
ejpam-6087	544	8	relatively	relatively	ADV
ejpam-6087	544	9	low	low	ADJ
ejpam-6087	544	10	in	in	ADP
ejpam-6087	544	11	all	all	DET
ejpam-6087	544	12	models	model	NOUN
ejpam-6087	544	13	compared	compare	VERB
ejpam-6087	544	14	to	to	ADP
ejpam-6087	544	15	other	other	ADJ
ejpam-6087	544	16	feature	feature	NOUN
ejpam-6087	544	17	selection	selection	NOUN
ejpam-6087	544	18	methods	method	NOUN
ejpam-6087	544	19	,	,	PUNCT
ejpam-6087	544	20	indicating	indicate	VERB
ejpam-6087	544	21	that	that	SCONJ
ejpam-6087	544	22	pca	pca	PROPN
ejpam-6087	544	23	does	do	AUX
ejpam-6087	544	24	not	not	PART
ejpam-6087	544	25	significantly	significantly	ADV
ejpam-6087	544	26	improve	improve	VERB
ejpam-6087	544	27	the	the	DET
ejpam-6087	544	28	ability	ability	NOUN
ejpam-6087	544	29	of	of	ADP
ejpam-6087	544	30	these	these	DET
ejpam-6087	544	31	models	model	NOUN
ejpam-6087	544	32	to	to	PART
ejpam-6087	544	33	identify	identify	VERB
ejpam-6087	544	34	true	true	ADJ
ejpam-6087	544	35	positives	positive	NOUN
ejpam-6087	544	36	.	.	PUNCT
ejpam-6087	545	1	random	random	ADJ
ejpam-6087	545	2	forest	forest	NOUN
ejpam-6087	545	3	is	be	AUX
ejpam-6087	545	4	the	the	DET
ejpam-6087	545	5	most	most	ADV
ejpam-6087	545	6	effective	effective	ADJ
ejpam-6087	545	7	in	in	ADP
ejpam-6087	545	8	terms	term	NOUN
ejpam-6087	545	9	of	of	ADP
ejpam-6087	545	10	recall	recall	NOUN
ejpam-6087	545	11	,	,	PUNCT
ejpam-6087	545	12	whereas	whereas	SCONJ
ejpam-6087	545	13	other	other	ADJ
ejpam-6087	545	14	models	model	NOUN
ejpam-6087	545	15	such	such	ADJ
ejpam-6087	545	16	as	as	ADP
ejpam-6087	545	17	svm	svm	PROPN
ejpam-6087	545	18	and	and	CCONJ
ejpam-6087	545	19	mlp	mlp	NOUN
ejpam-6087	545	20	show	show	VERB
ejpam-6087	545	21	relatively	relatively	ADV
ejpam-6087	545	22	poor	poor	ADJ
ejpam-6087	545	23	performance	performance	NOUN
ejpam-6087	545	24	.	.	PUNCT
ejpam-6087	546	1	figure	figure	NOUN
ejpam-6087	546	2	3d	3d	PROPN
ejpam-6087	546	3	illustrates	illustrate	VERB
ejpam-6087	546	4	the	the	DET
ejpam-6087	546	5	f-1	f-1	NOUN
ejpam-6087	546	6	scores	score	NOUN
ejpam-6087	546	7	for	for	ADP
ejpam-6087	546	8	various	various	ADJ
ejpam-6087	546	9	ml	ml	NOUN
ejpam-6087	546	10	models	model	NOUN
ejpam-6087	546	11	as	as	ADP
ejpam-6087	546	12	the	the	DET
ejpam-6087	546	13	number	number	NOUN
ejpam-6087	546	14	of	of	ADP
ejpam-6087	546	15	principal	principal	ADJ
ejpam-6087	546	16	components	component	NOUN
ejpam-6087	546	17	increases	increase	NOUN
ejpam-6087	546	18	from	from	ADP
ejpam-6087	546	19	5	5	NUM
ejpam-6087	546	20	to	to	PART
ejpam-6087	546	21	30	30	NUM
ejpam-6087	546	22	.	.	PUNCT
ejpam-6087	547	1	the	the	DET
ejpam-6087	547	2	f-1	f-1	NOUN
ejpam-6087	547	3	score	score	NOUN
ejpam-6087	547	4	,	,	PUNCT
ejpam-6087	547	5	which	which	PRON
ejpam-6087	547	6	balances	balance	VERB
ejpam-6087	547	7	precision	precision	NOUN
ejpam-6087	547	8	and	and	CCONJ
ejpam-6087	547	9	recall	recall	NOUN
ejpam-6087	547	10	,	,	PUNCT
ejpam-6087	547	11	highlights	highlight	VERB
ejpam-6087	547	12	the	the	DET
ejpam-6087	547	13	trade	trade	NOUN
ejpam-6087	547	14	-	-	PUNCT
ejpam-6087	547	15	offs	off	NOUN
ejpam-6087	547	16	between	between	ADP
ejpam-6087	547	17	these	these	DET
ejpam-6087	547	18	two	two	NUM
ejpam-6087	547	19	metrics	metric	NOUN
ejpam-6087	547	20	.	.	PUNCT
ejpam-6087	548	1	rf	rf	VERB
ejpam-6087	548	2	consistently	consistently	ADV
ejpam-6087	548	3	achieves	achieve	VERB
ejpam-6087	548	4	the	the	DET
ejpam-6087	548	5	highest	high	ADJ
ejpam-6087	548	6	f-1	f-1	NOUN
ejpam-6087	548	7	scores	score	NOUN
ejpam-6087	548	8	in	in	ADP
ejpam-6087	548	9	all	all	DET
ejpam-6087	548	10	principal	principal	ADJ
ejpam-6087	548	11	components	component	NOUN
ejpam-6087	548	12	,	,	PUNCT
ejpam-6087	548	13	starting	start	VERB
ejpam-6087	548	14	at	at	ADP
ejpam-6087	548	15	61.34	61.34	NUM
ejpam-6087	548	16	%	%	NOUN
ejpam-6087	548	17	with	with	ADP
ejpam-6087	548	18	5	5	NUM
ejpam-6087	548	19	components	component	NOUN
ejpam-6087	548	20	and	and	CCONJ
ejpam-6087	548	21	increasing	increase	VERB
ejpam-6087	548	22	to	to	ADP
ejpam-6087	548	23	64.59	64.59	NUM
ejpam-6087	548	24	%	%	NOUN
ejpam-6087	548	25	with	with	ADP
ejpam-6087	548	26	30	30	NUM
ejpam-6087	548	27	components	component	NOUN
ejpam-6087	548	28	.	.	PUNCT
ejpam-6087	549	1	this	this	DET
ejpam-6087	549	2	model	model	NOUN
ejpam-6087	549	3	demonstrates	demonstrate	VERB
ejpam-6087	549	4	high	high	ADJ
ejpam-6087	549	5	precision	precision	NOUN
ejpam-6087	549	6	and	and	CCONJ
ejpam-6087	549	7	recall	recall	NOUN
ejpam-6087	549	8	,	,	PUNCT
ejpam-6087	549	9	resulting	result	VERB
ejpam-6087	549	10	in	in	ADP
ejpam-6087	549	11	the	the	DET
ejpam-6087	549	12	most	most	ADV
ejpam-6087	549	13	balanced	balanced	ADJ
ejpam-6087	549	14	performance	performance	NOUN
ejpam-6087	549	15	among	among	ADP
ejpam-6087	549	16	the	the	DET
ejpam-6087	549	17	models	model	NOUN
ejpam-6087	549	18	tested	test	VERB
ejpam-6087	549	19	.	.	PUNCT
ejpam-6087	550	1	in	in	ADP
ejpam-6087	550	2	contrast	contrast	NOUN
ejpam-6087	550	3	,	,	PUNCT
ejpam-6087	550	4	knn	knn	PROPN
ejpam-6087	550	5	shows	show	VERB
ejpam-6087	550	6	a	a	DET
ejpam-6087	550	7	stable	stable	ADJ
ejpam-6087	550	8	f-1	f-1	NOUN
ejpam-6087	550	9	score	score	NOUN
ejpam-6087	550	10	,	,	PUNCT
ejpam-6087	550	11	remaining	remain	VERB
ejpam-6087	550	12	between	between	ADP
ejpam-6087	550	13	58.67	58.67	NUM
ejpam-6087	550	14	%	%	NOUN
ejpam-6087	550	15	and	and	CCONJ
ejpam-6087	550	16	58.74	58.74	NUM
ejpam-6087	550	17	%	%	NOUN
ejpam-6087	550	18	,	,	PUNCT
ejpam-6087	550	19	indicating	indicate	VERB
ejpam-6087	550	20	consistent	consistent	ADJ
ejpam-6087	550	21	performance	performance	NOUN
ejpam-6087	550	22	but	but	CCONJ
ejpam-6087	550	23	at	at	ADP
ejpam-6087	550	24	a	a	DET
ejpam-6087	550	25	lower	low	ADJ
ejpam-6087	550	26	level	level	NOUN
ejpam-6087	550	27	compared	compare	VERB
ejpam-6087	550	28	to	to	ADP
ejpam-6087	550	29	random	random	ADJ
ejpam-6087	550	30	forest	forest	NOUN
ejpam-6087	550	31	.	.	PUNCT
ejpam-6087	551	1	dt	dt	PUNCT
ejpam-6087	551	2	exhibits	exhibit	VERB
ejpam-6087	551	3	moderate	moderate	ADJ
ejpam-6087	551	4	performance	performance	NOUN
ejpam-6087	551	5	in	in	ADP
ejpam-6087	551	6	terms	term	NOUN
ejpam-6087	551	7	of	of	ADP
ejpam-6087	551	8	the	the	DET
ejpam-6087	551	9	f-1	f-1	NOUN
ejpam-6087	551	10	score	score	NOUN
ejpam-6087	551	11	,	,	PUNCT
ejpam-6087	551	12	with	with	ADP
ejpam-6087	551	13	values	value	NOUN
ejpam-6087	551	14	starting	start	VERB
ejpam-6087	551	15	at	at	ADP
ejpam-6087	551	16	58.57	58.57	NUM
ejpam-6087	551	17	%	%	NOUN
ejpam-6087	551	18	at	at	ADP
ejpam-6087	551	19	5	5	NUM
ejpam-6087	551	20	components	component	NOUN
ejpam-6087	551	21	and	and	CCONJ
ejpam-6087	551	22	reaching	reach	VERB
ejpam-6087	551	23	56.07	56.07	NUM
ejpam-6087	551	24	%	%	NOUN
ejpam-6087	551	25	at	at	ADP
ejpam-6087	551	26	30	30	NUM
ejpam-6087	551	27	components	component	NOUN
ejpam-6087	551	28	,	,	PUNCT
ejpam-6087	551	29	showing	show	VERB
ejpam-6087	551	30	a	a	DET
ejpam-6087	551	31	slight	slight	ADJ
ejpam-6087	551	32	decrease	decrease	NOUN
ejpam-6087	551	33	as	as	ADP
ejpam-6087	551	34	the	the	DET
ejpam-6087	551	35	number	number	NOUN
ejpam-6087	551	36	of	of	ADP
ejpam-6087	551	37	principal	principal	ADJ
ejpam-6087	551	38	m.	m.	NOUN
ejpam-6087	551	39	al	al	PROPN
ejpam-6087	551	40	-	-	PUNCT
ejpam-6087	551	41	tarawneh	tarawneh	PROPN
ejpam-6087	551	42	et	et	PROPN
ejpam-6087	551	43	al	al	PROPN
ejpam-6087	551	44	.	.	PUNCT
ejpam-6087	551	45	/	/	SYM
ejpam-6087	551	46	eur	eur	PROPN
ejpam-6087	551	47	.	.	PUNCT
ejpam-6087	552	1	j.	j.	PROPN
ejpam-6087	552	2	pure	pure	PROPN
ejpam-6087	552	3	appl	appl	PROPN
ejpam-6087	552	4	.	.	PROPN
ejpam-6087	552	5	math	math	PROPN
ejpam-6087	552	6	,	,	PUNCT
ejpam-6087	552	7	18	18	NUM
ejpam-6087	552	8	(	(	PUNCT
ejpam-6087	552	9	2	2	NUM
ejpam-6087	552	10	)	)	PUNCT
ejpam-6087	552	11	(	(	PUNCT
ejpam-6087	552	12	2025	2025	NUM
ejpam-6087	552	13	)	)	PUNCT
ejpam-6087	552	14	,	,	PUNCT
ejpam-6087	552	15	6087	6087	NUM
ejpam-6087	552	16	30	30	NUM
ejpam-6087	552	17	of	of	ADP
ejpam-6087	552	18	39	39	NUM
ejpam-6087	552	19	components	component	NOUN
ejpam-6087	552	20	increases	increase	NOUN
ejpam-6087	552	21	.	.	PUNCT
ejpam-6087	553	1	lr	lr	PROPN
ejpam-6087	553	2	shows	show	VERB
ejpam-6087	553	3	a	a	DET
ejpam-6087	553	4	steady	steady	ADJ
ejpam-6087	553	5	improvement	improvement	NOUN
ejpam-6087	553	6	from	from	ADP
ejpam-6087	553	7	54.03	54.03	NUM
ejpam-6087	553	8	%	%	NOUN
ejpam-6087	553	9	to	to	ADP
ejpam-6087	553	10	58.30	58.30	NUM
ejpam-6087	553	11	%	%	NOUN
ejpam-6087	553	12	,	,	PUNCT
ejpam-6087	553	13	reflecting	reflect	VERB
ejpam-6087	553	14	a	a	DET
ejpam-6087	553	15	positive	positive	ADJ
ejpam-6087	553	16	trend	trend	NOUN
ejpam-6087	553	17	in	in	ADP
ejpam-6087	553	18	its	its	PRON
ejpam-6087	553	19	ability	ability	NOUN
ejpam-6087	553	20	to	to	PART
ejpam-6087	553	21	balance	balance	VERB
ejpam-6087	553	22	precision	precision	NOUN
ejpam-6087	553	23	and	and	CCONJ
ejpam-6087	553	24	recall	recall	NOUN
ejpam-6087	553	25	.	.	PUNCT
ejpam-6087	554	1	mlp	mlp	NOUN
ejpam-6087	554	2	shows	show	VERB
ejpam-6087	554	3	relatively	relatively	ADV
ejpam-6087	554	4	poor	poor	ADJ
ejpam-6087	554	5	f-1	f-1	NOUN
ejpam-6087	554	6	scores	score	NOUN
ejpam-6087	554	7	,	,	PUNCT
ejpam-6087	554	8	with	with	ADP
ejpam-6087	554	9	values	value	NOUN
ejpam-6087	554	10	starting	start	VERB
ejpam-6087	554	11	at	at	ADP
ejpam-6087	554	12	44.65	44.65	NUM
ejpam-6087	554	13	%	%	NOUN
ejpam-6087	554	14	and	and	CCONJ
ejpam-6087	554	15	increasing	increase	VERB
ejpam-6087	554	16	to	to	ADP
ejpam-6087	554	17	45.33	45.33	NUM
ejpam-6087	554	18	%	%	NOUN
ejpam-6087	554	19	,	,	PUNCT
ejpam-6087	554	20	indicating	indicate	VERB
ejpam-6087	554	21	that	that	SCONJ
ejpam-6087	554	22	this	this	DET
ejpam-6087	554	23	model	model	NOUN
ejpam-6087	554	24	struggles	struggle	VERB
ejpam-6087	554	25	to	to	PART
ejpam-6087	554	26	effectively	effectively	ADV
ejpam-6087	554	27	balance	balance	VERB
ejpam-6087	554	28	precision	precision	NOUN
ejpam-6087	554	29	and	and	CCONJ
ejpam-6087	554	30	recall	recall	NOUN
ejpam-6087	554	31	.	.	PUNCT
ejpam-6087	555	1	nb	nb	PROPN
ejpam-6087	555	2	also	also	ADV
ejpam-6087	555	3	shows	show	VERB
ejpam-6087	555	4	low	low	ADJ
ejpam-6087	555	5	f-1	f-1	PROPN
ejpam-6087	555	6	scores	score	NOUN
ejpam-6087	555	7	,	,	PUNCT
ejpam-6087	555	8	starting	start	VERB
ejpam-6087	555	9	at	at	ADP
ejpam-6087	555	10	37.21	37.21	NUM
ejpam-6087	555	11	%	%	NOUN
ejpam-6087	555	12	and	and	CCONJ
ejpam-6087	555	13	increasing	increase	VERB
ejpam-6087	555	14	to	to	ADP
ejpam-6087	555	15	45.04	45.04	NUM
ejpam-6087	555	16	%	%	NOUN
ejpam-6087	555	17	,	,	PUNCT
ejpam-6087	555	18	with	with	ADP
ejpam-6087	555	19	a	a	DET
ejpam-6087	555	20	notable	notable	ADJ
ejpam-6087	555	21	but	but	CCONJ
ejpam-6087	555	22	still	still	ADV
ejpam-6087	555	23	limited	limited	ADJ
ejpam-6087	555	24	improvement	improvement	NOUN
ejpam-6087	555	25	in	in	ADP
ejpam-6087	555	26	performance	performance	NOUN
ejpam-6087	555	27	.	.	PUNCT
ejpam-6087	556	1	finally	finally	ADV
ejpam-6087	556	2	,	,	PUNCT
ejpam-6087	556	3	svm	svm	PROPN
ejpam-6087	556	4	maintains	maintain	VERB
ejpam-6087	556	5	low	low	ADJ
ejpam-6087	556	6	and	and	CCONJ
ejpam-6087	556	7	relatively	relatively	ADV
ejpam-6087	556	8	stable	stable	ADJ
ejpam-6087	556	9	f-1	f-1	NOUN
ejpam-6087	556	10	scores	score	NOUN
ejpam-6087	556	11	throughout	throughout	ADP
ejpam-6087	556	12	the	the	DET
ejpam-6087	556	13	range	range	NOUN
ejpam-6087	556	14	of	of	ADP
ejpam-6087	556	15	principal	principal	ADJ
ejpam-6087	556	16	components	component	NOUN
ejpam-6087	556	17	,	,	PUNCT
ejpam-6087	556	18	starting	start	VERB
ejpam-6087	556	19	at	at	ADP
ejpam-6087	556	20	51.74	51.74	NUM
ejpam-6087	556	21	%	%	NOUN
ejpam-6087	556	22	at	at	ADP
ejpam-6087	556	23	5	5	NUM
ejpam-6087	556	24	components	component	NOUN
ejpam-6087	556	25	and	and	CCONJ
ejpam-6087	556	26	remaining	remain	VERB
ejpam-6087	556	27	nearly	nearly	ADV
ejpam-6087	556	28	unchanged	unchanged	ADJ
ejpam-6087	556	29	up	up	ADP
ejpam-6087	556	30	to	to	PART
ejpam-6087	556	31	30	30	NUM
ejpam-6087	556	32	components	component	NOUN
ejpam-6087	556	33	.	.	PUNCT
ejpam-6087	557	1	this	this	PRON
ejpam-6087	557	2	reflects	reflect	VERB
ejpam-6087	557	3	a	a	DET
ejpam-6087	557	4	consistent	consistent	ADJ
ejpam-6087	557	5	,	,	PUNCT
ejpam-6087	557	6	yet	yet	ADV
ejpam-6087	557	7	underwhelming	underwhelming	ADJ
ejpam-6087	557	8	,	,	PUNCT
ejpam-6087	557	9	performance	performance	NOUN
ejpam-6087	557	10	in	in	ADP
ejpam-6087	557	11	achieving	achieve	VERB
ejpam-6087	557	12	a	a	DET
ejpam-6087	557	13	balance	balance	NOUN
ejpam-6087	557	14	between	between	ADP
ejpam-6087	557	15	precision	precision	NOUN
ejpam-6087	557	16	and	and	CCONJ
ejpam-6087	557	17	recall	recall	NOUN
ejpam-6087	557	18	.	.	PUNCT
ejpam-6087	558	1	in	in	ADP
ejpam-6087	558	2	summary	summary	NOUN
ejpam-6087	558	3	,	,	PUNCT
ejpam-6087	558	4	the	the	DET
ejpam-6087	558	5	f-1	f-1	NOUN
ejpam-6087	558	6	score	score	NOUN
ejpam-6087	558	7	analysis	analysis	NOUN
ejpam-6087	558	8	reveals	reveal	VERB
ejpam-6087	558	9	that	that	SCONJ
ejpam-6087	558	10	while	while	SCONJ
ejpam-6087	558	11	rf	rf	PRON
ejpam-6087	558	12	leads	lead	VERB
ejpam-6087	558	13	in	in	ADP
ejpam-6087	558	14	balanced	balanced	ADJ
ejpam-6087	558	15	performance	performance	NOUN
ejpam-6087	558	16	,	,	PUNCT
ejpam-6087	558	17	other	other	ADJ
ejpam-6087	558	18	models	model	NOUN
ejpam-6087	558	19	such	such	ADJ
ejpam-6087	558	20	as	as	ADP
ejpam-6087	558	21	knn	knn	PROPN
ejpam-6087	558	22	,	,	PUNCT
ejpam-6087	558	23	lr	lr	NOUN
ejpam-6087	558	24	,	,	PUNCT
ejpam-6087	558	25	and	and	CCONJ
ejpam-6087	558	26	dt	dt	AUX
ejpam-6087	558	27	show	show	VERB
ejpam-6087	558	28	varying	vary	VERB
ejpam-6087	558	29	levels	level	NOUN
ejpam-6087	558	30	of	of	ADP
ejpam-6087	558	31	effectiveness	effectiveness	NOUN
ejpam-6087	558	32	,	,	PUNCT
ejpam-6087	558	33	and	and	CCONJ
ejpam-6087	558	34	mlp	mlp	PROPN
ejpam-6087	558	35	and	and	CCONJ
ejpam-6087	558	36	nb	nb	PROPN
ejpam-6087	558	37	demonstrate	demonstrate	VERB
ejpam-6087	558	38	relatively	relatively	ADV
ejpam-6087	558	39	poor	poor	ADJ
ejpam-6087	558	40	performance	performance	NOUN
ejpam-6087	558	41	.	.	PUNCT
ejpam-6087	559	1	the	the	DET
ejpam-6087	559	2	use	use	NOUN
ejpam-6087	559	3	of	of	ADP
ejpam-6087	559	4	pca	pca	NOUN
ejpam-6087	559	5	for	for	ADP
ejpam-6087	559	6	feature	feature	NOUN
ejpam-6087	559	7	selection	selection	NOUN
ejpam-6087	559	8	does	do	AUX
ejpam-6087	559	9	not	not	PART
ejpam-6087	559	10	lead	lead	VERB
ejpam-6087	559	11	to	to	ADP
ejpam-6087	559	12	significant	significant	ADJ
ejpam-6087	559	13	improvements	improvement	NOUN
ejpam-6087	559	14	in	in	ADP
ejpam-6087	559	15	f-1	f-1	NOUN
ejpam-6087	559	16	scores	score	NOUN
ejpam-6087	559	17	in	in	ADP
ejpam-6087	559	18	most	most	ADJ
ejpam-6087	559	19	models	model	NOUN
ejpam-6087	559	20	.	.	PUNCT
ejpam-6087	560	1	figure	figure	NOUN
ejpam-6087	560	2	3e	3e	PROPN
ejpam-6087	560	3	displays	display	VERB
ejpam-6087	560	4	the	the	DET
ejpam-6087	560	5	auc	auc	NOUN
ejpam-6087	560	6	-	-	PUNCT
ejpam-6087	560	7	roc	roc	NOUN
ejpam-6087	560	8	scores	score	NOUN
ejpam-6087	560	9	for	for	ADP
ejpam-6087	560	10	various	various	ADJ
ejpam-6087	560	11	ml	ml	NOUN
ejpam-6087	560	12	models	model	NOUN
ejpam-6087	560	13	as	as	ADP
ejpam-6087	560	14	the	the	DET
ejpam-6087	560	15	number	number	NOUN
ejpam-6087	560	16	of	of	ADP
ejpam-6087	560	17	principal	principal	ADJ
ejpam-6087	560	18	components	component	NOUN
ejpam-6087	560	19	increases	increase	NOUN
ejpam-6087	560	20	from	from	ADP
ejpam-6087	560	21	5	5	NUM
ejpam-6087	560	22	to	to	PART
ejpam-6087	560	23	30	30	NUM
ejpam-6087	560	24	.	.	PUNCT
ejpam-6087	561	1	the	the	DET
ejpam-6087	561	2	auc	auc	NOUN
ejpam-6087	561	3	-	-	PUNCT
ejpam-6087	561	4	roc	roc	NOUN
ejpam-6087	561	5	score	score	NOUN
ejpam-6087	561	6	measures	measure	VERB
ejpam-6087	561	7	the	the	DET
ejpam-6087	561	8	performance	performance	NOUN
ejpam-6087	561	9	of	of	ADP
ejpam-6087	561	10	a	a	DET
ejpam-6087	561	11	model	model	NOUN
ejpam-6087	561	12	in	in	ADP
ejpam-6087	561	13	distinguishing	distinguish	VERB
ejpam-6087	561	14	between	between	ADP
ejpam-6087	561	15	classes	class	NOUN
ejpam-6087	561	16	,	,	PUNCT
ejpam-6087	561	17	with	with	ADP
ejpam-6087	561	18	higher	high	ADJ
ejpam-6087	561	19	values	value	NOUN
ejpam-6087	561	20	indicating	indicate	VERB
ejpam-6087	561	21	better	well	ADJ
ejpam-6087	561	22	performance	performance	NOUN
ejpam-6087	561	23	.	.	PUNCT
ejpam-6087	562	1	rf	rf	NOUN
ejpam-6087	562	2	demonstrates	demonstrate	VERB
ejpam-6087	562	3	the	the	DET
ejpam-6087	562	4	highest	high	ADJ
ejpam-6087	562	5	auc	auc	NOUN
ejpam-6087	562	6	-	-	PUNCT
ejpam-6087	562	7	roc	roc	NOUN
ejpam-6087	562	8	scores	score	NOUN
ejpam-6087	562	9	among	among	ADP
ejpam-6087	562	10	all	all	DET
ejpam-6087	562	11	models	model	NOUN
ejpam-6087	562	12	,	,	PUNCT
ejpam-6087	562	13	starting	start	VERB
ejpam-6087	562	14	at	at	ADP
ejpam-6087	562	15	68.10	68.10	NUM
ejpam-6087	562	16	%	%	NOUN
ejpam-6087	562	17	with	with	ADP
ejpam-6087	562	18	5	5	NUM
ejpam-6087	562	19	principal	principal	ADJ
ejpam-6087	562	20	components	component	NOUN
ejpam-6087	562	21	and	and	CCONJ
ejpam-6087	562	22	increasing	increase	VERB
ejpam-6087	562	23	to	to	ADP
ejpam-6087	562	24	70.18	70.18	NUM
ejpam-6087	562	25	%	%	NOUN
ejpam-6087	562	26	with	with	ADP
ejpam-6087	562	27	30	30	NUM
ejpam-6087	562	28	components	component	NOUN
ejpam-6087	562	29	.	.	PUNCT
ejpam-6087	563	1	this	this	PRON
ejpam-6087	563	2	indicates	indicate	VERB
ejpam-6087	563	3	that	that	SCONJ
ejpam-6087	563	4	rf	rf	PRON
ejpam-6087	563	5	consistently	consistently	ADV
ejpam-6087	563	6	performs	perform	VERB
ejpam-6087	563	7	well	well	ADV
ejpam-6087	563	8	in	in	ADP
ejpam-6087	563	9	terms	term	NOUN
ejpam-6087	563	10	of	of	ADP
ejpam-6087	563	11	class	class	NOUN
ejpam-6087	563	12	separation	separation	NOUN
ejpam-6087	563	13	as	as	SCONJ
ejpam-6087	563	14	more	more	ADJ
ejpam-6087	563	15	principal	principal	ADJ
ejpam-6087	563	16	components	component	NOUN
ejpam-6087	563	17	are	be	AUX
ejpam-6087	563	18	considered	consider	VERB
ejpam-6087	563	19	.	.	PUNCT
ejpam-6087	564	1	knn	knn	PROPN
ejpam-6087	564	2	also	also	ADV
ejpam-6087	564	3	performs	perform	VERB
ejpam-6087	564	4	well	well	ADV
ejpam-6087	564	5	in	in	ADP
ejpam-6087	564	6	auc	auc	NOUN
ejpam-6087	564	7	-	-	PUNCT
ejpam-6087	564	8	roc	roc	NOUN
ejpam-6087	564	9	,	,	PUNCT
ejpam-6087	564	10	showing	show	VERB
ejpam-6087	564	11	steady	steady	ADJ
ejpam-6087	564	12	and	and	CCONJ
ejpam-6087	564	13	high	high	ADJ
ejpam-6087	564	14	performance	performance	NOUN
ejpam-6087	564	15	from	from	ADP
ejpam-6087	564	16	62.60	62.60	NUM
ejpam-6087	564	17	%	%	NOUN
ejpam-6087	564	18	at	at	ADP
ejpam-6087	564	19	5	5	NUM
ejpam-6087	564	20	components	component	NOUN
ejpam-6087	564	21	to	to	ADP
ejpam-6087	564	22	62.60	62.60	NUM
ejpam-6087	564	23	%	%	NOUN
ejpam-6087	564	24	at	at	ADP
ejpam-6087	564	25	30	30	NUM
ejpam-6087	564	26	components	component	NOUN
ejpam-6087	564	27	.	.	PUNCT
ejpam-6087	565	1	this	this	PRON
ejpam-6087	565	2	indicates	indicate	VERB
ejpam-6087	565	3	that	that	SCONJ
ejpam-6087	565	4	knn	knn	PROPN
ejpam-6087	565	5	maintains	maintain	VERB
ejpam-6087	565	6	strong	strong	ADJ
ejpam-6087	565	7	performance	performance	NOUN
ejpam-6087	565	8	throughout	throughout	ADP
ejpam-6087	565	9	the	the	DET
ejpam-6087	565	10	feature	feature	NOUN
ejpam-6087	565	11	selection	selection	NOUN
ejpam-6087	565	12	process	process	NOUN
ejpam-6087	565	13	.	.	PUNCT
ejpam-6087	566	1	lr	lr	NOUN
ejpam-6087	566	2	and	and	CCONJ
ejpam-6087	566	3	nb	nb	PROPN
ejpam-6087	566	4	show	show	VERB
ejpam-6087	566	5	moderate	moderate	ADJ
ejpam-6087	566	6	auc	auc	NOUN
ejpam-6087	566	7	-	-	PUNCT
ejpam-6087	566	8	roc	roc	NOUN
ejpam-6087	566	9	scores	score	NOUN
ejpam-6087	566	10	.	.	PUNCT
ejpam-6087	567	1	lr	lr	X
ejpam-6087	567	2	starts	start	NOUN
ejpam-6087	567	3	at	at	ADP
ejpam-6087	567	4	55.95	55.95	NUM
ejpam-6087	567	5	%	%	NOUN
ejpam-6087	567	6	and	and	CCONJ
ejpam-6087	567	7	increases	increase	VERB
ejpam-6087	567	8	to	to	ADP
ejpam-6087	567	9	61.55	61.55	NUM
ejpam-6087	567	10	%	%	NOUN
ejpam-6087	567	11	,	,	PUNCT
ejpam-6087	567	12	reflecting	reflect	VERB
ejpam-6087	567	13	a	a	DET
ejpam-6087	567	14	trend	trend	NOUN
ejpam-6087	567	15	of	of	ADP
ejpam-6087	567	16	improved	improved	ADJ
ejpam-6087	567	17	performance	performance	NOUN
ejpam-6087	567	18	with	with	ADP
ejpam-6087	567	19	more	more	ADJ
ejpam-6087	567	20	principal	principal	ADJ
ejpam-6087	567	21	components	component	NOUN
ejpam-6087	567	22	.	.	PUNCT
ejpam-6087	568	1	nb	nb	PROPN
ejpam-6087	568	2	starts	start	VERB
ejpam-6087	568	3	at	at	ADP
ejpam-6087	568	4	55.04	55.04	NUM
ejpam-6087	568	5	%	%	NOUN
ejpam-6087	568	6	and	and	CCONJ
ejpam-6087	568	7	increases	increase	VERB
ejpam-6087	568	8	to	to	ADP
ejpam-6087	568	9	58.38	58.38	NUM
ejpam-6087	568	10	%	%	NOUN
ejpam-6087	568	11	,	,	PUNCT
ejpam-6087	568	12	showing	show	VERB
ejpam-6087	568	13	a	a	DET
ejpam-6087	568	14	more	more	ADV
ejpam-6087	568	15	modest	modest	ADJ
ejpam-6087	568	16	improvement	improvement	NOUN
ejpam-6087	568	17	in	in	ADP
ejpam-6087	568	18	performance	performance	NOUN
ejpam-6087	568	19	.	.	PUNCT
ejpam-6087	569	1	mlp	mlp	NOUN
ejpam-6087	569	2	shows	show	VERB
ejpam-6087	569	3	relatively	relatively	ADV
ejpam-6087	569	4	high	high	ADJ
ejpam-6087	569	5	auc	auc	NOUN
ejpam-6087	569	6	-	-	PUNCT
ejpam-6087	569	7	roc	roc	NOUN
ejpam-6087	569	8	values	value	NOUN
ejpam-6087	569	9	,	,	PUNCT
ejpam-6087	569	10	starting	start	VERB
ejpam-6087	569	11	at	at	ADP
ejpam-6087	569	12	48.11	48.11	NUM
ejpam-6087	569	13	%	%	NOUN
ejpam-6087	569	14	and	and	CCONJ
ejpam-6087	569	15	increasing	increase	VERB
ejpam-6087	569	16	to	to	ADP
ejpam-6087	569	17	58.47	58.47	NUM
ejpam-6087	569	18	%	%	NOUN
ejpam-6087	569	19	as	as	ADP
ejpam-6087	569	20	the	the	DET
ejpam-6087	569	21	number	number	NOUN
ejpam-6087	569	22	of	of	ADP
ejpam-6087	569	23	principal	principal	ADJ
ejpam-6087	569	24	components	component	NOUN
ejpam-6087	569	25	increases	increase	NOUN
ejpam-6087	569	26	.	.	PUNCT
ejpam-6087	570	1	this	this	PRON
ejpam-6087	570	2	shows	show	VERB
ejpam-6087	570	3	a	a	DET
ejpam-6087	570	4	gradual	gradual	ADJ
ejpam-6087	570	5	improvement	improvement	NOUN
ejpam-6087	570	6	in	in	ADP
ejpam-6087	570	7	performance	performance	NOUN
ejpam-6087	570	8	,	,	PUNCT
ejpam-6087	570	9	though	though	SCONJ
ejpam-6087	570	10	it	it	PRON
ejpam-6087	570	11	remains	remain	VERB
ejpam-6087	570	12	lower	low	ADJ
ejpam-6087	570	13	compared	compare	VERB
ejpam-6087	570	14	to	to	ADP
ejpam-6087	570	15	the	the	DET
ejpam-6087	570	16	top	top	ADV
ejpam-6087	570	17	-	-	PUNCT
ejpam-6087	570	18	performing	perform	VERB
ejpam-6087	570	19	models	model	NOUN
ejpam-6087	570	20	.	.	PUNCT
ejpam-6087	571	1	dt	dt	PROPN
ejpam-6087	571	2	exhibits	exhibit	VERB
ejpam-6087	571	3	the	the	DET
ejpam-6087	571	4	lowest	low	ADJ
ejpam-6087	571	5	auc	auc	NOUN
ejpam-6087	571	6	-	-	PUNCT
ejpam-6087	571	7	roc	roc	NOUN
ejpam-6087	571	8	scores	score	NOUN
ejpam-6087	571	9	among	among	ADP
ejpam-6087	571	10	the	the	DET
ejpam-6087	571	11	models	model	NOUN
ejpam-6087	571	12	,	,	PUNCT
ejpam-6087	571	13	with	with	ADP
ejpam-6087	571	14	values	value	NOUN
ejpam-6087	571	15	ranging	range	VERB
ejpam-6087	571	16	from	from	ADP
ejpam-6087	571	17	54.45	54.45	NUM
ejpam-6087	571	18	%	%	NOUN
ejpam-6087	571	19	to	to	ADP
ejpam-6087	571	20	56.04	56.04	NUM
ejpam-6087	571	21	%	%	NOUN
ejpam-6087	571	22	.	.	PUNCT
ejpam-6087	572	1	this	this	PRON
ejpam-6087	572	2	indicates	indicate	VERB
ejpam-6087	572	3	that	that	SCONJ
ejpam-6087	572	4	dt	dt	PROPN
ejpam-6087	572	5	struggles	struggle	VERB
ejpam-6087	572	6	more	more	ADJ
ejpam-6087	572	7	than	than	ADP
ejpam-6087	572	8	other	other	ADJ
ejpam-6087	572	9	models	model	NOUN
ejpam-6087	572	10	in	in	ADP
ejpam-6087	572	11	distinguishing	distinguish	VERB
ejpam-6087	572	12	between	between	ADP
ejpam-6087	572	13	classes	class	NOUN
ejpam-6087	572	14	effectively	effectively	ADV
ejpam-6087	572	15	.	.	PUNCT
ejpam-6087	573	1	svm	svm	PROPN
ejpam-6087	573	2	maintains	maintain	VERB
ejpam-6087	573	3	relatively	relatively	ADV
ejpam-6087	573	4	stable	stable	ADJ
ejpam-6087	573	5	but	but	CCONJ
ejpam-6087	573	6	low	low	ADJ
ejpam-6087	573	7	auc	auc	NOUN
ejpam-6087	573	8	-	-	PUNCT
ejpam-6087	573	9	roc	roc	NOUN
ejpam-6087	573	10	performance	performance	NOUN
ejpam-6087	573	11	,	,	PUNCT
ejpam-6087	573	12	starting	start	VERB
ejpam-6087	573	13	at	at	ADP
ejpam-6087	573	14	55.73	55.73	NUM
ejpam-6087	573	15	%	%	NOUN
ejpam-6087	573	16	and	and	CCONJ
ejpam-6087	573	17	showing	show	VERB
ejpam-6087	573	18	minimal	minimal	ADJ
ejpam-6087	573	19	variation	variation	NOUN
ejpam-6087	573	20	up	up	ADP
ejpam-6087	573	21	to	to	PART
ejpam-6087	573	22	30	30	NUM
ejpam-6087	573	23	principal	principal	ADJ
ejpam-6087	573	24	components	component	NOUN
ejpam-6087	573	25	.	.	PUNCT
ejpam-6087	574	1	this	this	PRON
ejpam-6087	574	2	indicates	indicate	VERB
ejpam-6087	574	3	that	that	SCONJ
ejpam-6087	574	4	svm	svm	PROPN
ejpam-6087	574	5	’s	’s	PART
ejpam-6087	574	6	ability	ability	NOUN
ejpam-6087	574	7	to	to	PART
ejpam-6087	574	8	distinguish	distinguish	VERB
ejpam-6087	574	9	between	between	ADP
ejpam-6087	574	10	classes	class	NOUN
ejpam-6087	574	11	is	be	AUX
ejpam-6087	574	12	limited	limit	VERB
ejpam-6087	574	13	compared	compare	VERB
ejpam-6087	574	14	to	to	ADP
ejpam-6087	574	15	other	other	ADJ
ejpam-6087	574	16	models	model	NOUN
ejpam-6087	574	17	.	.	PUNCT
ejpam-6087	575	1	in	in	ADP
ejpam-6087	575	2	summary	summary	NOUN
ejpam-6087	575	3	,	,	PUNCT
ejpam-6087	575	4	rf	rf	PRON
ejpam-6087	575	5	leads	lead	VERB
ejpam-6087	575	6	in	in	ADP
ejpam-6087	575	7	auc	auc	NOUN
ejpam-6087	575	8	-	-	PUNCT
ejpam-6087	575	9	roc	roc	NOUN
ejpam-6087	575	10	performance	performance	NOUN
ejpam-6087	575	11	,	,	PUNCT
ejpam-6087	575	12	showing	show	VERB
ejpam-6087	575	13	the	the	DET
ejpam-6087	575	14	best	good	ADJ
ejpam-6087	575	15	results	result	NOUN
ejpam-6087	575	16	among	among	ADP
ejpam-6087	575	17	all	all	DET
ejpam-6087	575	18	models	model	NOUN
ejpam-6087	575	19	as	as	ADP
ejpam-6087	575	20	the	the	DET
ejpam-6087	575	21	number	number	NOUN
ejpam-6087	575	22	of	of	ADP
ejpam-6087	575	23	principal	principal	ADJ
ejpam-6087	575	24	components	component	NOUN
ejpam-6087	575	25	increases	increase	NOUN
ejpam-6087	575	26	.	.	PUNCT
ejpam-6087	576	1	knn	knn	PROPN
ejpam-6087	576	2	also	also	ADV
ejpam-6087	576	3	performs	perform	VERB
ejpam-6087	576	4	well	well	ADV
ejpam-6087	576	5	,	,	PUNCT
ejpam-6087	576	6	while	while	SCONJ
ejpam-6087	576	7	other	other	ADJ
ejpam-6087	576	8	models	model	NOUN
ejpam-6087	576	9	like	like	ADP
ejpam-6087	576	10	dt	dt	PUNCT
ejpam-6087	576	11	and	and	CCONJ
ejpam-6087	576	12	svm	svm	PROPN
ejpam-6087	576	13	show	show	VERB
ejpam-6087	576	14	relatively	relatively	ADV
ejpam-6087	576	15	poor	poor	ADJ
ejpam-6087	576	16	auc	auc	NOUN
ejpam-6087	576	17	-	-	PUNCT
ejpam-6087	576	18	roc	roc	NOUN
ejpam-6087	576	19	scores	score	NOUN
ejpam-6087	576	20	.	.	PUNCT
ejpam-6087	577	1	most	most	ADJ
ejpam-6087	577	2	models	model	NOUN
ejpam-6087	577	3	exhibit	exhibit	VERB
ejpam-6087	577	4	some	some	DET
ejpam-6087	577	5	improvement	improvement	NOUN
ejpam-6087	577	6	with	with	ADP
ejpam-6087	577	7	more	more	ADJ
ejpam-6087	577	8	principal	principal	ADJ
ejpam-6087	577	9	components	component	NOUN
ejpam-6087	577	10	,	,	PUNCT
ejpam-6087	577	11	but	but	CCONJ
ejpam-6087	577	12	overall	overall	ADJ
ejpam-6087	577	13	performance	performance	NOUN
ejpam-6087	577	14	is	be	AUX
ejpam-6087	577	15	modest	modest	ADJ
ejpam-6087	577	16	compared	compare	VERB
ejpam-6087	577	17	to	to	ADP
ejpam-6087	577	18	other	other	ADJ
ejpam-6087	577	19	feature	feature	NOUN
ejpam-6087	577	20	selection	selection	NOUN
ejpam-6087	577	21	methods	method	NOUN
ejpam-6087	577	22	.	.	PUNCT
ejpam-6087	578	1	principal	principal	ADJ
ejpam-6087	578	2	component	component	NOUN
ejpam-6087	578	3	analysis	analysis	NOUN
ejpam-6087	578	4	(	(	PUNCT
ejpam-6087	578	5	pca	pca	NOUN
ejpam-6087	578	6	)	)	PUNCT
ejpam-6087	578	7	consistently	consistently	ADV
ejpam-6087	578	8	yielded	yield	VERB
ejpam-6087	578	9	lower	low	ADJ
ejpam-6087	578	10	performance	performance	NOUN
ejpam-6087	578	11	compared	compare	VERB
ejpam-6087	578	12	to	to	ADP
ejpam-6087	578	13	other	other	ADJ
ejpam-6087	578	14	feature	feature	NOUN
ejpam-6087	578	15	selection	selection	NOUN
ejpam-6087	578	16	methods	method	NOUN
ejpam-6087	578	17	in	in	ADP
ejpam-6087	578	18	all	all	DET
ejpam-6087	578	19	classifiers	classifier	NOUN
ejpam-6087	578	20	.	.	PUNCT
ejpam-6087	579	1	this	this	DET
ejpam-6087	579	2	underperformance	underperformance	NOUN
ejpam-6087	579	3	can	can	AUX
ejpam-6087	579	4	theoretically	theoretically	ADV
ejpam-6087	579	5	be	be	AUX
ejpam-6087	579	6	explained	explain	VERB
ejpam-6087	579	7	by	by	ADP
ejpam-6087	579	8	the	the	DET
ejpam-6087	579	9	nature	nature	NOUN
ejpam-6087	579	10	of	of	ADP
ejpam-6087	579	11	pca	pca	PROPN
ejpam-6087	579	12	itself	itself	PRON
ejpam-6087	579	13	.	.	PUNCT
ejpam-6087	580	1	as	as	ADP
ejpam-6087	580	2	an	an	DET
ejpam-6087	580	3	unsupervised	unsupervised	ADJ
ejpam-6087	580	4	dimensionality	dimensionality	NOUN
ejpam-6087	580	5	reduction	reduction	NOUN
ejpam-6087	580	6	technique	technique	NOUN
ejpam-6087	580	7	,	,	PUNCT
ejpam-6087	580	8	pca	pca	PROPN
ejpam-6087	580	9	transforms	transform	VERB
ejpam-6087	580	10	the	the	DET
ejpam-6087	580	11	original	original	ADJ
ejpam-6087	580	12	feature	feature	NOUN
ejpam-6087	580	13	space	space	NOUN
ejpam-6087	580	14	into	into	ADP
ejpam-6087	580	15	orthogonal	orthogonal	ADJ
ejpam-6087	580	16	components	component	NOUN
ejpam-6087	580	17	based	base	VERB
ejpam-6087	580	18	on	on	ADP
ejpam-6087	580	19	variance	variance	NOUN
ejpam-6087	580	20	,	,	PUNCT
ejpam-6087	580	21	without	without	ADP
ejpam-6087	580	22	considering	consider	VERB
ejpam-6087	580	23	class	class	NOUN
ejpam-6087	580	24	labels	label	NOUN
ejpam-6087	580	25	.	.	PUNCT
ejpam-6087	581	1	although	although	SCONJ
ejpam-6087	581	2	this	this	PRON
ejpam-6087	581	3	helps	help	VERB
ejpam-6087	581	4	to	to	PART
ejpam-6087	581	5	capture	capture	VERB
ejpam-6087	581	6	m.	m.	PROPN
ejpam-6087	581	7	al	al	PROPN
ejpam-6087	581	8	-	-	PUNCT
ejpam-6087	581	9	tarawneh	tarawneh	PROPN
ejpam-6087	581	10	et	et	PROPN
ejpam-6087	581	11	al	al	PROPN
ejpam-6087	581	12	.	.	PUNCT
ejpam-6087	581	13	/	/	SYM
ejpam-6087	581	14	eur	eur	PROPN
ejpam-6087	581	15	.	.	PUNCT
ejpam-6087	582	1	j.	j.	PROPN
ejpam-6087	582	2	pure	pure	PROPN
ejpam-6087	582	3	appl	appl	PROPN
ejpam-6087	582	4	.	.	PROPN
ejpam-6087	582	5	math	math	PROPN
ejpam-6087	582	6	,	,	PUNCT
ejpam-6087	582	7	18	18	NUM
ejpam-6087	582	8	(	(	PUNCT
ejpam-6087	582	9	2	2	NUM
ejpam-6087	582	10	)	)	PUNCT
ejpam-6087	582	11	(	(	PUNCT
ejpam-6087	582	12	2025	2025	NUM
ejpam-6087	582	13	)	)	PUNCT
ejpam-6087	582	14	,	,	PUNCT
ejpam-6087	582	15	6087	6087	NUM
ejpam-6087	582	16	31	31	NUM
ejpam-6087	582	17	of	of	ADP
ejpam-6087	582	18	39	39	NUM
ejpam-6087	582	19	table	table	NOUN
ejpam-6087	582	20	6	6	NUM
ejpam-6087	582	21	:	:	PUNCT
ejpam-6087	582	22	selected	select	VERB
ejpam-6087	582	23	features	feature	NOUN
ejpam-6087	582	24	by	by	ADP
ejpam-6087	582	25	the	the	DET
ejpam-6087	582	26	genetic	genetic	ADJ
ejpam-6087	582	27	algorithm	algorithm	NOUN
ejpam-6087	582	28	(	(	PUNCT
ejpam-6087	582	29	ga	ga	NOUN
ejpam-6087	582	30	)	)	PUNCT
ejpam-6087	582	31	no	no	NOUN
ejpam-6087	583	1	.	.	PUNCT
ejpam-6087	583	2	selected	select	VERB
ejpam-6087	583	3	feature	feature	NOUN
ejpam-6087	583	4	category	category	NOUN
ejpam-6087	583	5	1	1	NUM
ejpam-6087	583	6	binarynumtarget	binarynumtarget	VERB
ejpam-6087	583	7	meta	meta	ADJ
ejpam-6087	583	8	-	-	PUNCT
ejpam-6087	583	9	data	data	NOUN
ejpam-6087	583	10	2	2	NUM
ejpam-6087	583	11	statuses	status	NOUN
ejpam-6087	583	12	count	count	VERB
ejpam-6087	583	13	meta	meta	ADJ
ejpam-6087	583	14	-	-	PUNCT
ejpam-6087	583	15	data	data	NOUN
ejpam-6087	583	16	3	3	NUM
ejpam-6087	583	17	following	follow	VERB
ejpam-6087	583	18	meta	meta	ADJ
ejpam-6087	583	19	-	-	PUNCT
ejpam-6087	583	20	data	data	NOUN
ejpam-6087	583	21	4	4	NUM
ejpam-6087	583	22	botscorebinary	botscorebinary	ADJ
ejpam-6087	583	23	meta	meta	ADJ
ejpam-6087	583	24	-	-	PUNCT
ejpam-6087	583	25	data	data	NOUN
ejpam-6087	583	26	5	5	NUM
ejpam-6087	583	27	normalize	normalize	VERB
ejpam-6087	583	28	influence	influence	NOUN
ejpam-6087	583	29	meta	meta	NOUN
ejpam-6087	583	30	-	-	PUNCT
ejpam-6087	583	31	data	data	NOUN
ejpam-6087	583	32	6	6	NUM
ejpam-6087	583	33	mentions	mention	VERB
ejpam-6087	583	34	meta	meta	ADJ
ejpam-6087	583	35	-	-	PUNCT
ejpam-6087	583	36	data	data	NOUN
ejpam-6087	583	37	7	7	NUM
ejpam-6087	583	38	replies	reply	NOUN
ejpam-6087	583	39	meta	meta	ADJ
ejpam-6087	583	40	-	-	PUNCT
ejpam-6087	583	41	data	data	NOUN
ejpam-6087	583	42	8	8	NUM
ejpam-6087	583	43	favourites	favourite	NOUN
ejpam-6087	583	44	meta	meta	ADJ
ejpam-6087	583	45	-	-	PUNCT
ejpam-6087	583	46	data	data	NOUN
ejpam-6087	583	47	9	9	NUM
ejpam-6087	583	48	org	org	NOUN
ejpam-6087	583	49	percentage	percentage	NOUN
ejpam-6087	583	50	text	text	NOUN
ejpam-6087	583	51	10	10	NUM
ejpam-6087	583	52	date	date	NOUN
ejpam-6087	583	53	percentage	percentage	NOUN
ejpam-6087	583	54	text	text	NOUN
ejpam-6087	583	55	11	11	NUM
ejpam-6087	583	56	fac	fac	NOUN
ejpam-6087	583	57	percentage	percentage	NOUN
ejpam-6087	583	58	text	text	NOUN
ejpam-6087	583	59	12	12	NUM
ejpam-6087	583	60	law	law	NOUN
ejpam-6087	583	61	percentage	percentage	NOUN
ejpam-6087	583	62	text	text	NOUN
ejpam-6087	583	63	13	13	NUM
ejpam-6087	583	64	product	product	NOUN
ejpam-6087	583	65	percentage	percentage	NOUN
ejpam-6087	583	66	text	text	NOUN
ejpam-6087	583	67	14	14	NUM
ejpam-6087	583	68	loc	loc	NOUN
ejpam-6087	583	69	percentage	percentage	NOUN
ejpam-6087	583	70	text	text	NOUN
ejpam-6087	583	71	15	15	NUM
ejpam-6087	583	72	language	language	NOUN
ejpam-6087	583	73	percentage	percentage	NOUN
ejpam-6087	583	74	text	text	NOUN
ejpam-6087	583	75	16	16	NUM
ejpam-6087	583	76	word	word	NOUN
ejpam-6087	583	77	count	count	NOUN
ejpam-6087	583	78	text	text	NOUN
ejpam-6087	583	79	17	17	NUM
ejpam-6087	583	80	min	min	NOUN
ejpam-6087	583	81	word	word	NOUN
ejpam-6087	583	82	length	length	NOUN
ejpam-6087	583	83	text	text	NOUN
ejpam-6087	583	84	18	18	NUM
ejpam-6087	583	85	present	present	ADJ
ejpam-6087	583	86	verbs	verb	NOUN
ejpam-6087	583	87	lexical	lexical	ADJ
ejpam-6087	583	88	19	19	NUM
ejpam-6087	583	89	past	past	ADJ
ejpam-6087	583	90	verbs	verbs	PROPN
ejpam-6087	583	91	lexical	lexical	ADJ
ejpam-6087	583	92	20	20	NUM
ejpam-6087	583	93	adverbs	adverb	NOUN
ejpam-6087	583	94	lexical	lexical	ADJ
ejpam-6087	583	95	21	21	NUM
ejpam-6087	583	96	questions	question	NOUN
ejpam-6087	583	97	lexical	lexical	VERB
ejpam-6087	583	98	22	22	NUM
ejpam-6087	583	99	ampersand	ampersand	ADJ
ejpam-6087	583	100	lexical	lexical	ADJ
ejpam-6087	583	101	23	23	NUM
ejpam-6087	583	102	long	long	ADJ
ejpam-6087	583	103	word	word	NOUN
ejpam-6087	583	104	freq	freq	NOUN
ejpam-6087	583	105	lexical	lexical	VERB
ejpam-6087	583	106	the	the	DET
ejpam-6087	583	107	overall	overall	ADJ
ejpam-6087	583	108	structure	structure	NOUN
ejpam-6087	583	109	of	of	ADP
ejpam-6087	583	110	the	the	DET
ejpam-6087	583	111	data	datum	NOUN
ejpam-6087	583	112	,	,	PUNCT
ejpam-6087	583	113	it	it	PRON
ejpam-6087	583	114	does	do	AUX
ejpam-6087	583	115	not	not	PART
ejpam-6087	583	116	necessarily	necessarily	ADV
ejpam-6087	583	117	preserve	preserve	VERB
ejpam-6087	583	118	the	the	DET
ejpam-6087	583	119	most	most	ADV
ejpam-6087	583	120	relevant	relevant	ADJ
ejpam-6087	583	121	features	feature	NOUN
ejpam-6087	583	122	for	for	ADP
ejpam-6087	583	123	distinguishing	distinguish	VERB
ejpam-6087	583	124	between	between	ADP
ejpam-6087	583	125	real	real	ADJ
ejpam-6087	583	126	and	and	CCONJ
ejpam-6087	583	127	fake	fake	ADJ
ejpam-6087	583	128	content	content	NOUN
ejpam-6087	583	129	.	.	PUNCT
ejpam-6087	584	1	in	in	ADP
ejpam-6087	584	2	contrast	contrast	NOUN
ejpam-6087	584	3	,	,	PUNCT
ejpam-6087	584	4	methods	method	NOUN
ejpam-6087	584	5	such	such	ADJ
ejpam-6087	584	6	as	as	ADP
ejpam-6087	584	7	rfe	rfe	NOUN
ejpam-6087	584	8	and	and	CCONJ
ejpam-6087	584	9	selectkbest	selectkb	ADJ
ejpam-6087	584	10	directly	directly	ADV
ejpam-6087	584	11	evaluate	evaluate	VERB
ejpam-6087	584	12	the	the	DET
ejpam-6087	584	13	relevance	relevance	NOUN
ejpam-6087	584	14	of	of	ADP
ejpam-6087	584	15	each	each	DET
ejpam-6087	584	16	feature	feature	NOUN
ejpam-6087	584	17	with	with	ADP
ejpam-6087	584	18	respect	respect	NOUN
ejpam-6087	584	19	to	to	ADP
ejpam-6087	584	20	the	the	DET
ejpam-6087	584	21	classification	classification	NOUN
ejpam-6087	584	22	target	target	NOUN
ejpam-6087	584	23	,	,	PUNCT
ejpam-6087	584	24	making	make	VERB
ejpam-6087	584	25	them	they	PRON
ejpam-6087	584	26	more	more	ADV
ejpam-6087	584	27	effective	effective	ADJ
ejpam-6087	584	28	in	in	ADP
ejpam-6087	584	29	contexts	context	NOUN
ejpam-6087	584	30	where	where	SCONJ
ejpam-6087	584	31	feature	feature	NOUN
ejpam-6087	584	32	-	-	PUNCT
ejpam-6087	584	33	label	label	NOUN
ejpam-6087	584	34	relationships	relationship	NOUN
ejpam-6087	584	35	are	be	AUX
ejpam-6087	584	36	critical	critical	ADJ
ejpam-6087	584	37	.	.	PUNCT
ejpam-6087	585	1	4.3.4	4.3.4	X
ejpam-6087	585	2	.	.	PUNCT
ejpam-6087	586	1	genetic	genetic	ADJ
ejpam-6087	586	2	algorithm	algorithm	NOUN
ejpam-6087	586	3	results	result	VERB
ejpam-6087	586	4	this	this	DET
ejpam-6087	586	5	section	section	NOUN
ejpam-6087	586	6	presents	present	VERB
ejpam-6087	586	7	the	the	DET
ejpam-6087	586	8	results	result	NOUN
ejpam-6087	586	9	obtained	obtain	VERB
ejpam-6087	586	10	by	by	ADP
ejpam-6087	586	11	various	various	ADJ
ejpam-6087	586	12	ml	ml	NOUN
ejpam-6087	586	13	models	model	NOUN
ejpam-6087	586	14	,	,	PUNCT
ejpam-6087	586	15	following	follow	VERB
ejpam-6087	586	16	a	a	DET
ejpam-6087	586	17	feature	feature	NOUN
ejpam-6087	586	18	selection	selection	NOUN
ejpam-6087	586	19	step	step	NOUN
ejpam-6087	586	20	performed	perform	VERB
ejpam-6087	586	21	using	use	VERB
ejpam-6087	586	22	the	the	DET
ejpam-6087	586	23	ga	ga	PROPN
ejpam-6087	586	24	.	.	PUNCT
ejpam-6087	587	1	the	the	DET
ejpam-6087	587	2	ga	ga	PROPN
ejpam-6087	587	3	algorithm	algorithm	PROPN
ejpam-6087	587	4	optimally	optimally	ADV
ejpam-6087	587	5	selected	select	VERB
ejpam-6087	587	6	23	23	NUM
ejpam-6087	587	7	features	feature	NOUN
ejpam-6087	587	8	,	,	PUNCT
ejpam-6087	587	9	as	as	SCONJ
ejpam-6087	587	10	shown	show	VERB
ejpam-6087	587	11	in	in	ADP
ejpam-6087	587	12	table	table	NOUN
ejpam-6087	587	13	6	6	NUM
ejpam-6087	587	14	.	.	PUNCT
ejpam-6087	588	1	this	this	DET
ejpam-6087	588	2	set	set	NOUN
ejpam-6087	588	3	of	of	ADP
ejpam-6087	588	4	features	feature	NOUN
ejpam-6087	588	5	was	be	AUX
ejpam-6087	588	6	used	use	VERB
ejpam-6087	588	7	to	to	PART
ejpam-6087	588	8	train	train	VERB
ejpam-6087	588	9	and	and	CCONJ
ejpam-6087	588	10	evaluate	evaluate	VERB
ejpam-6087	588	11	multiple	multiple	ADJ
ejpam-6087	588	12	ml	ml	NOUN
ejpam-6087	588	13	models	model	NOUN
ejpam-6087	588	14	,	,	PUNCT
ejpam-6087	588	15	and	and	CCONJ
ejpam-6087	588	16	their	their	PRON
ejpam-6087	588	17	performance	performance	NOUN
ejpam-6087	588	18	was	be	AUX
ejpam-6087	588	19	assessed	assess	VERB
ejpam-6087	588	20	based	base	VERB
ejpam-6087	588	21	on	on	ADP
ejpam-6087	588	22	metrics	metric	NOUN
ejpam-6087	588	23	including	include	VERB
ejpam-6087	588	24	accuracy	accuracy	NOUN
ejpam-6087	588	25	,	,	PUNCT
ejpam-6087	588	26	precision	precision	NOUN
ejpam-6087	588	27	,	,	PUNCT
ejpam-6087	588	28	recall	recall	NOUN
ejpam-6087	588	29	,	,	PUNCT
ejpam-6087	588	30	f1	f1	NOUN
ejpam-6087	588	31	score	score	NOUN
ejpam-6087	588	32	,	,	PUNCT
ejpam-6087	588	33	and	and	CCONJ
ejpam-6087	588	34	auc	auc	NOUN
ejpam-6087	588	35	-	-	PUNCT
ejpam-6087	588	36	roc	roc	NOUN
ejpam-6087	588	37	,	,	PUNCT
ejpam-6087	588	38	as	as	SCONJ
ejpam-6087	588	39	shown	show	VERB
ejpam-6087	588	40	in	in	ADP
ejpam-6087	588	41	figure	figure	NOUN
ejpam-6087	588	42	4	4	NUM
ejpam-6087	588	43	.	.	PUNCT
ejpam-6087	588	44	to	to	PART
ejpam-6087	588	45	minimize	minimize	VERB
ejpam-6087	588	46	the	the	DET
ejpam-6087	588	47	risk	risk	NOUN
ejpam-6087	588	48	of	of	ADP
ejpam-6087	588	49	overfitting	overfitte	VERB
ejpam-6087	588	50	during	during	ADP
ejpam-6087	588	51	the	the	DET
ejpam-6087	588	52	selection	selection	NOUN
ejpam-6087	588	53	of	of	ADP
ejpam-6087	588	54	features	feature	NOUN
ejpam-6087	588	55	based	base	VERB
ejpam-6087	588	56	on	on	ADP
ejpam-6087	588	57	genetic	genetic	ADJ
ejpam-6087	588	58	algorithms	algorithm	NOUN
ejpam-6087	588	59	,	,	PUNCT
ejpam-6087	588	60	a	a	DET
ejpam-6087	588	61	5	5	NUM
ejpam-6087	588	62	-	-	ADJ
ejpam-6087	588	63	fold	fold	ADJ
ejpam-6087	588	64	cross	cross	ADJ
ejpam-6087	588	65	-	-	ADJ
ejpam-6087	588	66	validation	validation	ADJ
ejpam-6087	588	67	strategy	strategy	NOUN
ejpam-6087	588	68	was	be	AUX
ejpam-6087	588	69	integrated	integrate	VERB
ejpam-6087	588	70	into	into	ADP
ejpam-6087	588	71	the	the	DET
ejpam-6087	588	72	ga	ga	PROPN
ejpam-6087	588	73	fitness	fitness	PROPN
ejpam-6087	588	74	evaluation	evaluation	NOUN
ejpam-6087	588	75	.	.	PUNCT
ejpam-6087	589	1	following	follow	VERB
ejpam-6087	589	2	the	the	DET
ejpam-6087	589	3	feature	feature	NOUN
ejpam-6087	589	4	selection	selection	NOUN
ejpam-6087	589	5	process	process	NOUN
ejpam-6087	589	6	,	,	PUNCT
ejpam-6087	589	7	all	all	DET
ejpam-6087	589	8	classifiers	classifier	NOUN
ejpam-6087	589	9	were	be	AUX
ejpam-6087	589	10	re	re	VERB
ejpam-6087	589	11	-	-	VERB
ejpam-6087	589	12	trained	train	VERB
ejpam-6087	589	13	and	and	CCONJ
ejpam-6087	589	14	evaluated	evaluate	VERB
ejpam-6087	589	15	on	on	ADP
ejpam-6087	589	16	a	a	DET
ejpam-6087	589	17	separate	separate	ADJ
ejpam-6087	589	18	unseen	unseen	ADJ
ejpam-6087	589	19	test	test	NOUN
ejpam-6087	589	20	set	set	VERB
ejpam-6087	589	21	to	to	PART
ejpam-6087	589	22	ensure	ensure	VERB
ejpam-6087	589	23	robust	robust	ADJ
ejpam-6087	589	24	and	and	CCONJ
ejpam-6087	589	25	unbiased	unbiased	ADJ
ejpam-6087	589	26	performance	performance	NOUN
ejpam-6087	589	27	assessment	assessment	NOUN
ejpam-6087	589	28	.	.	PUNCT
ejpam-6087	590	1	figure	figure	NOUN
ejpam-6087	590	2	4a	4a	NOUN
ejpam-6087	590	3	illustrates	illustrate	VERB
ejpam-6087	590	4	the	the	DET
ejpam-6087	590	5	accuracy	accuracy	NOUN
ejpam-6087	590	6	scores	score	NOUN
ejpam-6087	590	7	of	of	ADP
ejpam-6087	590	8	various	various	ADJ
ejpam-6087	590	9	ml	ml	NOUN
ejpam-6087	590	10	models	model	NOUN
ejpam-6087	590	11	following	follow	VERB
ejpam-6087	590	12	the	the	DET
ejpam-6087	590	13	feature	feature	NOUN
ejpam-6087	590	14	selection	selection	NOUN
ejpam-6087	590	15	step	step	NOUN
ejpam-6087	590	16	performed	perform	VERB
ejpam-6087	590	17	by	by	ADP
ejpam-6087	590	18	the	the	DET
ejpam-6087	590	19	ga	ga	PROPN
ejpam-6087	590	20	algorithm	algorithm	NOUN
ejpam-6087	590	21	.	.	PUNCT
ejpam-6087	591	1	the	the	DET
ejpam-6087	591	2	dt	dt	PROPN
ejpam-6087	591	3	and	and	CCONJ
ejpam-6087	591	4	rf	rf	NOUN
ejpam-6087	591	5	models	model	NOUN
ejpam-6087	591	6	exhibit	exhibit	VERB
ejpam-6087	591	7	the	the	DET
ejpam-6087	591	8	highest	high	ADJ
ejpam-6087	591	9	accuracy	accuracy	NOUN
ejpam-6087	591	10	among	among	ADP
ejpam-6087	591	11	all	all	DET
ejpam-6087	591	12	models	model	NOUN
ejpam-6087	591	13	,	,	PUNCT
ejpam-6087	591	14	achieving	achieve	VERB
ejpam-6087	591	15	nearly	nearly	ADV
ejpam-6087	591	16	91	91	NUM
ejpam-6087	591	17	%	%	NOUN
ejpam-6087	591	18	and	and	CCONJ
ejpam-6087	591	19	96	96	NUM
ejpam-6087	591	20	%	%	NOUN
ejpam-6087	591	21	,	,	PUNCT
ejpam-6087	591	22	respectively	respectively	ADV
ejpam-6087	591	23	.	.	PUNCT
ejpam-6087	592	1	lr	lr	PROPN
ejpam-6087	592	2	also	also	ADV
ejpam-6087	592	3	shows	show	VERB
ejpam-6087	592	4	a	a	DET
ejpam-6087	592	5	high	high	ADJ
ejpam-6087	592	6	accuracy	accuracy	NOUN
ejpam-6087	592	7	of	of	ADP
ejpam-6087	592	8	approximately	approximately	ADV
ejpam-6087	592	9	96	96	NUM
ejpam-6087	592	10	%	%	NOUN
ejpam-6087	592	11	,	,	PUNCT
ejpam-6087	592	12	indicating	indicate	VERB
ejpam-6087	592	13	strong	strong	ADJ
ejpam-6087	592	14	performance	performance	NOUN
ejpam-6087	592	15	on	on	ADP
ejpam-6087	592	16	the	the	DET
ejpam-6087	592	17	test	test	NOUN
ejpam-6087	592	18	set	set	NOUN
ejpam-6087	592	19	.	.	PUNCT
ejpam-6087	593	1	in	in	ADP
ejpam-6087	593	2	contrast	contrast	NOUN
ejpam-6087	593	3	,	,	PUNCT
ejpam-6087	593	4	the	the	DET
ejpam-6087	593	5	knn	knn	PROPN
ejpam-6087	593	6	and	and	CCONJ
ejpam-6087	593	7	svm	svm	ADJ
ejpam-6087	593	8	models	model	NOUN
ejpam-6087	593	9	demonstrate	demonstrate	VERB
ejpam-6087	593	10	significantly	significantly	ADV
ejpam-6087	593	11	lower	low	ADJ
ejpam-6087	593	12	accuracy	accuracy	NOUN
ejpam-6087	593	13	,	,	PUNCT
ejpam-6087	593	14	with	with	ADP
ejpam-6087	593	15	values	value	NOUN
ejpam-6087	593	16	of	of	ADP
ejpam-6087	593	17	around	around	ADP
ejpam-6087	593	18	56	56	NUM
ejpam-6087	593	19	%	%	NOUN
ejpam-6087	593	20	and	and	CCONJ
ejpam-6087	593	21	54	54	NUM
ejpam-6087	593	22	%	%	NOUN
ejpam-6087	593	23	,	,	PUNCT
ejpam-6087	593	24	respectively	respectively	ADV
ejpam-6087	593	25	.	.	PUNCT
ejpam-6087	594	1	m.	m.	PROPN
ejpam-6087	594	2	al	al	PROPN
ejpam-6087	594	3	-	-	PUNCT
ejpam-6087	594	4	tarawneh	tarawneh	PROPN
ejpam-6087	594	5	et	et	PROPN
ejpam-6087	594	6	al	al	PROPN
ejpam-6087	594	7	.	.	PUNCT
ejpam-6087	594	8	/	/	SYM
ejpam-6087	594	9	eur	eur	PROPN
ejpam-6087	594	10	.	.	PUNCT
ejpam-6087	595	1	j.	j.	PROPN
ejpam-6087	595	2	pure	pure	PROPN
ejpam-6087	595	3	appl	appl	PROPN
ejpam-6087	595	4	.	.	PROPN
ejpam-6087	595	5	math	math	PROPN
ejpam-6087	595	6	,	,	PUNCT
ejpam-6087	595	7	18	18	NUM
ejpam-6087	595	8	(	(	PUNCT
ejpam-6087	595	9	2	2	NUM
ejpam-6087	595	10	)	)	PUNCT
ejpam-6087	595	11	(	(	PUNCT
ejpam-6087	595	12	2025	2025	NUM
ejpam-6087	595	13	)	)	PUNCT
ejpam-6087	595	14	,	,	PUNCT
ejpam-6087	595	15	6087	6087	NUM
ejpam-6087	595	16	32	32	NUM
ejpam-6087	595	17	of	of	ADP
ejpam-6087	595	18	39	39	NUM
ejpam-6087	595	19	a	a	DET
ejpam-6087	595	20	accuracy	accuracy	NOUN
ejpam-6087	596	1	b	b	PROPN
ejpam-6087	596	2	precision	precision	NOUN
ejpam-6087	596	3	c	c	NOUN
ejpam-6087	596	4	recall	recall	PROPN
ejpam-6087	596	5	d	d	PROPN
ejpam-6087	596	6	f1	f1	PROPN
ejpam-6087	596	7	score	score	NOUN
ejpam-6087	596	8	e	e	NOUN
ejpam-6087	596	9	auc	auc	ADJ
ejpam-6087	596	10	-	-	PUNCT
ejpam-6087	596	11	roc	roc	NOUN
ejpam-6087	596	12	figure	figure	NOUN
ejpam-6087	596	13	4	4	NUM
ejpam-6087	596	14	:	:	PUNCT
ejpam-6087	596	15	performance	performance	NOUN
ejpam-6087	596	16	metrics	metric	NOUN
ejpam-6087	596	17	of	of	ADP
ejpam-6087	596	18	various	various	ADJ
ejpam-6087	596	19	models	model	NOUN
ejpam-6087	596	20	under	under	ADP
ejpam-6087	596	21	ga	ga	PROPN
ejpam-6087	596	22	feature	feature	NOUN
ejpam-6087	596	23	selection	selection	NOUN
ejpam-6087	596	24	.	.	PUNCT
ejpam-6087	597	1	mlp	mlp	NOUN
ejpam-6087	597	2	performs	perform	VERB
ejpam-6087	597	3	moderately	moderately	ADV
ejpam-6087	597	4	with	with	ADP
ejpam-6087	597	5	an	an	DET
ejpam-6087	597	6	accuracy	accuracy	NOUN
ejpam-6087	597	7	of	of	ADP
ejpam-6087	597	8	approximately	approximately	ADV
ejpam-6087	597	9	70	70	NUM
ejpam-6087	597	10	%	%	NOUN
ejpam-6087	597	11	,	,	PUNCT
ejpam-6087	597	12	while	while	SCONJ
ejpam-6087	597	13	nb	nb	PROPN
ejpam-6087	597	14	shows	show	VERB
ejpam-6087	597	15	the	the	DET
ejpam-6087	597	16	lowest	low	ADJ
ejpam-6087	597	17	accuracy	accuracy	NOUN
ejpam-6087	597	18	at	at	ADP
ejpam-6087	597	19	approximately	approximately	ADV
ejpam-6087	597	20	52	52	NUM
ejpam-6087	597	21	%	%	NOUN
ejpam-6087	597	22	.	.	PUNCT
ejpam-6087	598	1	overall	overall	ADV
ejpam-6087	598	2	,	,	PUNCT
ejpam-6087	598	3	the	the	DET
ejpam-6087	598	4	accuracy	accuracy	NOUN
ejpam-6087	598	5	results	result	VERB
ejpam-6087	598	6	highlight	highlight	VERB
ejpam-6087	598	7	that	that	SCONJ
ejpam-6087	598	8	ensemble	ensemble	ADJ
ejpam-6087	598	9	methods	method	NOUN
ejpam-6087	598	10	such	such	ADJ
ejpam-6087	598	11	as	as	ADP
ejpam-6087	598	12	dt	dt	PROPN
ejpam-6087	598	13	and	and	CCONJ
ejpam-6087	598	14	rf	rf	PRON
ejpam-6087	598	15	perform	perform	VERB
ejpam-6087	598	16	significantly	significantly	ADV
ejpam-6087	598	17	better	well	ADV
ejpam-6087	598	18	compared	compare	VERB
ejpam-6087	598	19	to	to	ADP
ejpam-6087	598	20	other	other	ADJ
ejpam-6087	598	21	models	model	NOUN
ejpam-6087	598	22	following	follow	VERB
ejpam-6087	598	23	the	the	DET
ejpam-6087	598	24	feature	feature	NOUN
ejpam-6087	598	25	selection	selection	NOUN
ejpam-6087	598	26	step	step	NOUN
ejpam-6087	598	27	.	.	PUNCT
ejpam-6087	599	1	figure	figure	VERB
ejpam-6087	599	2	4b	4b	PROPN
ejpam-6087	599	3	presents	present	VERB
ejpam-6087	599	4	the	the	DET
ejpam-6087	599	5	precision	precision	NOUN
ejpam-6087	599	6	scores	score	NOUN
ejpam-6087	599	7	for	for	ADP
ejpam-6087	599	8	m.	m.	NOUN
ejpam-6087	599	9	al	al	PROPN
ejpam-6087	599	10	-	-	PUNCT
ejpam-6087	599	11	tarawneh	tarawneh	PROPN
ejpam-6087	599	12	et	et	PROPN
ejpam-6087	599	13	al	al	PROPN
ejpam-6087	599	14	.	.	PUNCT
ejpam-6087	599	15	/	/	SYM
ejpam-6087	599	16	eur	eur	PROPN
ejpam-6087	599	17	.	.	PUNCT
ejpam-6087	600	1	j.	j.	PROPN
ejpam-6087	600	2	pure	pure	PROPN
ejpam-6087	600	3	appl	appl	PROPN
ejpam-6087	600	4	.	.	PROPN
ejpam-6087	600	5	math	math	PROPN
ejpam-6087	600	6	,	,	PUNCT
ejpam-6087	600	7	18	18	NUM
ejpam-6087	600	8	(	(	PUNCT
ejpam-6087	600	9	2	2	NUM
ejpam-6087	600	10	)	)	PUNCT
ejpam-6087	600	11	(	(	PUNCT
ejpam-6087	600	12	2025	2025	NUM
ejpam-6087	600	13	)	)	PUNCT
ejpam-6087	600	14	,	,	PUNCT
ejpam-6087	600	15	6087	6087	NUM
ejpam-6087	600	16	33	33	NUM
ejpam-6087	600	17	of	of	ADP
ejpam-6087	600	18	39	39	NUM
ejpam-6087	600	19	different	different	ADJ
ejpam-6087	600	20	models	model	NOUN
ejpam-6087	600	21	after	after	ADP
ejpam-6087	600	22	applying	apply	VERB
ejpam-6087	600	23	the	the	DET
ejpam-6087	600	24	ga	ga	NOUN
ejpam-6087	600	25	feature	feature	NOUN
ejpam-6087	600	26	selection	selection	NOUN
ejpam-6087	600	27	algorithm	algorithm	NOUN
ejpam-6087	600	28	.	.	PUNCT
ejpam-6087	601	1	lr	lr	NOUN
ejpam-6087	601	2	and	and	CCONJ
ejpam-6087	601	3	rf	rf	PRON
ejpam-6087	601	4	achieve	achieve	VERB
ejpam-6087	601	5	the	the	DET
ejpam-6087	601	6	highest	high	ADJ
ejpam-6087	601	7	precision	precision	NOUN
ejpam-6087	601	8	scores	score	NOUN
ejpam-6087	601	9	of	of	ADP
ejpam-6087	601	10	approximately	approximately	ADV
ejpam-6087	601	11	96	96	NUM
ejpam-6087	601	12	%	%	NOUN
ejpam-6087	601	13	and	and	CCONJ
ejpam-6087	601	14	96	96	NUM
ejpam-6087	601	15	%	%	NOUN
ejpam-6087	601	16	,	,	PUNCT
ejpam-6087	601	17	respectively	respectively	ADV
ejpam-6087	601	18	,	,	PUNCT
ejpam-6087	601	19	indicating	indicate	VERB
ejpam-6087	601	20	that	that	SCONJ
ejpam-6087	601	21	these	these	DET
ejpam-6087	601	22	models	model	NOUN
ejpam-6087	601	23	are	be	AUX
ejpam-6087	601	24	effective	effective	ADJ
ejpam-6087	601	25	at	at	ADP
ejpam-6087	601	26	correctly	correctly	ADV
ejpam-6087	601	27	identifying	identify	VERB
ejpam-6087	601	28	positive	positive	ADJ
ejpam-6087	601	29	instances	instance	NOUN
ejpam-6087	601	30	.	.	PUNCT
ejpam-6087	602	1	the	the	DET
ejpam-6087	602	2	dt	dt	PROPN
ejpam-6087	602	3	model	model	NOUN
ejpam-6087	602	4	also	also	ADV
ejpam-6087	602	5	shows	show	VERB
ejpam-6087	602	6	high	high	ADJ
ejpam-6087	602	7	precision	precision	NOUN
ejpam-6087	602	8	,	,	PUNCT
ejpam-6087	602	9	around	around	ADV
ejpam-6087	602	10	91	91	NUM
ejpam-6087	602	11	%	%	NOUN
ejpam-6087	602	12	,	,	PUNCT
ejpam-6087	602	13	demonstrating	demonstrate	VERB
ejpam-6087	602	14	its	its	PRON
ejpam-6087	602	15	effectiveness	effectiveness	NOUN
ejpam-6087	602	16	in	in	ADP
ejpam-6087	602	17	distinguishing	distinguish	VERB
ejpam-6087	602	18	positive	positive	ADJ
ejpam-6087	602	19	cases	case	NOUN
ejpam-6087	602	20	.	.	PUNCT
ejpam-6087	603	1	mlp	mlp	NOUN
ejpam-6087	603	2	shows	show	VERB
ejpam-6087	603	3	a	a	DET
ejpam-6087	603	4	good	good	ADJ
ejpam-6087	603	5	precision	precision	NOUN
ejpam-6087	603	6	of	of	ADP
ejpam-6087	603	7	about	about	ADV
ejpam-6087	603	8	80	80	NUM
ejpam-6087	603	9	%	%	NOUN
ejpam-6087	603	10	.	.	PUNCT
ejpam-6087	604	1	in	in	ADP
ejpam-6087	604	2	contrast	contrast	NOUN
ejpam-6087	604	3	,	,	PUNCT
ejpam-6087	604	4	the	the	DET
ejpam-6087	604	5	knn	knn	PROPN
ejpam-6087	604	6	and	and	CCONJ
ejpam-6087	604	7	svm	svm	ADJ
ejpam-6087	604	8	models	model	NOUN
ejpam-6087	604	9	exhibit	exhibit	VERB
ejpam-6087	604	10	lower	low	ADJ
ejpam-6087	604	11	precision	precision	NOUN
ejpam-6087	604	12	scores	score	NOUN
ejpam-6087	604	13	,	,	PUNCT
ejpam-6087	604	14	around	around	ADV
ejpam-6087	604	15	56	56	NUM
ejpam-6087	604	16	%	%	NOUN
ejpam-6087	604	17	and	and	CCONJ
ejpam-6087	604	18	54	54	NUM
ejpam-6087	604	19	%	%	NOUN
ejpam-6087	604	20	,	,	PUNCT
ejpam-6087	604	21	respectively	respectively	ADV
ejpam-6087	604	22	.	.	PUNCT
ejpam-6087	605	1	the	the	DET
ejpam-6087	605	2	nb	nb	PROPN
ejpam-6087	605	3	model	model	NOUN
ejpam-6087	605	4	achieves	achieve	VERB
ejpam-6087	605	5	a	a	DET
ejpam-6087	605	6	moderate	moderate	ADJ
ejpam-6087	605	7	precision	precision	NOUN
ejpam-6087	605	8	score	score	NOUN
ejpam-6087	605	9	of	of	ADP
ejpam-6087	605	10	approximately	approximately	ADV
ejpam-6087	605	11	58	58	NUM
ejpam-6087	605	12	%	%	NOUN
ejpam-6087	605	13	.	.	PUNCT
ejpam-6087	606	1	this	this	DET
ejpam-6087	606	2	figure	figure	NOUN
ejpam-6087	606	3	underscores	underscore	VERB
ejpam-6087	606	4	that	that	SCONJ
ejpam-6087	606	5	lr	lr	NOUN
ejpam-6087	606	6	,	,	PUNCT
ejpam-6087	606	7	rf	rf	ADJ
ejpam-6087	606	8	,	,	PUNCT
ejpam-6087	606	9	and	and	CCONJ
ejpam-6087	606	10	dt	dt	PROPN
ejpam-6087	606	11	are	be	AUX
ejpam-6087	606	12	highly	highly	ADV
ejpam-6087	606	13	precise	precise	ADJ
ejpam-6087	606	14	,	,	PUNCT
ejpam-6087	606	15	while	while	SCONJ
ejpam-6087	606	16	other	other	ADJ
ejpam-6087	606	17	models	model	NOUN
ejpam-6087	606	18	like	like	ADP
ejpam-6087	606	19	knn	knn	PROPN
ejpam-6087	606	20	and	and	CCONJ
ejpam-6087	606	21	svm	svm	PROPN
ejpam-6087	606	22	show	show	VERB
ejpam-6087	606	23	relatively	relatively	ADV
ejpam-6087	606	24	poor	poor	ADJ
ejpam-6087	606	25	performance	performance	NOUN
ejpam-6087	606	26	in	in	ADP
ejpam-6087	606	27	terms	term	NOUN
ejpam-6087	606	28	of	of	ADP
ejpam-6087	606	29	precision	precision	NOUN
ejpam-6087	606	30	.	.	PUNCT
ejpam-6087	607	1	figure	figure	NOUN
ejpam-6087	607	2	4c	4c	NOUN
ejpam-6087	607	3	shows	show	VERB
ejpam-6087	607	4	the	the	DET
ejpam-6087	607	5	recall	recall	NOUN
ejpam-6087	607	6	scores	score	NOUN
ejpam-6087	607	7	for	for	ADP
ejpam-6087	607	8	various	various	ADJ
ejpam-6087	607	9	models	model	NOUN
ejpam-6087	607	10	after	after	ADP
ejpam-6087	607	11	feature	feature	NOUN
ejpam-6087	607	12	selection	selection	NOUN
ejpam-6087	607	13	using	use	VERB
ejpam-6087	607	14	the	the	DET
ejpam-6087	607	15	ga	ga	PROPN
ejpam-6087	607	16	algorithm	algorithm	NOUN
ejpam-6087	607	17	.	.	PUNCT
ejpam-6087	608	1	lr	lr	NOUN
ejpam-6087	608	2	and	and	CCONJ
ejpam-6087	608	3	rf	rf	PRON
ejpam-6087	608	4	achieve	achieve	VERB
ejpam-6087	608	5	the	the	DET
ejpam-6087	608	6	highest	high	ADJ
ejpam-6087	608	7	recall	recall	NOUN
ejpam-6087	608	8	scores	score	NOUN
ejpam-6087	608	9	,	,	PUNCT
ejpam-6087	608	10	reaching	reach	VERB
ejpam-6087	608	11	around	around	ADP
ejpam-6087	608	12	96	96	NUM
ejpam-6087	608	13	%	%	NOUN
ejpam-6087	608	14	and	and	CCONJ
ejpam-6087	608	15	96	96	NUM
ejpam-6087	608	16	%	%	NOUN
ejpam-6087	608	17	,	,	PUNCT
ejpam-6087	608	18	respectively	respectively	ADV
ejpam-6087	608	19	.	.	PUNCT
ejpam-6087	609	1	dt	dt	PUNCT
ejpam-6087	609	2	follows	follow	VERB
ejpam-6087	609	3	closely	closely	ADV
ejpam-6087	609	4	with	with	ADP
ejpam-6087	609	5	a	a	DET
ejpam-6087	609	6	recall	recall	NOUN
ejpam-6087	609	7	score	score	NOUN
ejpam-6087	609	8	of	of	ADP
ejpam-6087	609	9	approximately	approximately	ADV
ejpam-6087	609	10	91	91	NUM
ejpam-6087	609	11	%	%	NOUN
ejpam-6087	609	12	,	,	PUNCT
ejpam-6087	609	13	indicating	indicate	VERB
ejpam-6087	609	14	an	an	DET
ejpam-6087	609	15	effective	effective	ADJ
ejpam-6087	609	16	identification	identification	NOUN
ejpam-6087	609	17	of	of	ADP
ejpam-6087	609	18	positive	positive	ADJ
ejpam-6087	609	19	instances	instance	NOUN
ejpam-6087	609	20	.	.	PUNCT
ejpam-6087	610	1	mlp	mlp	PROPN
ejpam-6087	610	2	shows	show	VERB
ejpam-6087	610	3	a	a	DET
ejpam-6087	610	4	moderate	moderate	ADJ
ejpam-6087	610	5	recall	recall	NOUN
ejpam-6087	610	6	score	score	NOUN
ejpam-6087	610	7	of	of	ADP
ejpam-6087	610	8	about	about	ADP
ejpam-6087	610	9	70	70	NUM
ejpam-6087	610	10	%	%	NOUN
ejpam-6087	610	11	,	,	PUNCT
ejpam-6087	610	12	while	while	SCONJ
ejpam-6087	610	13	knn	knn	PROPN
ejpam-6087	610	14	and	and	CCONJ
ejpam-6087	610	15	svm	svm	PROPN
ejpam-6087	610	16	exhibit	exhibit	VERB
ejpam-6087	610	17	significantly	significantly	ADV
ejpam-6087	610	18	lower	low	ADJ
ejpam-6087	610	19	recall	recall	NOUN
ejpam-6087	610	20	scores	score	NOUN
ejpam-6087	610	21	,	,	PUNCT
ejpam-6087	610	22	approximately	approximately	ADV
ejpam-6087	610	23	56	56	NUM
ejpam-6087	610	24	%	%	NOUN
ejpam-6087	610	25	and	and	CCONJ
ejpam-6087	610	26	54	54	NUM
ejpam-6087	610	27	%	%	NOUN
ejpam-6087	610	28	,	,	PUNCT
ejpam-6087	610	29	respectively	respectively	ADV
ejpam-6087	610	30	.	.	PUNCT
ejpam-6087	611	1	nb	nb	PROPN
ejpam-6087	611	2	has	have	VERB
ejpam-6087	611	3	the	the	DET
ejpam-6087	611	4	lowest	low	ADJ
ejpam-6087	611	5	recall	recall	NOUN
ejpam-6087	611	6	score	score	NOUN
ejpam-6087	611	7	of	of	ADP
ejpam-6087	611	8	around	around	ADP
ejpam-6087	611	9	52	52	NUM
ejpam-6087	611	10	%	%	NOUN
ejpam-6087	611	11	.	.	PUNCT
ejpam-6087	612	1	recall	recall	VERB
ejpam-6087	612	2	analysis	analysis	NOUN
ejpam-6087	612	3	reveals	reveal	VERB
ejpam-6087	612	4	that	that	SCONJ
ejpam-6087	612	5	lr	lr	NOUN
ejpam-6087	612	6	and	and	CCONJ
ejpam-6087	612	7	rf	rf	NOUN
ejpam-6087	612	8	are	be	AUX
ejpam-6087	612	9	the	the	DET
ejpam-6087	612	10	most	most	ADV
ejpam-6087	612	11	effective	effective	ADJ
ejpam-6087	612	12	in	in	ADP
ejpam-6087	612	13	detecting	detect	VERB
ejpam-6087	612	14	positive	positive	ADJ
ejpam-6087	612	15	instances	instance	NOUN
ejpam-6087	612	16	,	,	PUNCT
ejpam-6087	612	17	while	while	SCONJ
ejpam-6087	612	18	other	other	ADJ
ejpam-6087	612	19	models	model	NOUN
ejpam-6087	612	20	such	such	ADJ
ejpam-6087	612	21	as	as	ADP
ejpam-6087	612	22	knn	knn	PROPN
ejpam-6087	612	23	and	and	CCONJ
ejpam-6087	612	24	svm	svm	PROPN
ejpam-6087	612	25	perform	perform	VERB
ejpam-6087	612	26	poorly	poorly	ADV
ejpam-6087	612	27	in	in	ADP
ejpam-6087	612	28	this	this	DET
ejpam-6087	612	29	regard	regard	NOUN
ejpam-6087	612	30	.	.	PUNCT
ejpam-6087	613	1	figure	figure	NOUN
ejpam-6087	613	2	4d	4d	NOUN
ejpam-6087	613	3	displays	display	VERB
ejpam-6087	613	4	the	the	DET
ejpam-6087	613	5	f1	f1	PROPN
ejpam-6087	613	6	scores	score	NOUN
ejpam-6087	613	7	for	for	ADP
ejpam-6087	613	8	different	different	ADJ
ejpam-6087	613	9	ml	ml	NOUN
ejpam-6087	613	10	models	model	NOUN
ejpam-6087	613	11	after	after	ADP
ejpam-6087	613	12	the	the	DET
ejpam-6087	613	13	ga	ga	PROPN
ejpam-6087	613	14	feature	feature	NOUN
ejpam-6087	613	15	selection	selection	NOUN
ejpam-6087	613	16	process	process	NOUN
ejpam-6087	613	17	.	.	PUNCT
ejpam-6087	614	1	the	the	DET
ejpam-6087	614	2	lr	lr	PROPN
ejpam-6087	614	3	and	and	CCONJ
ejpam-6087	614	4	rf	rf	NOUN
ejpam-6087	614	5	models	model	NOUN
ejpam-6087	614	6	lead	lead	VERB
ejpam-6087	614	7	with	with	ADP
ejpam-6087	614	8	the	the	DET
ejpam-6087	614	9	highest	high	ADJ
ejpam-6087	614	10	f1	f1	NOUN
ejpam-6087	614	11	scores	score	NOUN
ejpam-6087	614	12	,	,	PUNCT
ejpam-6087	614	13	approximately	approximately	ADV
ejpam-6087	614	14	96	96	NUM
ejpam-6087	614	15	%	%	NOUN
ejpam-6087	614	16	and	and	CCONJ
ejpam-6087	614	17	96	96	NUM
ejpam-6087	614	18	%	%	NOUN
ejpam-6087	614	19	,	,	PUNCT
ejpam-6087	614	20	respectively	respectively	ADV
ejpam-6087	614	21	.	.	PUNCT
ejpam-6087	615	1	dt	dt	PUNCT
ejpam-6087	615	2	also	also	ADV
ejpam-6087	615	3	shows	show	VERB
ejpam-6087	615	4	a	a	DET
ejpam-6087	615	5	high	high	ADJ
ejpam-6087	615	6	f1	f1	NOUN
ejpam-6087	615	7	score	score	NOUN
ejpam-6087	615	8	of	of	ADP
ejpam-6087	615	9	about	about	ADV
ejpam-6087	615	10	91	91	NUM
ejpam-6087	615	11	%	%	NOUN
ejpam-6087	615	12	,	,	PUNCT
ejpam-6087	615	13	demonstrating	demonstrate	VERB
ejpam-6087	615	14	balanced	balanced	ADJ
ejpam-6087	615	15	performance	performance	NOUN
ejpam-6087	615	16	between	between	ADP
ejpam-6087	615	17	precision	precision	NOUN
ejpam-6087	615	18	and	and	CCONJ
ejpam-6087	615	19	recall	recall	NOUN
ejpam-6087	615	20	.	.	PUNCT
ejpam-6087	616	1	mlp	mlp	NOUN
ejpam-6087	616	2	exhibits	exhibit	VERB
ejpam-6087	616	3	a	a	DET
ejpam-6087	616	4	moderate	moderate	ADJ
ejpam-6087	616	5	f1	f1	NOUN
ejpam-6087	616	6	score	score	NOUN
ejpam-6087	616	7	of	of	ADP
ejpam-6087	616	8	around	around	ADV
ejpam-6087	616	9	68	68	NUM
ejpam-6087	616	10	%	%	NOUN
ejpam-6087	616	11	.	.	PUNCT
ejpam-6087	617	1	in	in	ADP
ejpam-6087	617	2	contrast	contrast	NOUN
ejpam-6087	617	3	,	,	PUNCT
ejpam-6087	617	4	the	the	DET
ejpam-6087	617	5	knn	knn	PROPN
ejpam-6087	617	6	and	and	CCONJ
ejpam-6087	617	7	svm	svm	ADJ
ejpam-6087	617	8	models	model	NOUN
ejpam-6087	617	9	show	show	VERB
ejpam-6087	617	10	relatively	relatively	ADV
ejpam-6087	617	11	low	low	ADJ
ejpam-6087	617	12	f1	f1	NOUN
ejpam-6087	617	13	scores	score	NOUN
ejpam-6087	617	14	,	,	PUNCT
ejpam-6087	617	15	about	about	ADV
ejpam-6087	617	16	56	56	NUM
ejpam-6087	617	17	%	%	NOUN
ejpam-6087	617	18	and	and	CCONJ
ejpam-6087	617	19	54	54	NUM
ejpam-6087	617	20	%	%	NOUN
ejpam-6087	617	21	,	,	PUNCT
ejpam-6087	617	22	respectively	respectively	ADV
ejpam-6087	617	23	.	.	PUNCT
ejpam-6087	618	1	nb	nb	PROPN
ejpam-6087	618	2	achieves	achieve	VERB
ejpam-6087	618	3	the	the	DET
ejpam-6087	618	4	lowest	low	ADJ
ejpam-6087	618	5	f1	f1	NOUN
ejpam-6087	618	6	score	score	NOUN
ejpam-6087	618	7	of	of	ADP
ejpam-6087	618	8	approximately	approximately	ADV
ejpam-6087	618	9	36	36	NUM
ejpam-6087	618	10	%	%	NOUN
ejpam-6087	618	11	.	.	PUNCT
ejpam-6087	619	1	this	this	DET
ejpam-6087	619	2	figure	figure	NOUN
ejpam-6087	619	3	highlights	highlight	NOUN
ejpam-6087	619	4	that	that	SCONJ
ejpam-6087	619	5	the	the	DET
ejpam-6087	619	6	ensemble	ensemble	ADJ
ejpam-6087	619	7	methods	method	NOUN
ejpam-6087	619	8	,	,	PUNCT
ejpam-6087	619	9	specifically	specifically	ADV
ejpam-6087	619	10	dt	dt	X
ejpam-6087	619	11	and	and	CCONJ
ejpam-6087	619	12	rf	rf	VERB
ejpam-6087	619	13	,	,	PUNCT
ejpam-6087	619	14	along	along	ADP
ejpam-6087	619	15	with	with	ADP
ejpam-6087	619	16	lr	lr	NOUN
ejpam-6087	619	17	,	,	PUNCT
ejpam-6087	619	18	demonstrate	demonstrate	VERB
ejpam-6087	619	19	superior	superior	ADJ
ejpam-6087	619	20	overall	overall	ADJ
ejpam-6087	619	21	performance	performance	NOUN
ejpam-6087	619	22	compared	compare	VERB
ejpam-6087	619	23	to	to	ADP
ejpam-6087	619	24	other	other	ADJ
ejpam-6087	619	25	models	model	NOUN
ejpam-6087	619	26	.	.	PUNCT
ejpam-6087	620	1	figure	figure	NOUN
ejpam-6087	620	2	4e	4e	NOUN
ejpam-6087	620	3	presents	present	VERB
ejpam-6087	620	4	the	the	DET
ejpam-6087	620	5	auc	auc	NOUN
ejpam-6087	620	6	-	-	PUNCT
ejpam-6087	620	7	roc	roc	NOUN
ejpam-6087	620	8	scores	score	NOUN
ejpam-6087	620	9	for	for	ADP
ejpam-6087	620	10	different	different	ADJ
ejpam-6087	620	11	models	model	NOUN
ejpam-6087	620	12	following	follow	VERB
ejpam-6087	620	13	the	the	DET
ejpam-6087	620	14	ga	ga	PROPN
ejpam-6087	620	15	feature	feature	NOUN
ejpam-6087	620	16	selection	selection	NOUN
ejpam-6087	620	17	process	process	NOUN
ejpam-6087	620	18	.	.	PUNCT
ejpam-6087	621	1	rf	rf	PRON
ejpam-6087	621	2	achieves	achieve	VERB
ejpam-6087	621	3	the	the	DET
ejpam-6087	621	4	highest	high	ADJ
ejpam-6087	621	5	auc	auc	NOUN
ejpam-6087	621	6	-	-	PUNCT
ejpam-6087	621	7	roc	roc	NOUN
ejpam-6087	621	8	score	score	NOUN
ejpam-6087	621	9	of	of	ADP
ejpam-6087	621	10	approximately	approximately	ADV
ejpam-6087	621	11	95.76	95.76	NUM
ejpam-6087	621	12	%	%	NOUN
ejpam-6087	621	13	,	,	PUNCT
ejpam-6087	621	14	demonstrating	demonstrate	VERB
ejpam-6087	621	15	the	the	DET
ejpam-6087	621	16	best	good	ADJ
ejpam-6087	621	17	performance	performance	NOUN
ejpam-6087	621	18	in	in	ADP
ejpam-6087	621	19	distinguishing	distinguish	VERB
ejpam-6087	621	20	between	between	ADP
ejpam-6087	621	21	positive	positive	ADJ
ejpam-6087	621	22	and	and	CCONJ
ejpam-6087	621	23	negative	negative	ADJ
ejpam-6087	621	24	instances	instance	NOUN
ejpam-6087	621	25	.	.	PUNCT
ejpam-6087	622	1	lr	lr	PROPN
ejpam-6087	622	2	follows	follow	VERB
ejpam-6087	622	3	closely	closely	ADV
ejpam-6087	622	4	with	with	ADP
ejpam-6087	622	5	a	a	DET
ejpam-6087	622	6	score	score	NOUN
ejpam-6087	622	7	of	of	ADP
ejpam-6087	622	8	around	around	ADP
ejpam-6087	622	9	95.85	95.85	NUM
ejpam-6087	622	10	%	%	NOUN
ejpam-6087	622	11	,	,	PUNCT
ejpam-6087	622	12	which	which	PRON
ejpam-6087	622	13	also	also	ADV
ejpam-6087	622	14	indicates	indicate	VERB
ejpam-6087	622	15	strong	strong	ADJ
ejpam-6087	622	16	performance	performance	NOUN
ejpam-6087	622	17	.	.	PUNCT
ejpam-6087	623	1	on	on	ADP
ejpam-6087	623	2	the	the	DET
ejpam-6087	623	3	other	other	ADJ
ejpam-6087	623	4	hand	hand	NOUN
ejpam-6087	623	5	,	,	PUNCT
ejpam-6087	623	6	knn	knn	PROPN
ejpam-6087	623	7	and	and	CCONJ
ejpam-6087	623	8	svm	svm	PROPN
ejpam-6087	623	9	exhibit	exhibit	VERB
ejpam-6087	623	10	lower	low	ADJ
ejpam-6087	623	11	auc	auc	NOUN
ejpam-6087	623	12	-	-	PUNCT
ejpam-6087	623	13	roc	roc	NOUN
ejpam-6087	623	14	scores	score	NOUN
ejpam-6087	623	15	of	of	ADP
ejpam-6087	623	16	about	about	ADP
ejpam-6087	623	17	58.99	58.99	NUM
ejpam-6087	623	18	%	%	NOUN
ejpam-6087	623	19	and	and	CCONJ
ejpam-6087	623	20	55.53	55.53	NUM
ejpam-6087	623	21	%	%	NOUN
ejpam-6087	623	22	,	,	PUNCT
ejpam-6087	623	23	respectively	respectively	ADV
ejpam-6087	623	24	,	,	PUNCT
ejpam-6087	623	25	reflecting	reflect	VERB
ejpam-6087	623	26	their	their	PRON
ejpam-6087	623	27	relatively	relatively	ADV
ejpam-6087	623	28	weaker	weak	ADJ
ejpam-6087	623	29	performance	performance	NOUN
ejpam-6087	623	30	in	in	ADP
ejpam-6087	623	31	class	class	NOUN
ejpam-6087	623	32	discrimination	discrimination	NOUN
ejpam-6087	623	33	.	.	PUNCT
ejpam-6087	624	1	nb	nb	PROPN
ejpam-6087	624	2	achieves	achieve	VERB
ejpam-6087	624	3	a	a	DET
ejpam-6087	624	4	moderate	moderate	ADJ
ejpam-6087	624	5	aucroc	aucroc	NOUN
ejpam-6087	624	6	score	score	NOUN
ejpam-6087	624	7	of	of	ADP
ejpam-6087	624	8	approximately	approximately	ADV
ejpam-6087	624	9	69.07	69.07	NUM
ejpam-6087	624	10	%	%	NOUN
ejpam-6087	624	11	,	,	PUNCT
ejpam-6087	624	12	while	while	SCONJ
ejpam-6087	624	13	dt	dt	PROPN
ejpam-6087	624	14	shows	show	VERB
ejpam-6087	624	15	an	an	DET
ejpam-6087	624	16	auc	auc	NOUN
ejpam-6087	624	17	-	-	PUNCT
ejpam-6087	624	18	roc	roc	NOUN
ejpam-6087	624	19	score	score	NOUN
ejpam-6087	624	20	of	of	ADP
ejpam-6087	624	21	approximately	approximately	ADV
ejpam-6087	624	22	90.72	90.72	NUM
ejpam-6087	624	23	%	%	NOUN
ejpam-6087	624	24	,	,	PUNCT
ejpam-6087	624	25	which	which	PRON
ejpam-6087	624	26	is	be	AUX
ejpam-6087	624	27	lower	low	ADJ
ejpam-6087	624	28	compared	compare	VERB
ejpam-6087	624	29	to	to	ADP
ejpam-6087	624	30	lr	lr	NOUN
ejpam-6087	624	31	and	and	CCONJ
ejpam-6087	624	32	rf	rf	NOUN
ejpam-6087	624	33	but	but	CCONJ
ejpam-6087	624	34	still	still	ADV
ejpam-6087	624	35	reflects	reflect	VERB
ejpam-6087	624	36	reasonable	reasonable	ADJ
ejpam-6087	624	37	performance	performance	NOUN
ejpam-6087	624	38	.	.	PUNCT
ejpam-6087	625	1	in	in	ADP
ejpam-6087	625	2	general	general	ADJ
ejpam-6087	625	3	,	,	PUNCT
ejpam-6087	625	4	the	the	DET
ejpam-6087	625	5	results	result	NOUN
ejpam-6087	625	6	highlight	highlight	VERB
ejpam-6087	625	7	that	that	SCONJ
ejpam-6087	625	8	lr	lr	PROPN
ejpam-6087	625	9	and	and	CCONJ
ejpam-6087	625	10	rf	rf	NOUN
ejpam-6087	625	11	demonstrate	demonstrate	VERB
ejpam-6087	625	12	the	the	DET
ejpam-6087	625	13	highest	high	ADJ
ejpam-6087	625	14	effectiveness	effectiveness	NOUN
ejpam-6087	625	15	in	in	ADP
ejpam-6087	625	16	classifying	classify	VERB
ejpam-6087	625	17	instances	instance	NOUN
ejpam-6087	625	18	,	,	PUNCT
ejpam-6087	625	19	whereas	whereas	SCONJ
ejpam-6087	625	20	knn	knn	PROPN
ejpam-6087	625	21	and	and	CCONJ
ejpam-6087	625	22	svm	svm	PROPN
ejpam-6087	625	23	show	show	VERB
ejpam-6087	625	24	relatively	relatively	ADV
ejpam-6087	625	25	poor	poor	ADJ
ejpam-6087	625	26	performance	performance	NOUN
ejpam-6087	625	27	.	.	PUNCT
ejpam-6087	626	1	4.4	4.4	NUM
ejpam-6087	626	2	.	.	PUNCT
ejpam-6087	627	1	summary	summary	NOUN
ejpam-6087	627	2	of	of	ADP
ejpam-6087	627	3	key	key	ADJ
ejpam-6087	627	4	insights	insight	NOUN
ejpam-6087	627	5	the	the	DET
ejpam-6087	627	6	evaluation	evaluation	NOUN
ejpam-6087	627	7	of	of	ADP
ejpam-6087	627	8	feature	feature	NOUN
ejpam-6087	627	9	selection	selection	NOUN
ejpam-6087	627	10	methods	method	NOUN
ejpam-6087	627	11	—	—	PUNCT
ejpam-6087	627	12	rfe	rfe	NOUN
ejpam-6087	627	13	,	,	PUNCT
ejpam-6087	627	14	selectkbest	selectkb	ADJ
ejpam-6087	627	15	,	,	PUNCT
ejpam-6087	627	16	pca	pca	PROPN
ejpam-6087	627	17	,	,	PUNCT
ejpam-6087	627	18	and	and	CCONJ
ejpam-6087	627	19	ga	ga	PROPN
ejpam-6087	627	20	—	—	PUNCT
ejpam-6087	627	21	reveals	reveal	VERB
ejpam-6087	627	22	several	several	ADJ
ejpam-6087	627	23	important	important	ADJ
ejpam-6087	627	24	trends	trend	NOUN
ejpam-6087	627	25	in	in	ADP
ejpam-6087	627	26	how	how	SCONJ
ejpam-6087	627	27	these	these	DET
ejpam-6087	627	28	techniques	technique	NOUN
ejpam-6087	627	29	influence	influence	NOUN
ejpam-6087	627	30	model	model	NOUN
ejpam-6087	627	31	performance	performance	NOUN
ejpam-6087	627	32	.	.	PUNCT
ejpam-6087	628	1	this	this	DET
ejpam-6087	628	2	section	section	NOUN
ejpam-6087	628	3	discusses	discuss	VERB
ejpam-6087	628	4	the	the	DET
ejpam-6087	628	5	general	general	ADJ
ejpam-6087	628	6	insights	insight	NOUN
ejpam-6087	628	7	gained	gain	VERB
ejpam-6087	628	8	from	from	ADP
ejpam-6087	628	9	these	these	DET
ejpam-6087	628	10	methods	method	NOUN
ejpam-6087	628	11	and	and	CCONJ
ejpam-6087	628	12	highlights	highlight	NOUN
ejpam-6087	628	13	the	the	DET
ejpam-6087	628	14	stability	stability	NOUN
ejpam-6087	628	15	of	of	ADP
ejpam-6087	628	16	different	different	ADJ
ejpam-6087	628	17	ml	ml	NOUN
ejpam-6087	628	18	models	model	NOUN
ejpam-6087	628	19	across	across	ADP
ejpam-6087	628	20	various	various	ADJ
ejpam-6087	628	21	feature	feature	NOUN
ejpam-6087	628	22	selection	selection	NOUN
ejpam-6087	628	23	techniques	technique	NOUN
ejpam-6087	628	24	.	.	PUNCT
ejpam-6087	629	1	in	in	ADP
ejpam-6087	629	2	general	general	ADJ
ejpam-6087	629	3	,	,	PUNCT
ejpam-6087	629	4	the	the	DET
ejpam-6087	629	5	findings	finding	NOUN
ejpam-6087	629	6	of	of	ADP
ejpam-6087	629	7	this	this	DET
ejpam-6087	629	8	study	study	NOUN
ejpam-6087	629	9	can	can	AUX
ejpam-6087	629	10	be	be	AUX
ejpam-6087	629	11	summarized	summarize	VERB
ejpam-6087	629	12	as	as	ADP
ejpam-6087	629	13	the	the	DET
ejpam-6087	629	14	following	follow	VERB
ejpam-6087	629	15	key	key	ADJ
ejpam-6087	629	16	insights	insight	NOUN
ejpam-6087	629	17	.	.	PUNCT
ejpam-6087	630	1	m.	m.	PROPN
ejpam-6087	630	2	al	al	PROPN
ejpam-6087	630	3	-	-	PUNCT
ejpam-6087	630	4	tarawneh	tarawneh	PROPN
ejpam-6087	630	5	et	et	PROPN
ejpam-6087	630	6	al	al	PROPN
ejpam-6087	630	7	.	.	PUNCT
ejpam-6087	630	8	/	/	SYM
ejpam-6087	630	9	eur	eur	PROPN
ejpam-6087	630	10	.	.	PUNCT
ejpam-6087	631	1	j.	j.	PROPN
ejpam-6087	631	2	pure	pure	PROPN
ejpam-6087	631	3	appl	appl	PROPN
ejpam-6087	631	4	.	.	PROPN
ejpam-6087	631	5	math	math	PROPN
ejpam-6087	631	6	,	,	PUNCT
ejpam-6087	631	7	18	18	NUM
ejpam-6087	631	8	(	(	PUNCT
ejpam-6087	631	9	2	2	NUM
ejpam-6087	631	10	)	)	PUNCT
ejpam-6087	631	11	(	(	PUNCT
ejpam-6087	631	12	2025	2025	NUM
ejpam-6087	631	13	)	)	PUNCT
ejpam-6087	631	14	,	,	PUNCT
ejpam-6087	631	15	6087	6087	NUM
ejpam-6087	631	16	34	34	NUM
ejpam-6087	631	17	of	of	ADP
ejpam-6087	631	18	39	39	NUM
ejpam-6087	631	19	•	•	NOUN
ejpam-6087	631	20	performance	performance	NOUN
ejpam-6087	631	21	stability	stability	NOUN
ejpam-6087	631	22	across	across	ADP
ejpam-6087	631	23	feature	feature	NOUN
ejpam-6087	631	24	selection	selection	NOUN
ejpam-6087	631	25	methods	method	NOUN
ejpam-6087	631	26	:	:	PUNCT
ejpam-6087	631	27	rf	rf	NUM
ejpam-6087	631	28	and	and	CCONJ
ejpam-6087	631	29	lr	lr	NOUN
ejpam-6087	631	30	are	be	AUX
ejpam-6087	631	31	notable	notable	ADJ
ejpam-6087	631	32	for	for	ADP
ejpam-6087	631	33	their	their	PRON
ejpam-6087	631	34	stable	stable	ADJ
ejpam-6087	631	35	performance	performance	NOUN
ejpam-6087	631	36	across	across	ADP
ejpam-6087	631	37	different	different	ADJ
ejpam-6087	631	38	feature	feature	NOUN
ejpam-6087	631	39	selection	selection	NOUN
ejpam-6087	631	40	methods	method	NOUN
ejpam-6087	631	41	.	.	PUNCT
ejpam-6087	632	1	rf	rf	VERB
ejpam-6087	632	2	,	,	PUNCT
ejpam-6087	632	3	in	in	ADP
ejpam-6087	632	4	particular	particular	ADJ
ejpam-6087	632	5	,	,	PUNCT
ejpam-6087	632	6	shows	show	VERB
ejpam-6087	632	7	high	high	ADJ
ejpam-6087	632	8	and	and	CCONJ
ejpam-6087	632	9	consistent	consistent	ADJ
ejpam-6087	632	10	performance	performance	NOUN
ejpam-6087	632	11	across	across	ADP
ejpam-6087	632	12	rfe	rfe	NOUN
ejpam-6087	632	13	,	,	PUNCT
ejpam-6087	632	14	selectkbest	selectkb	ADJ
ejpam-6087	632	15	,	,	PUNCT
ejpam-6087	632	16	and	and	CCONJ
ejpam-6087	632	17	ga	ga	NOUN
ejpam-6087	632	18	methods	method	NOUN
ejpam-6087	632	19	.	.	PUNCT
ejpam-6087	633	1	for	for	ADP
ejpam-6087	633	2	example	example	NOUN
ejpam-6087	633	3	,	,	PUNCT
ejpam-6087	633	4	rf	rf	PRON
ejpam-6087	633	5	achieves	achieve	VERB
ejpam-6087	633	6	the	the	DET
ejpam-6087	633	7	highest	high	ADJ
ejpam-6087	633	8	auc	auc	NOUN
ejpam-6087	633	9	-	-	PUNCT
ejpam-6087	633	10	roc	roc	NOUN
ejpam-6087	633	11	score	score	NOUN
ejpam-6087	633	12	of	of	ADP
ejpam-6087	633	13	95.85	95.85	NUM
ejpam-6087	633	14	%	%	NOUN
ejpam-6087	633	15	under	under	ADP
ejpam-6087	633	16	the	the	DET
ejpam-6087	633	17	ga	ga	PROPN
ejpam-6087	633	18	method	method	NOUN
ejpam-6087	633	19	,	,	PUNCT
ejpam-6087	633	20	demonstrating	demonstrate	VERB
ejpam-6087	633	21	its	its	PRON
ejpam-6087	633	22	robustness	robustness	NOUN
ejpam-6087	633	23	in	in	ADP
ejpam-6087	633	24	handling	handle	VERB
ejpam-6087	633	25	different	different	ADJ
ejpam-6087	633	26	subsets	subset	NOUN
ejpam-6087	633	27	of	of	ADP
ejpam-6087	633	28	features	feature	NOUN
ejpam-6087	633	29	.	.	PUNCT
ejpam-6087	634	1	however	however	ADV
ejpam-6087	634	2	,	,	PUNCT
ejpam-6087	634	3	while	while	SCONJ
ejpam-6087	634	4	lr	lr	NOUN
ejpam-6087	634	5	maintains	maintain	VERB
ejpam-6087	634	6	high	high	ADJ
ejpam-6087	634	7	performance	performance	NOUN
ejpam-6087	634	8	under	under	ADP
ejpam-6087	634	9	selectkbest	selectkb	ADJ
ejpam-6087	634	10	,	,	PUNCT
ejpam-6087	634	11	it	it	PRON
ejpam-6087	634	12	is	be	AUX
ejpam-6087	634	13	less	less	ADV
ejpam-6087	634	14	consistent	consistent	ADJ
ejpam-6087	634	15	under	under	ADP
ejpam-6087	634	16	rfe	rfe	NOUN
ejpam-6087	634	17	and	and	CCONJ
ejpam-6087	634	18	pca	pca	PROPN
ejpam-6087	634	19	,	,	PUNCT
ejpam-6087	634	20	indicating	indicate	VERB
ejpam-6087	634	21	that	that	SCONJ
ejpam-6087	634	22	its	its	PRON
ejpam-6087	634	23	effectiveness	effectiveness	NOUN
ejpam-6087	634	24	is	be	AUX
ejpam-6087	634	25	more	more	ADV
ejpam-6087	634	26	dependent	dependent	ADJ
ejpam-6087	634	27	on	on	ADP
ejpam-6087	634	28	the	the	DET
ejpam-6087	634	29	feature	feature	NOUN
ejpam-6087	634	30	selection	selection	NOUN
ejpam-6087	634	31	process	process	NOUN
ejpam-6087	634	32	.	.	PUNCT
ejpam-6087	635	1	•	•	NUM
ejpam-6087	635	2	effectiveness	effectiveness	NOUN
ejpam-6087	635	3	of	of	ADP
ejpam-6087	635	4	dimensionality	dimensionality	NOUN
ejpam-6087	635	5	reduction	reduction	NOUN
ejpam-6087	635	6	techniques	technique	NOUN
ejpam-6087	635	7	:	:	PUNCT
ejpam-6087	635	8	the	the	DET
ejpam-6087	635	9	pca	pca	NOUN
ejpam-6087	635	10	technique	technique	NOUN
ejpam-6087	635	11	generally	generally	ADV
ejpam-6087	635	12	proves	prove	VERB
ejpam-6087	635	13	less	less	ADV
ejpam-6087	635	14	effective	effective	ADJ
ejpam-6087	635	15	compared	compare	VERB
ejpam-6087	635	16	to	to	ADP
ejpam-6087	635	17	other	other	ADJ
ejpam-6087	635	18	feature	feature	NOUN
ejpam-6087	635	19	selection	selection	NOUN
ejpam-6087	635	20	methods	method	NOUN
ejpam-6087	635	21	.	.	PUNCT
ejpam-6087	636	1	pca	pca	PROPN
ejpam-6087	636	2	’s	’s	PART
ejpam-6087	636	3	dimensionality	dimensionality	NOUN
ejpam-6087	636	4	reduction	reduction	NOUN
ejpam-6087	636	5	often	often	ADV
ejpam-6087	636	6	leads	lead	VERB
ejpam-6087	636	7	to	to	ADP
ejpam-6087	636	8	decreased	decrease	VERB
ejpam-6087	636	9	model	model	NOUN
ejpam-6087	636	10	performance	performance	NOUN
ejpam-6087	636	11	due	due	ADP
ejpam-6087	636	12	to	to	ADP
ejpam-6087	636	13	the	the	DET
ejpam-6087	636	14	loss	loss	NOUN
ejpam-6087	636	15	of	of	ADP
ejpam-6087	636	16	important	important	ADJ
ejpam-6087	636	17	feature	feature	NOUN
ejpam-6087	636	18	information	information	NOUN
ejpam-6087	636	19	and	and	CCONJ
ejpam-6087	636	20	interpretability	interpretability	NOUN
ejpam-6087	636	21	.	.	PUNCT
ejpam-6087	637	1	this	this	PRON
ejpam-6087	637	2	is	be	AUX
ejpam-6087	637	3	in	in	ADP
ejpam-6087	637	4	contrast	contrast	NOUN
ejpam-6087	637	5	to	to	ADP
ejpam-6087	637	6	rfe	rfe	VERB
ejpam-6087	637	7	and	and	CCONJ
ejpam-6087	637	8	selectkbest	selectkb	ADJ
ejpam-6087	637	9	,	,	PUNCT
ejpam-6087	637	10	which	which	PRON
ejpam-6087	637	11	focus	focus	VERB
ejpam-6087	637	12	on	on	ADP
ejpam-6087	637	13	selecting	select	VERB
ejpam-6087	637	14	the	the	DET
ejpam-6087	637	15	most	most	ADV
ejpam-6087	637	16	relevant	relevant	ADJ
ejpam-6087	637	17	features	feature	NOUN
ejpam-6087	637	18	and	and	CCONJ
ejpam-6087	637	19	thereby	thereby	ADV
ejpam-6087	637	20	improve	improve	VERB
ejpam-6087	637	21	model	model	NOUN
ejpam-6087	637	22	performance	performance	NOUN
ejpam-6087	637	23	.	.	PUNCT
ejpam-6087	638	1	in	in	ADP
ejpam-6087	638	2	this	this	DET
ejpam-6087	638	3	study	study	NOUN
ejpam-6087	638	4	,	,	PUNCT
ejpam-6087	638	5	pca	pca	PROPN
ejpam-6087	638	6	fails	fail	VERB
ejpam-6087	638	7	to	to	PART
ejpam-6087	638	8	effectively	effectively	ADV
ejpam-6087	638	9	enhance	enhance	VERB
ejpam-6087	638	10	model	model	NOUN
ejpam-6087	638	11	performance	performance	NOUN
ejpam-6087	638	12	,	,	PUNCT
ejpam-6087	638	13	reinforcing	reinforce	VERB
ejpam-6087	638	14	the	the	DET
ejpam-6087	638	15	idea	idea	NOUN
ejpam-6087	638	16	that	that	SCONJ
ejpam-6087	638	17	pca	pca	NOUN
ejpam-6087	638	18	is	be	AUX
ejpam-6087	638	19	less	less	ADV
ejpam-6087	638	20	suitable	suitable	ADJ
ejpam-6087	638	21	for	for	ADP
ejpam-6087	638	22	tasks	task	NOUN
ejpam-6087	638	23	that	that	PRON
ejpam-6087	638	24	require	require	VERB
ejpam-6087	638	25	feature	feature	NOUN
ejpam-6087	638	26	relevance	relevance	NOUN
ejpam-6087	638	27	and	and	CCONJ
ejpam-6087	638	28	interpretability	interpretability	NOUN
ejpam-6087	638	29	.	.	PUNCT
ejpam-6087	639	1	while	while	SCONJ
ejpam-6087	639	2	principal	principal	ADJ
ejpam-6087	639	3	component	component	NOUN
ejpam-6087	639	4	analysis	analysis	NOUN
ejpam-6087	639	5	is	be	AUX
ejpam-6087	639	6	well	well	ADV
ejpam-6087	639	7	regarded	regard	VERB
ejpam-6087	639	8	for	for	ADP
ejpam-6087	639	9	its	its	PRON
ejpam-6087	639	10	dimensionality	dimensionality	NOUN
ejpam-6087	639	11	reduction	reduction	NOUN
ejpam-6087	639	12	capabilities	capability	NOUN
ejpam-6087	639	13	,	,	PUNCT
ejpam-6087	639	14	it	it	PRON
ejpam-6087	639	15	inherently	inherently	ADV
ejpam-6087	639	16	compromises	compromise	VERB
ejpam-6087	639	17	feature	feature	NOUN
ejpam-6087	639	18	interpretability	interpretability	NOUN
ejpam-6087	639	19	,	,	PUNCT
ejpam-6087	639	20	which	which	PRON
ejpam-6087	639	21	can	can	AUX
ejpam-6087	639	22	hinder	hinder	VERB
ejpam-6087	639	23	explainability	explainability	NOUN
ejpam-6087	639	24	in	in	ADP
ejpam-6087	639	25	applications	application	NOUN
ejpam-6087	639	26	such	such	ADJ
ejpam-6087	639	27	as	as	ADP
ejpam-6087	639	28	fake	fake	ADJ
ejpam-6087	639	29	news	news	NOUN
ejpam-6087	639	30	detection	detection	NOUN
ejpam-6087	639	31	.	.	PUNCT
ejpam-6087	640	1	in	in	ADP
ejpam-6087	640	2	this	this	DET
ejpam-6087	640	3	study	study	NOUN
ejpam-6087	640	4	,	,	PUNCT
ejpam-6087	640	5	pca	pca	PROPN
ejpam-6087	640	6	was	be	AUX
ejpam-6087	640	7	intentionally	intentionally	ADV
ejpam-6087	640	8	included	include	VERB
ejpam-6087	640	9	as	as	ADP
ejpam-6087	640	10	a	a	DET
ejpam-6087	640	11	benchmarking	benchmarke	VERB
ejpam-6087	640	12	tool	tool	NOUN
ejpam-6087	640	13	to	to	PART
ejpam-6087	640	14	examine	examine	VERB
ejpam-6087	640	15	the	the	DET
ejpam-6087	640	16	trade	trade	NOUN
ejpam-6087	640	17	-	-	PUNCT
ejpam-6087	640	18	offs	off	NOUN
ejpam-6087	640	19	between	between	ADP
ejpam-6087	640	20	dimensionality	dimensionality	NOUN
ejpam-6087	640	21	reduction	reduction	NOUN
ejpam-6087	640	22	and	and	CCONJ
ejpam-6087	640	23	interpretability	interpretability	NOUN
ejpam-6087	640	24	.	.	PUNCT
ejpam-6087	641	1	although	although	SCONJ
ejpam-6087	641	2	pca	pca	PROPN
ejpam-6087	641	3	contributed	contribute	VERB
ejpam-6087	641	4	to	to	ADP
ejpam-6087	641	5	performance	performance	NOUN
ejpam-6087	641	6	improvement	improvement	NOUN
ejpam-6087	641	7	in	in	ADP
ejpam-6087	641	8	limited	limited	ADJ
ejpam-6087	641	9	cases	case	NOUN
ejpam-6087	641	10	,	,	PUNCT
ejpam-6087	641	11	it	it	PRON
ejpam-6087	641	12	was	be	AUX
ejpam-6087	641	13	consistently	consistently	ADV
ejpam-6087	641	14	outperformed	outperform	VERB
ejpam-6087	641	15	by	by	ADP
ejpam-6087	641	16	rfe	rfe	PROPN
ejpam-6087	641	17	and	and	CCONJ
ejpam-6087	641	18	ga	ga	PROPN
ejpam-6087	641	19	,	,	PUNCT
ejpam-6087	641	20	which	which	PRON
ejpam-6087	641	21	not	not	PART
ejpam-6087	641	22	only	only	ADV
ejpam-6087	641	23	yielded	yield	VERB
ejpam-6087	641	24	better	well	ADJ
ejpam-6087	641	25	classification	classification	NOUN
ejpam-6087	641	26	results	result	NOUN
ejpam-6087	641	27	,	,	PUNCT
ejpam-6087	641	28	but	but	CCONJ
ejpam-6087	641	29	also	also	ADV
ejpam-6087	641	30	preserved	preserve	VERB
ejpam-6087	641	31	transparency	transparency	NOUN
ejpam-6087	641	32	at	at	ADP
ejpam-6087	641	33	the	the	DET
ejpam-6087	641	34	feature	feature	NOUN
ejpam-6087	641	35	level	level	NOUN
ejpam-6087	641	36	.	.	PUNCT
ejpam-6087	642	1	these	these	DET
ejpam-6087	642	2	findings	finding	NOUN
ejpam-6087	642	3	reinforce	reinforce	VERB
ejpam-6087	642	4	the	the	DET
ejpam-6087	642	5	practical	practical	ADJ
ejpam-6087	642	6	advantages	advantage	NOUN
ejpam-6087	642	7	of	of	ADP
ejpam-6087	642	8	using	use	VERB
ejpam-6087	642	9	interpretable	interpretable	ADJ
ejpam-6087	642	10	feature	feature	NOUN
ejpam-6087	642	11	selection	selection	NOUN
ejpam-6087	642	12	techniques	technique	NOUN
ejpam-6087	642	13	for	for	ADP
ejpam-6087	642	14	fake	fake	ADJ
ejpam-6087	642	15	news	news	NOUN
ejpam-6087	642	16	detection	detection	NOUN
ejpam-6087	642	17	.	.	PUNCT
ejpam-6087	643	1	•	•	NOUN
ejpam-6087	643	2	success	success	NOUN
ejpam-6087	643	3	of	of	ADP
ejpam-6087	643	4	genetic	genetic	ADJ
ejpam-6087	643	5	algorithms	algorithm	NOUN
ejpam-6087	643	6	:	:	PUNCT
ejpam-6087	643	7	ga	ga	NOUN
ejpam-6087	643	8	emerges	emerge	VERB
ejpam-6087	643	9	as	as	ADP
ejpam-6087	643	10	a	a	DET
ejpam-6087	643	11	highly	highly	ADV
ejpam-6087	643	12	effective	effective	ADJ
ejpam-6087	643	13	feature	feature	NOUN
ejpam-6087	643	14	selection	selection	NOUN
ejpam-6087	643	15	technique	technique	NOUN
ejpam-6087	643	16	.	.	PUNCT
ejpam-6087	644	1	the	the	DET
ejpam-6087	644	2	ability	ability	NOUN
ejpam-6087	644	3	of	of	ADP
ejpam-6087	644	4	ga	ga	PROPN
ejpam-6087	644	5	to	to	PART
ejpam-6087	644	6	explore	explore	VERB
ejpam-6087	644	7	a	a	DET
ejpam-6087	644	8	wide	wide	ADJ
ejpam-6087	644	9	range	range	NOUN
ejpam-6087	644	10	of	of	ADP
ejpam-6087	644	11	feature	feature	NOUN
ejpam-6087	644	12	combinations	combination	NOUN
ejpam-6087	644	13	allows	allow	VERB
ejpam-6087	644	14	it	it	PRON
ejpam-6087	644	15	to	to	PART
ejpam-6087	644	16	identify	identify	VERB
ejpam-6087	644	17	optimal	optimal	ADJ
ejpam-6087	644	18	subsets	subset	NOUN
ejpam-6087	644	19	that	that	PRON
ejpam-6087	644	20	significantly	significantly	ADV
ejpam-6087	644	21	enhance	enhance	VERB
ejpam-6087	644	22	the	the	DET
ejpam-6087	644	23	performance	performance	NOUN
ejpam-6087	644	24	of	of	ADP
ejpam-6087	644	25	the	the	DET
ejpam-6087	644	26	model	model	NOUN
ejpam-6087	644	27	.	.	PUNCT
ejpam-6087	645	1	this	this	DET
ejpam-6087	645	2	effectiveness	effectiveness	NOUN
ejpam-6087	645	3	is	be	AUX
ejpam-6087	645	4	demonstrated	demonstrate	VERB
ejpam-6087	645	5	by	by	ADP
ejpam-6087	645	6	rf	rf	PRON
ejpam-6087	645	7	achieving	achieve	VERB
ejpam-6087	645	8	the	the	DET
ejpam-6087	645	9	highest	high	ADJ
ejpam-6087	645	10	aucroc	aucroc	NOUN
ejpam-6087	645	11	score	score	NOUN
ejpam-6087	645	12	of	of	ADP
ejpam-6087	645	13	95.85	95.85	NUM
ejpam-6087	645	14	%	%	NOUN
ejpam-6087	645	15	with	with	ADP
ejpam-6087	645	16	the	the	DET
ejpam-6087	645	17	ga	ga	PROPN
ejpam-6087	645	18	method	method	NOUN
ejpam-6087	645	19	.	.	PUNCT
ejpam-6087	646	1	ga	ga	PROPN
ejpam-6087	646	2	’s	’s	PART
ejpam-6087	646	3	success	success	NOUN
ejpam-6087	646	4	highlights	highlight	VERB
ejpam-6087	646	5	its	its	PRON
ejpam-6087	646	6	potential	potential	NOUN
ejpam-6087	646	7	as	as	ADP
ejpam-6087	646	8	a	a	DET
ejpam-6087	646	9	powerful	powerful	ADJ
ejpam-6087	646	10	tool	tool	NOUN
ejpam-6087	646	11	for	for	ADP
ejpam-6087	646	12	feature	feature	NOUN
ejpam-6087	646	13	selection	selection	NOUN
ejpam-6087	646	14	,	,	PUNCT
ejpam-6087	646	15	providing	provide	VERB
ejpam-6087	646	16	a	a	DET
ejpam-6087	646	17	robust	robust	ADJ
ejpam-6087	646	18	approach	approach	NOUN
ejpam-6087	646	19	to	to	PART
ejpam-6087	646	20	optimize	optimize	VERB
ejpam-6087	646	21	feature	feature	NOUN
ejpam-6087	646	22	subsets	subset	NOUN
ejpam-6087	646	23	and	and	CCONJ
ejpam-6087	646	24	improve	improve	VERB
ejpam-6087	646	25	classification	classification	NOUN
ejpam-6087	646	26	results	result	NOUN
ejpam-6087	646	27	.	.	PUNCT
ejpam-6087	647	1	in	in	ADP
ejpam-6087	647	2	summary	summary	NOUN
ejpam-6087	647	3	,	,	PUNCT
ejpam-6087	647	4	this	this	DET
ejpam-6087	647	5	study	study	NOUN
ejpam-6087	647	6	reveals	reveal	VERB
ejpam-6087	647	7	that	that	SCONJ
ejpam-6087	647	8	different	different	ADJ
ejpam-6087	647	9	feature	feature	NOUN
ejpam-6087	647	10	selection	selection	NOUN
ejpam-6087	647	11	methods	method	NOUN
ejpam-6087	647	12	have	have	VERB
ejpam-6087	647	13	different	different	ADJ
ejpam-6087	647	14	impacts	impact	NOUN
ejpam-6087	647	15	on	on	ADP
ejpam-6087	647	16	model	model	NOUN
ejpam-6087	647	17	performance	performance	NOUN
ejpam-6087	647	18	.	.	PUNCT
ejpam-6087	648	1	rfe	rfe	NOUN
ejpam-6087	648	2	and	and	CCONJ
ejpam-6087	648	3	selectkbest	selectkb	ADJ
ejpam-6087	648	4	demonstrate	demonstrate	NOUN
ejpam-6087	648	5	effectiveness	effectiveness	NOUN
ejpam-6087	648	6	in	in	ADP
ejpam-6087	648	7	improving	improve	VERB
ejpam-6087	648	8	model	model	NOUN
ejpam-6087	648	9	outcomes	outcome	NOUN
ejpam-6087	648	10	,	,	PUNCT
ejpam-6087	648	11	whereas	whereas	SCONJ
ejpam-6087	648	12	pca	pca	PROPN
ejpam-6087	648	13	tends	tend	VERB
ejpam-6087	648	14	to	to	PART
ejpam-6087	648	15	degrade	degrade	VERB
ejpam-6087	648	16	performance	performance	NOUN
ejpam-6087	648	17	due	due	ADP
ejpam-6087	648	18	to	to	ADP
ejpam-6087	648	19	reduced	reduce	VERB
ejpam-6087	648	20	feature	feature	NOUN
ejpam-6087	648	21	relevance	relevance	NOUN
ejpam-6087	648	22	and	and	CCONJ
ejpam-6087	648	23	interpretability	interpretability	NOUN
ejpam-6087	648	24	.	.	PUNCT
ejpam-6087	649	1	among	among	ADP
ejpam-6087	649	2	the	the	DET
ejpam-6087	649	3	models	model	NOUN
ejpam-6087	649	4	evaluated	evaluate	VERB
ejpam-6087	649	5	,	,	PUNCT
ejpam-6087	649	6	random	random	ADJ
ejpam-6087	649	7	forest	forest	NOUN
ejpam-6087	649	8	(	(	PUNCT
ejpam-6087	649	9	rf	rf	NOUN
ejpam-6087	649	10	)	)	PUNCT
ejpam-6087	649	11	consistently	consistently	ADV
ejpam-6087	649	12	shows	show	VERB
ejpam-6087	649	13	high	high	ADJ
ejpam-6087	649	14	reliability	reliability	NOUN
ejpam-6087	649	15	and	and	CCONJ
ejpam-6087	649	16	strong	strong	ADJ
ejpam-6087	649	17	performance	performance	NOUN
ejpam-6087	649	18	across	across	ADP
ejpam-6087	649	19	all	all	DET
ejpam-6087	649	20	feature	feature	NOUN
ejpam-6087	649	21	selection	selection	NOUN
ejpam-6087	649	22	techniques	technique	NOUN
ejpam-6087	649	23	,	,	PUNCT
ejpam-6087	649	24	highlighting	highlight	VERB
ejpam-6087	649	25	its	its	PRON
ejpam-6087	649	26	robustness	robustness	NOUN
ejpam-6087	649	27	.	.	PUNCT
ejpam-6087	650	1	logistic	logistic	ADJ
ejpam-6087	650	2	regression	regression	NOUN
ejpam-6087	650	3	(	(	PUNCT
ejpam-6087	650	4	lr	lr	NOUN
ejpam-6087	650	5	)	)	PUNCT
ejpam-6087	650	6	also	also	ADV
ejpam-6087	650	7	performs	perform	VERB
ejpam-6087	650	8	competitively	competitively	ADV
ejpam-6087	650	9	,	,	PUNCT
ejpam-6087	650	10	but	but	CCONJ
ejpam-6087	650	11	exhibits	exhibit	VERB
ejpam-6087	650	12	greater	great	ADJ
ejpam-6087	650	13	sensitivity	sensitivity	NOUN
ejpam-6087	650	14	to	to	ADP
ejpam-6087	650	15	the	the	DET
ejpam-6087	650	16	selected	select	VERB
ejpam-6087	650	17	feature	feature	NOUN
ejpam-6087	650	18	selection	selection	NOUN
ejpam-6087	650	19	method	method	NOUN
ejpam-6087	650	20	.	.	PUNCT
ejpam-6087	651	1	genetic	genetic	ADJ
ejpam-6087	651	2	algorithms	algorithm	NOUN
ejpam-6087	651	3	(	(	PUNCT
ejpam-6087	651	4	gas	gas	NOUN
ejpam-6087	651	5	)	)	PUNCT
ejpam-6087	651	6	further	far	ADV
ejpam-6087	651	7	prove	prove	VERB
ejpam-6087	651	8	effective	effective	ADJ
ejpam-6087	651	9	in	in	ADP
ejpam-6087	651	10	optimizing	optimize	VERB
ejpam-6087	651	11	feature	feature	NOUN
ejpam-6087	651	12	subsets	subset	NOUN
ejpam-6087	651	13	,	,	PUNCT
ejpam-6087	651	14	as	as	SCONJ
ejpam-6087	651	15	evidenced	evidence	VERB
ejpam-6087	651	16	by	by	ADP
ejpam-6087	651	17	the	the	DET
ejpam-6087	651	18	m.	m.	NOUN
ejpam-6087	651	19	al	al	PROPN
ejpam-6087	651	20	-	-	PUNCT
ejpam-6087	651	21	tarawneh	tarawneh	PROPN
ejpam-6087	651	22	et	et	PROPN
ejpam-6087	651	23	al	al	PROPN
ejpam-6087	651	24	.	.	PUNCT
ejpam-6087	651	25	/	/	SYM
ejpam-6087	651	26	eur	eur	PROPN
ejpam-6087	651	27	.	.	PUNCT
ejpam-6087	652	1	j.	j.	PROPN
ejpam-6087	652	2	pure	pure	PROPN
ejpam-6087	652	3	appl	appl	PROPN
ejpam-6087	652	4	.	.	PROPN
ejpam-6087	652	5	math	math	PROPN
ejpam-6087	652	6	,	,	PUNCT
ejpam-6087	652	7	18	18	NUM
ejpam-6087	652	8	(	(	PUNCT
ejpam-6087	652	9	2	2	NUM
ejpam-6087	652	10	)	)	PUNCT
ejpam-6087	652	11	(	(	PUNCT
ejpam-6087	652	12	2025	2025	NUM
ejpam-6087	652	13	)	)	PUNCT
ejpam-6087	652	14	,	,	PUNCT
ejpam-6087	652	15	6087	6087	NUM
ejpam-6087	652	16	35	35	NUM
ejpam-6087	652	17	of	of	ADP
ejpam-6087	652	18	39	39	NUM
ejpam-6087	652	19	superior	superior	ADJ
ejpam-6087	652	20	results	result	NOUN
ejpam-6087	652	21	obtained	obtain	VERB
ejpam-6087	652	22	when	when	SCONJ
ejpam-6087	652	23	combined	combine	VERB
ejpam-6087	652	24	with	with	ADP
ejpam-6087	652	25	rf	rf	NOUN
ejpam-6087	652	26	.	.	PUNCT
ejpam-6087	653	1	these	these	DET
ejpam-6087	653	2	findings	finding	NOUN
ejpam-6087	653	3	emphasize	emphasize	VERB
ejpam-6087	653	4	the	the	DET
ejpam-6087	653	5	importance	importance	NOUN
ejpam-6087	653	6	of	of	ADP
ejpam-6087	653	7	pairing	pair	VERB
ejpam-6087	653	8	strong	strong	ADJ
ejpam-6087	653	9	classifiers	classifier	NOUN
ejpam-6087	653	10	with	with	ADP
ejpam-6087	653	11	interpretable	interpretable	ADJ
ejpam-6087	653	12	and	and	CCONJ
ejpam-6087	653	13	computationally	computationally	ADV
ejpam-6087	653	14	efficient	efficient	ADJ
ejpam-6087	653	15	feature	feature	NOUN
ejpam-6087	653	16	selection	selection	NOUN
ejpam-6087	653	17	strategies	strategy	NOUN
ejpam-6087	653	18	for	for	ADP
ejpam-6087	653	19	fake	fake	ADJ
ejpam-6087	653	20	news	news	NOUN
ejpam-6087	653	21	detection	detection	NOUN
ejpam-6087	653	22	tasks	task	NOUN
ejpam-6087	653	23	.	.	PUNCT
ejpam-6087	654	1	furthermore	furthermore	ADV
ejpam-6087	654	2	,	,	PUNCT
ejpam-6087	654	3	the	the	DET
ejpam-6087	654	4	evolving	evolve	VERB
ejpam-6087	654	5	nature	nature	NOUN
ejpam-6087	654	6	of	of	ADP
ejpam-6087	654	7	fake	fake	ADJ
ejpam-6087	654	8	news	news	NOUN
ejpam-6087	654	9	poses	pose	VERB
ejpam-6087	654	10	a	a	DET
ejpam-6087	654	11	significant	significant	ADJ
ejpam-6087	654	12	challenge	challenge	NOUN
ejpam-6087	654	13	to	to	ADP
ejpam-6087	654	14	static	static	ADJ
ejpam-6087	654	15	models	model	NOUN
ejpam-6087	654	16	,	,	PUNCT
ejpam-6087	654	17	particularly	particularly	ADV
ejpam-6087	654	18	in	in	ADP
ejpam-6087	654	19	the	the	DET
ejpam-6087	654	20	face	face	NOUN
ejpam-6087	654	21	of	of	ADP
ejpam-6087	654	22	adversarial	adversarial	ADJ
ejpam-6087	654	23	content	content	NOUN
ejpam-6087	654	24	designed	design	VERB
ejpam-6087	654	25	to	to	PART
ejpam-6087	654	26	bypass	bypass	VERB
ejpam-6087	654	27	traditional	traditional	ADJ
ejpam-6087	654	28	detection	detection	NOUN
ejpam-6087	654	29	mechanisms	mechanism	NOUN
ejpam-6087	654	30	.	.	PUNCT
ejpam-6087	655	1	although	although	SCONJ
ejpam-6087	655	2	the	the	DET
ejpam-6087	655	3	current	current	ADJ
ejpam-6087	655	4	approach	approach	NOUN
ejpam-6087	655	5	relies	rely	VERB
ejpam-6087	655	6	on	on	ADP
ejpam-6087	655	7	offline	offline	ADJ
ejpam-6087	655	8	learning	learning	NOUN
ejpam-6087	655	9	,	,	PUNCT
ejpam-6087	655	10	these	these	DET
ejpam-6087	655	11	observations	observation	NOUN
ejpam-6087	655	12	point	point	VERB
ejpam-6087	655	13	to	to	ADP
ejpam-6087	655	14	the	the	DET
ejpam-6087	655	15	need	need	NOUN
ejpam-6087	655	16	to	to	PART
ejpam-6087	655	17	explore	explore	VERB
ejpam-6087	655	18	incremental	incremental	ADJ
ejpam-6087	655	19	and	and	CCONJ
ejpam-6087	655	20	adaptive	adaptive	ADJ
ejpam-6087	655	21	machine	machine	NOUN
ejpam-6087	655	22	learning	learning	NOUN
ejpam-6087	655	23	methods	method	NOUN
ejpam-6087	655	24	that	that	PRON
ejpam-6087	655	25	can	can	AUX
ejpam-6087	655	26	dynamically	dynamically	ADV
ejpam-6087	655	27	adjust	adjust	VERB
ejpam-6087	655	28	to	to	ADP
ejpam-6087	655	29	new	new	ADJ
ejpam-6087	655	30	patterns	pattern	NOUN
ejpam-6087	655	31	and	and	CCONJ
ejpam-6087	655	32	threats	threat	NOUN
ejpam-6087	655	33	.	.	PUNCT
ejpam-6087	656	1	such	such	ADJ
ejpam-6087	656	2	approaches	approach	NOUN
ejpam-6087	656	3	have	have	VERB
ejpam-6087	656	4	the	the	DET
ejpam-6087	656	5	potential	potential	NOUN
ejpam-6087	656	6	to	to	PART
ejpam-6087	656	7	improve	improve	VERB
ejpam-6087	656	8	the	the	DET
ejpam-6087	656	9	long	long	ADJ
ejpam-6087	656	10	-	-	PUNCT
ejpam-6087	656	11	term	term	NOUN
ejpam-6087	656	12	effectiveness	effectiveness	NOUN
ejpam-6087	656	13	and	and	CCONJ
ejpam-6087	656	14	resilience	resilience	NOUN
ejpam-6087	656	15	of	of	ADP
ejpam-6087	656	16	fake	fake	ADJ
ejpam-6087	656	17	news	news	NOUN
ejpam-6087	656	18	detection	detection	NOUN
ejpam-6087	656	19	systems	system	NOUN
ejpam-6087	656	20	.	.	PUNCT
ejpam-6087	657	1	in	in	ADP
ejpam-6087	657	2	addition	addition	NOUN
ejpam-6087	657	3	to	to	ADP
ejpam-6087	657	4	technical	technical	ADJ
ejpam-6087	657	5	considerations	consideration	NOUN
ejpam-6087	657	6	,	,	PUNCT
ejpam-6087	657	7	ethical	ethical	ADJ
ejpam-6087	657	8	implications	implication	NOUN
ejpam-6087	657	9	also	also	ADV
ejpam-6087	657	10	play	play	VERB
ejpam-6087	657	11	a	a	DET
ejpam-6087	657	12	crucial	crucial	ADJ
ejpam-6087	657	13	role	role	NOUN
ejpam-6087	657	14	in	in	ADP
ejpam-6087	657	15	the	the	DET
ejpam-6087	657	16	development	development	NOUN
ejpam-6087	657	17	and	and	CCONJ
ejpam-6087	657	18	deployment	deployment	NOUN
ejpam-6087	657	19	of	of	ADP
ejpam-6087	657	20	fake	fake	ADJ
ejpam-6087	657	21	news	news	NOUN
ejpam-6087	657	22	detection	detection	NOUN
ejpam-6087	657	23	systems	system	NOUN
ejpam-6087	657	24	.	.	PUNCT
ejpam-6087	658	1	a	a	DET
ejpam-6087	658	2	major	major	ADJ
ejpam-6087	658	3	concern	concern	NOUN
ejpam-6087	658	4	is	be	AUX
ejpam-6087	658	5	the	the	DET
ejpam-6087	658	6	possibility	possibility	NOUN
ejpam-6087	658	7	of	of	ADP
ejpam-6087	658	8	false	false	ADJ
ejpam-6087	658	9	positives	positive	NOUN
ejpam-6087	658	10	,	,	PUNCT
ejpam-6087	658	11	cases	case	NOUN
ejpam-6087	658	12	in	in	ADP
ejpam-6087	658	13	which	which	PRON
ejpam-6087	658	14	true	true	ADJ
ejpam-6087	658	15	information	information	NOUN
ejpam-6087	658	16	is	be	AUX
ejpam-6087	658	17	incorrectly	incorrectly	ADV
ejpam-6087	658	18	flagged	flag	VERB
ejpam-6087	658	19	as	as	ADP
ejpam-6087	658	20	fake	fake	ADJ
ejpam-6087	658	21	,	,	PUNCT
ejpam-6087	658	22	which	which	PRON
ejpam-6087	658	23	can	can	AUX
ejpam-6087	658	24	suppress	suppress	VERB
ejpam-6087	658	25	valid	valid	ADJ
ejpam-6087	658	26	discourse	discourse	NOUN
ejpam-6087	658	27	,	,	PUNCT
ejpam-6087	658	28	damage	damage	NOUN
ejpam-6087	658	29	reputations	reputation	NOUN
ejpam-6087	658	30	,	,	PUNCT
ejpam-6087	658	31	and	and	CCONJ
ejpam-6087	658	32	undermine	undermine	VERB
ejpam-6087	658	33	public	public	ADJ
ejpam-6087	658	34	trust	trust	NOUN
ejpam-6087	658	35	.	.	PUNCT
ejpam-6087	659	1	to	to	PART
ejpam-6087	659	2	address	address	VERB
ejpam-6087	659	3	this	this	PRON
ejpam-6087	659	4	,	,	PUNCT
ejpam-6087	659	5	the	the	DET
ejpam-6087	659	6	study	study	NOUN
ejpam-6087	659	7	incorporated	incorporate	VERB
ejpam-6087	659	8	a	a	DET
ejpam-6087	659	9	rigorous	rigorous	ADJ
ejpam-6087	659	10	evaluation	evaluation	NOUN
ejpam-6087	659	11	framework	framework	NOUN
ejpam-6087	659	12	using	use	VERB
ejpam-6087	659	13	a	a	DET
ejpam-6087	659	14	comprehensive	comprehensive	ADJ
ejpam-6087	659	15	set	set	NOUN
ejpam-6087	659	16	of	of	ADP
ejpam-6087	659	17	performance	performance	NOUN
ejpam-6087	659	18	metrics	metric	NOUN
ejpam-6087	659	19	,	,	PUNCT
ejpam-6087	659	20	including	include	VERB
ejpam-6087	659	21	precision	precision	NOUN
ejpam-6087	659	22	,	,	PUNCT
ejpam-6087	659	23	recall	recall	NOUN
ejpam-6087	659	24	,	,	PUNCT
ejpam-6087	659	25	f1	f1	NOUN
ejpam-6087	659	26	score	score	NOUN
ejpam-6087	659	27	,	,	PUNCT
ejpam-6087	659	28	and	and	CCONJ
ejpam-6087	659	29	auc	auc	NOUN
ejpam-6087	659	30	-	-	PUNCT
ejpam-6087	659	31	roc	roc	NOUN
ejpam-6087	659	32	,	,	PUNCT
ejpam-6087	659	33	to	to	PART
ejpam-6087	659	34	ensure	ensure	VERB
ejpam-6087	659	35	a	a	DET
ejpam-6087	659	36	balanced	balanced	ADJ
ejpam-6087	659	37	understanding	understanding	NOUN
ejpam-6087	659	38	of	of	ADP
ejpam-6087	659	39	model	model	NOUN
ejpam-6087	659	40	behavior	behavior	NOUN
ejpam-6087	659	41	.	.	PUNCT
ejpam-6087	660	1	however	however	ADV
ejpam-6087	660	2	,	,	PUNCT
ejpam-6087	660	3	technical	technical	ADJ
ejpam-6087	660	4	accuracy	accuracy	NOUN
ejpam-6087	660	5	alone	alone	ADV
ejpam-6087	660	6	is	be	AUX
ejpam-6087	660	7	not	not	PART
ejpam-6087	660	8	sufficient	sufficient	ADJ
ejpam-6087	660	9	.	.	PUNCT
ejpam-6087	661	1	responsible	responsible	ADJ
ejpam-6087	661	2	deployment	deployment	NOUN
ejpam-6087	661	3	should	should	AUX
ejpam-6087	661	4	consider	consider	VERB
ejpam-6087	661	5	transparency	transparency	NOUN
ejpam-6087	661	6	,	,	PUNCT
ejpam-6087	661	7	accountability	accountability	NOUN
ejpam-6087	661	8	,	,	PUNCT
ejpam-6087	661	9	and	and	CCONJ
ejpam-6087	661	10	,	,	PUNCT
ejpam-6087	661	11	where	where	SCONJ
ejpam-6087	661	12	appropriate	appropriate	ADJ
ejpam-6087	661	13	,	,	PUNCT
ejpam-6087	661	14	human	human	ADJ
ejpam-6087	661	15	oversight	oversight	NOUN
ejpam-6087	661	16	.	.	PUNCT
ejpam-6087	662	1	future	future	ADJ
ejpam-6087	662	2	work	work	NOUN
ejpam-6087	662	3	could	could	AUX
ejpam-6087	662	4	benefit	benefit	VERB
ejpam-6087	662	5	from	from	ADP
ejpam-6087	662	6	incorporating	incorporate	VERB
ejpam-6087	662	7	fairness	fairness	NOUN
ejpam-6087	662	8	-	-	PUNCT
ejpam-6087	662	9	aware	aware	ADJ
ejpam-6087	662	10	techniques	technique	NOUN
ejpam-6087	662	11	and	and	CCONJ
ejpam-6087	662	12	ethical	ethical	ADJ
ejpam-6087	662	13	design	design	NOUN
ejpam-6087	662	14	principles	principle	NOUN
ejpam-6087	662	15	to	to	PART
ejpam-6087	662	16	ensure	ensure	VERB
ejpam-6087	662	17	that	that	SCONJ
ejpam-6087	662	18	detection	detection	NOUN
ejpam-6087	662	19	systems	system	NOUN
ejpam-6087	662	20	are	be	AUX
ejpam-6087	662	21	not	not	PART
ejpam-6087	662	22	only	only	ADV
ejpam-6087	662	23	accurate	accurate	ADJ
ejpam-6087	662	24	but	but	CCONJ
ejpam-6087	662	25	also	also	ADV
ejpam-6087	662	26	aligned	align	VERB
ejpam-6087	662	27	with	with	ADP
ejpam-6087	662	28	societal	societal	ADJ
ejpam-6087	662	29	values	value	NOUN
ejpam-6087	662	30	.	.	PUNCT
ejpam-6087	663	1	5	5	X
ejpam-6087	663	2	.	.	X
ejpam-6087	663	3	conclusions	conclusion	NOUN
ejpam-6087	663	4	and	and	CCONJ
ejpam-6087	663	5	future	future	ADJ
ejpam-6087	663	6	work	work	NOUN
ejpam-6087	663	7	this	this	DET
ejpam-6087	663	8	study	study	NOUN
ejpam-6087	663	9	provided	provide	VERB
ejpam-6087	663	10	a	a	DET
ejpam-6087	663	11	comprehensive	comprehensive	ADJ
ejpam-6087	663	12	analysis	analysis	NOUN
ejpam-6087	663	13	of	of	ADP
ejpam-6087	663	14	various	various	ADJ
ejpam-6087	663	15	machine	machine	NOUN
ejpam-6087	663	16	learning	learn	VERB
ejpam-6087	663	17	classifiers	classifier	NOUN
ejpam-6087	663	18	and	and	CCONJ
ejpam-6087	663	19	feature	feature	NOUN
ejpam-6087	663	20	selection	selection	NOUN
ejpam-6087	663	21	techniques	technique	NOUN
ejpam-6087	663	22	to	to	PART
ejpam-6087	663	23	enhance	enhance	VERB
ejpam-6087	663	24	the	the	DET
ejpam-6087	663	25	detection	detection	NOUN
ejpam-6087	663	26	of	of	ADP
ejpam-6087	663	27	fake	fake	ADJ
ejpam-6087	663	28	news	news	NOUN
ejpam-6087	663	29	.	.	PUNCT
ejpam-6087	664	1	multiple	multiple	ADJ
ejpam-6087	664	2	models	model	NOUN
ejpam-6087	664	3	,	,	PUNCT
ejpam-6087	664	4	including	include	VERB
ejpam-6087	664	5	rf	rf	NOUN
ejpam-6087	664	6	,	,	PUNCT
ejpam-6087	664	7	gradient	gradient	NOUN
ejpam-6087	664	8	boosting	boosting	NOUN
ejpam-6087	664	9	,	,	PUNCT
ejpam-6087	664	10	and	and	CCONJ
ejpam-6087	664	11	svm	svm	ADJ
ejpam-6087	664	12	,	,	PUNCT
ejpam-6087	664	13	were	be	AUX
ejpam-6087	664	14	evaluated	evaluate	VERB
ejpam-6087	664	15	to	to	PART
ejpam-6087	664	16	identify	identify	VERB
ejpam-6087	664	17	the	the	DET
ejpam-6087	664	18	strengths	strength	NOUN
ejpam-6087	664	19	and	and	CCONJ
ejpam-6087	664	20	limitations	limitation	NOUN
ejpam-6087	664	21	of	of	ADP
ejpam-6087	664	22	each	each	DET
ejpam-6087	664	23	approach	approach	NOUN
ejpam-6087	664	24	in	in	ADP
ejpam-6087	664	25	the	the	DET
ejpam-6087	664	26	context	context	NOUN
ejpam-6087	664	27	of	of	ADP
ejpam-6087	664	28	detecting	detect	VERB
ejpam-6087	664	29	fake	fake	ADJ
ejpam-6087	664	30	news	news	NOUN
ejpam-6087	664	31	.	.	PUNCT
ejpam-6087	665	1	the	the	DET
ejpam-6087	665	2	findings	finding	NOUN
ejpam-6087	665	3	indicated	indicate	VERB
ejpam-6087	665	4	that	that	SCONJ
ejpam-6087	665	5	ensemble	ensemble	ADJ
ejpam-6087	665	6	learning	learning	NOUN
ejpam-6087	665	7	methods	method	NOUN
ejpam-6087	665	8	,	,	PUNCT
ejpam-6087	665	9	particularly	particularly	ADV
ejpam-6087	665	10	rf	rf	ADJ
ejpam-6087	665	11	and	and	CCONJ
ejpam-6087	665	12	gradient	gradient	ADJ
ejpam-6087	665	13	boosting	boosting	NOUN
ejpam-6087	665	14	,	,	PUNCT
ejpam-6087	665	15	achieved	achieve	VERB
ejpam-6087	665	16	superior	superior	ADJ
ejpam-6087	665	17	performance	performance	NOUN
ejpam-6087	665	18	in	in	ADP
ejpam-6087	665	19	terms	term	NOUN
ejpam-6087	665	20	of	of	ADP
ejpam-6087	665	21	accuracy	accuracy	NOUN
ejpam-6087	665	22	and	and	CCONJ
ejpam-6087	665	23	robustness	robustness	NOUN
ejpam-6087	665	24	.	.	PUNCT
ejpam-6087	666	1	this	this	DET
ejpam-6087	666	2	improvement	improvement	NOUN
ejpam-6087	666	3	was	be	AUX
ejpam-6087	666	4	especially	especially	ADV
ejpam-6087	666	5	pronounced	pronounce	VERB
ejpam-6087	666	6	when	when	SCONJ
ejpam-6087	666	7	these	these	DET
ejpam-6087	666	8	models	model	NOUN
ejpam-6087	666	9	were	be	AUX
ejpam-6087	666	10	paired	pair	VERB
ejpam-6087	666	11	with	with	ADP
ejpam-6087	666	12	effective	effective	ADJ
ejpam-6087	666	13	feature	feature	NOUN
ejpam-6087	666	14	selection	selection	NOUN
ejpam-6087	666	15	techniques	technique	NOUN
ejpam-6087	666	16	,	,	PUNCT
ejpam-6087	666	17	such	such	ADJ
ejpam-6087	666	18	as	as	ADP
ejpam-6087	666	19	rfe	rfe	NOUN
ejpam-6087	666	20	and	and	CCONJ
ejpam-6087	666	21	selectkbest	selectkb	ADJ
ejpam-6087	666	22	,	,	PUNCT
ejpam-6087	666	23	which	which	PRON
ejpam-6087	666	24	reduced	reduce	VERB
ejpam-6087	666	25	dimensionality	dimensionality	NOUN
ejpam-6087	666	26	and	and	CCONJ
ejpam-6087	666	27	enhanced	enhanced	ADJ
ejpam-6087	666	28	model	model	NOUN
ejpam-6087	666	29	performance	performance	NOUN
ejpam-6087	666	30	by	by	ADP
ejpam-6087	666	31	focusing	focus	VERB
ejpam-6087	666	32	on	on	ADP
ejpam-6087	666	33	the	the	DET
ejpam-6087	666	34	most	most	ADV
ejpam-6087	666	35	relevant	relevant	ADJ
ejpam-6087	666	36	features	feature	NOUN
ejpam-6087	666	37	.	.	PUNCT
ejpam-6087	667	1	the	the	DET
ejpam-6087	667	2	impact	impact	NOUN
ejpam-6087	667	3	of	of	ADP
ejpam-6087	667	4	different	different	ADJ
ejpam-6087	667	5	feature	feature	NOUN
ejpam-6087	667	6	selection	selection	NOUN
ejpam-6087	667	7	methods	method	NOUN
ejpam-6087	667	8	on	on	ADP
ejpam-6087	667	9	the	the	DET
ejpam-6087	667	10	effectiveness	effectiveness	NOUN
ejpam-6087	667	11	of	of	ADP
ejpam-6087	667	12	various	various	ADJ
ejpam-6087	667	13	classifiers	classifier	NOUN
ejpam-6087	667	14	was	be	AUX
ejpam-6087	667	15	thoroughly	thoroughly	ADV
ejpam-6087	667	16	examined	examine	VERB
ejpam-6087	667	17	,	,	PUNCT
ejpam-6087	667	18	demonstrating	demonstrate	VERB
ejpam-6087	667	19	that	that	SCONJ
ejpam-6087	667	20	techniques	technique	NOUN
ejpam-6087	667	21	like	like	ADP
ejpam-6087	667	22	rfe	rfe	NOUN
ejpam-6087	667	23	and	and	CCONJ
ejpam-6087	667	24	selectkbest	selectkb	ADJ
ejpam-6087	667	25	significantly	significantly	ADV
ejpam-6087	667	26	improved	improve	VERB
ejpam-6087	667	27	performance	performance	NOUN
ejpam-6087	667	28	.	.	PUNCT
ejpam-6087	668	1	in	in	ADP
ejpam-6087	668	2	contrast	contrast	NOUN
ejpam-6087	668	3	,	,	PUNCT
ejpam-6087	668	4	pca	pca	PROPN
ejpam-6087	668	5	was	be	AUX
ejpam-6087	668	6	found	find	VERB
ejpam-6087	668	7	to	to	PART
ejpam-6087	668	8	be	be	AUX
ejpam-6087	668	9	less	less	ADV
ejpam-6087	668	10	effective	effective	ADJ
ejpam-6087	668	11	,	,	PUNCT
ejpam-6087	668	12	underscoring	underscore	VERB
ejpam-6087	668	13	the	the	DET
ejpam-6087	668	14	importance	importance	NOUN
ejpam-6087	668	15	of	of	ADP
ejpam-6087	668	16	carefully	carefully	ADV
ejpam-6087	668	17	selecting	select	VERB
ejpam-6087	668	18	feature	feature	NOUN
ejpam-6087	668	19	extraction	extraction	NOUN
ejpam-6087	668	20	methods	method	NOUN
ejpam-6087	668	21	tailored	tailor	VERB
ejpam-6087	668	22	to	to	ADP
ejpam-6087	668	23	the	the	DET
ejpam-6087	668	24	specific	specific	ADJ
ejpam-6087	668	25	application	application	NOUN
ejpam-6087	668	26	.	.	PUNCT
ejpam-6087	669	1	in	in	ADP
ejpam-6087	669	2	conclusion	conclusion	NOUN
ejpam-6087	669	3	,	,	PUNCT
ejpam-6087	669	4	the	the	DET
ejpam-6087	669	5	integration	integration	NOUN
ejpam-6087	669	6	of	of	ADP
ejpam-6087	669	7	robust	robust	ADJ
ejpam-6087	669	8	ml	ml	NOUN
ejpam-6087	669	9	models	model	NOUN
ejpam-6087	669	10	with	with	ADP
ejpam-6087	669	11	optimized	optimize	VERB
ejpam-6087	669	12	feature	feature	NOUN
ejpam-6087	669	13	selection	selection	NOUN
ejpam-6087	669	14	strategies	strategy	NOUN
ejpam-6087	669	15	was	be	AUX
ejpam-6087	669	16	shown	show	VERB
ejpam-6087	669	17	to	to	PART
ejpam-6087	669	18	be	be	AUX
ejpam-6087	669	19	critical	critical	ADJ
ejpam-6087	669	20	to	to	ADP
ejpam-6087	669	21	achieving	achieve	VERB
ejpam-6087	669	22	high	high	ADJ
ejpam-6087	669	23	accuracy	accuracy	NOUN
ejpam-6087	669	24	and	and	CCONJ
ejpam-6087	669	25	reliability	reliability	NOUN
ejpam-6087	669	26	in	in	ADP
ejpam-6087	669	27	fake	fake	ADJ
ejpam-6087	669	28	news	news	NOUN
ejpam-6087	669	29	detection	detection	NOUN
ejpam-6087	669	30	.	.	PUNCT
ejpam-6087	670	1	this	this	DET
ejpam-6087	670	2	study	study	NOUN
ejpam-6087	670	3	provided	provide	VERB
ejpam-6087	670	4	valuable	valuable	ADJ
ejpam-6087	670	5	information	information	NOUN
ejpam-6087	670	6	on	on	ADP
ejpam-6087	670	7	the	the	DET
ejpam-6087	670	8	development	development	NOUN
ejpam-6087	670	9	of	of	ADP
ejpam-6087	670	10	more	more	ADV
ejpam-6087	670	11	effective	effective	ADJ
ejpam-6087	670	12	fake	fake	ADJ
ejpam-6087	670	13	news	news	NOUN
ejpam-6087	670	14	detection	detection	NOUN
ejpam-6087	670	15	systems	system	NOUN
ejpam-6087	670	16	and	and	CCONJ
ejpam-6087	670	17	contributed	contribute	VERB
ejpam-6087	670	18	to	to	ADP
ejpam-6087	670	19	the	the	DET
ejpam-6087	670	20	advancement	advancement	NOUN
ejpam-6087	670	21	of	of	ADP
ejpam-6087	670	22	methodologies	methodology	NOUN
ejpam-6087	670	23	in	in	ADP
ejpam-6087	670	24	this	this	DET
ejpam-6087	670	25	field	field	NOUN
ejpam-6087	670	26	.	.	PUNCT
ejpam-6087	671	1	while	while	SCONJ
ejpam-6087	671	2	this	this	DET
ejpam-6087	671	3	study	study	NOUN
ejpam-6087	671	4	demonstrated	demonstrate	VERB
ejpam-6087	671	5	the	the	DET
ejpam-6087	671	6	effectiveness	effectiveness	NOUN
ejpam-6087	671	7	of	of	ADP
ejpam-6087	671	8	various	various	ADJ
ejpam-6087	671	9	traditional	traditional	ADJ
ejpam-6087	671	10	and	and	CCONJ
ejpam-6087	671	11	ensemble	ensemble	ADJ
ejpam-6087	671	12	m.	m.	NOUN
ejpam-6087	671	13	al	al	PROPN
ejpam-6087	671	14	-	-	PUNCT
ejpam-6087	671	15	tarawneh	tarawneh	PROPN
ejpam-6087	671	16	et	et	PROPN
ejpam-6087	671	17	al	al	PROPN
ejpam-6087	671	18	.	.	PUNCT
ejpam-6087	671	19	/	/	SYM
ejpam-6087	671	20	eur	eur	PROPN
ejpam-6087	671	21	.	.	PUNCT
ejpam-6087	672	1	j.	j.	PROPN
ejpam-6087	672	2	pure	pure	PROPN
ejpam-6087	672	3	appl	appl	PROPN
ejpam-6087	672	4	.	.	PROPN
ejpam-6087	672	5	math	math	PROPN
ejpam-6087	672	6	,	,	PUNCT
ejpam-6087	672	7	18	18	NUM
ejpam-6087	672	8	(	(	PUNCT
ejpam-6087	672	9	2	2	NUM
ejpam-6087	672	10	)	)	PUNCT
ejpam-6087	672	11	(	(	PUNCT
ejpam-6087	672	12	2025	2025	NUM
ejpam-6087	672	13	)	)	PUNCT
ejpam-6087	672	14	,	,	PUNCT
ejpam-6087	672	15	6087	6087	NUM
ejpam-6087	672	16	36	36	NUM
ejpam-6087	672	17	of	of	ADP
ejpam-6087	672	18	39	39	NUM
ejpam-6087	672	19	classifiers	classifier	NOUN
ejpam-6087	672	20	combined	combine	VERB
ejpam-6087	672	21	with	with	ADP
ejpam-6087	672	22	feature	feature	NOUN
ejpam-6087	672	23	selection	selection	NOUN
ejpam-6087	672	24	techniques	technique	NOUN
ejpam-6087	672	25	,	,	PUNCT
ejpam-6087	672	26	future	future	ADJ
ejpam-6087	672	27	research	research	NOUN
ejpam-6087	672	28	could	could	AUX
ejpam-6087	672	29	explore	explore	VERB
ejpam-6087	672	30	additional	additional	ADJ
ejpam-6087	672	31	avenues	avenue	NOUN
ejpam-6087	672	32	to	to	PART
ejpam-6087	672	33	further	far	ADV
ejpam-6087	672	34	enhance	enhance	VERB
ejpam-6087	672	35	the	the	DET
ejpam-6087	672	36	detection	detection	NOUN
ejpam-6087	672	37	of	of	ADP
ejpam-6087	672	38	fake	fake	ADJ
ejpam-6087	672	39	news	news	NOUN
ejpam-6087	672	40	.	.	PUNCT
ejpam-6087	673	1	one	one	NUM
ejpam-6087	673	2	promising	promise	VERB
ejpam-6087	673	3	direction	direction	NOUN
ejpam-6087	673	4	involves	involve	VERB
ejpam-6087	673	5	the	the	DET
ejpam-6087	673	6	investigation	investigation	NOUN
ejpam-6087	673	7	of	of	ADP
ejpam-6087	673	8	incremental	incremental	ADJ
ejpam-6087	673	9	and	and	CCONJ
ejpam-6087	673	10	online	online	ADJ
ejpam-6087	673	11	learning	learning	NOUN
ejpam-6087	673	12	approaches	approach	NOUN
ejpam-6087	673	13	that	that	PRON
ejpam-6087	673	14	allow	allow	VERB
ejpam-6087	673	15	models	model	NOUN
ejpam-6087	673	16	to	to	PART
ejpam-6087	673	17	continuously	continuously	ADV
ejpam-6087	673	18	adapt	adapt	VERB
ejpam-6087	673	19	to	to	ADP
ejpam-6087	673	20	new	new	ADJ
ejpam-6087	673	21	data	datum	NOUN
ejpam-6087	673	22	.	.	PUNCT
ejpam-6087	674	1	given	give	VERB
ejpam-6087	674	2	the	the	DET
ejpam-6087	674	3	dynamic	dynamic	ADJ
ejpam-6087	674	4	and	and	CCONJ
ejpam-6087	674	5	evolving	evolve	VERB
ejpam-6087	674	6	nature	nature	NOUN
ejpam-6087	674	7	of	of	ADP
ejpam-6087	674	8	misinformation	misinformation	NOUN
ejpam-6087	674	9	,	,	PUNCT
ejpam-6087	674	10	such	such	ADJ
ejpam-6087	674	11	adaptive	adaptive	ADJ
ejpam-6087	674	12	techniques	technique	NOUN
ejpam-6087	674	13	can	can	AUX
ejpam-6087	674	14	improve	improve	VERB
ejpam-6087	674	15	response	response	NOUN
ejpam-6087	674	16	to	to	ADP
ejpam-6087	674	17	emerging	emerge	VERB
ejpam-6087	674	18	trends	trend	NOUN
ejpam-6087	674	19	without	without	ADP
ejpam-6087	674	20	the	the	DET
ejpam-6087	674	21	need	need	NOUN
ejpam-6087	674	22	for	for	ADP
ejpam-6087	674	23	complete	complete	ADJ
ejpam-6087	674	24	retraining	retraining	NOUN
ejpam-6087	674	25	.	.	PUNCT
ejpam-6087	675	1	additionally	additionally	ADV
ejpam-6087	675	2	,	,	PUNCT
ejpam-6087	675	3	evaluating	evaluate	VERB
ejpam-6087	675	4	model	model	NOUN
ejpam-6087	675	5	performance	performance	NOUN
ejpam-6087	675	6	across	across	ADP
ejpam-6087	675	7	multiple	multiple	ADJ
ejpam-6087	675	8	datasets	dataset	NOUN
ejpam-6087	675	9	from	from	ADP
ejpam-6087	675	10	different	different	ADJ
ejpam-6087	675	11	domains	domain	NOUN
ejpam-6087	675	12	and	and	CCONJ
ejpam-6087	675	13	platforms	platform	NOUN
ejpam-6087	675	14	could	could	AUX
ejpam-6087	675	15	help	help	VERB
ejpam-6087	675	16	assess	assess	VERB
ejpam-6087	675	17	generalizability	generalizability	NOUN
ejpam-6087	675	18	.	.	PUNCT
ejpam-6087	676	1	incorporating	incorporate	VERB
ejpam-6087	676	2	explainable	explainable	ADJ
ejpam-6087	676	3	ai	ai	NOUN
ejpam-6087	676	4	(	(	PUNCT
ejpam-6087	676	5	xai	xai	PROPN
ejpam-6087	676	6	)	)	PUNCT
ejpam-6087	676	7	techniques	technique	NOUN
ejpam-6087	676	8	can	can	AUX
ejpam-6087	676	9	also	also	ADV
ejpam-6087	676	10	strengthen	strengthen	VERB
ejpam-6087	676	11	the	the	DET
ejpam-6087	676	12	transparency	transparency	NOUN
ejpam-6087	676	13	and	and	CCONJ
ejpam-6087	676	14	trustworthiness	trustworthiness	NOUN
ejpam-6087	676	15	of	of	ADP
ejpam-6087	676	16	detection	detection	NOUN
ejpam-6087	676	17	systems	system	NOUN
ejpam-6087	676	18	,	,	PUNCT
ejpam-6087	676	19	particularly	particularly	ADV
ejpam-6087	676	20	in	in	ADP
ejpam-6087	676	21	high	high	ADJ
ejpam-6087	676	22	-	-	PUNCT
ejpam-6087	676	23	stakes	stake	NOUN
ejpam-6087	676	24	information	information	NOUN
ejpam-6087	676	25	environments	environment	NOUN
ejpam-6087	676	26	.	.	PUNCT
ejpam-6087	677	1	these	these	DET
ejpam-6087	677	2	directions	direction	NOUN
ejpam-6087	677	3	aim	aim	VERB
ejpam-6087	677	4	to	to	PART
ejpam-6087	677	5	build	build	VERB
ejpam-6087	677	6	on	on	ADP
ejpam-6087	677	7	the	the	DET
ejpam-6087	677	8	current	current	ADJ
ejpam-6087	677	9	work	work	NOUN
ejpam-6087	677	10	and	and	CCONJ
ejpam-6087	677	11	extend	extend	VERB
ejpam-6087	677	12	its	its	PRON
ejpam-6087	677	13	applicability	applicability	NOUN
ejpam-6087	677	14	to	to	ADP
ejpam-6087	677	15	real	real	ADJ
ejpam-6087	677	16	-	-	PUNCT
ejpam-6087	677	17	world	world	NOUN
ejpam-6087	677	18	,	,	PUNCT
ejpam-6087	677	19	evolving	evolve	VERB
ejpam-6087	677	20	contexts	contexts	NOUN
ejpam-6087	677	21	.	.	PUNCT
ejpam-6087	678	1	another	another	DET
ejpam-6087	678	2	potential	potential	ADJ
ejpam-6087	678	3	extension	extension	NOUN
ejpam-6087	678	4	is	be	AUX
ejpam-6087	678	5	the	the	DET
ejpam-6087	678	6	inclusion	inclusion	NOUN
ejpam-6087	678	7	of	of	ADP
ejpam-6087	678	8	a	a	DET
ejpam-6087	678	9	comparative	comparative	ADJ
ejpam-6087	678	10	analysis	analysis	NOUN
ejpam-6087	678	11	of	of	ADP
ejpam-6087	678	12	the	the	DET
ejpam-6087	678	13	computational	computational	ADJ
ejpam-6087	678	14	cost	cost	NOUN
ejpam-6087	678	15	associated	associate	VERB
ejpam-6087	678	16	with	with	ADP
ejpam-6087	678	17	different	different	ADJ
ejpam-6087	678	18	feature	feature	NOUN
ejpam-6087	678	19	selection	selection	NOUN
ejpam-6087	678	20	techniques	technique	NOUN
ejpam-6087	678	21	in	in	ADP
ejpam-6087	678	22	various	various	ADJ
ejpam-6087	678	23	types	type	NOUN
ejpam-6087	678	24	of	of	ADP
ejpam-6087	678	25	classifiers	classifier	NOUN
ejpam-6087	678	26	.	.	PUNCT
ejpam-6087	679	1	although	although	SCONJ
ejpam-6087	679	2	the	the	DET
ejpam-6087	679	3	present	present	ADJ
ejpam-6087	679	4	study	study	NOUN
ejpam-6087	679	5	focused	focus	VERB
ejpam-6087	679	6	primarily	primarily	ADV
ejpam-6087	679	7	on	on	ADP
ejpam-6087	679	8	classification	classification	NOUN
ejpam-6087	679	9	performance	performance	NOUN
ejpam-6087	679	10	and	and	CCONJ
ejpam-6087	679	11	given	give	VERB
ejpam-6087	679	12	that	that	SCONJ
ejpam-6087	679	13	the	the	DET
ejpam-6087	679	14	work	work	NOUN
ejpam-6087	679	15	does	do	AUX
ejpam-6087	679	16	not	not	PART
ejpam-6087	679	17	target	target	VERB
ejpam-6087	679	18	deployment	deployment	NOUN
ejpam-6087	679	19	in	in	ADP
ejpam-6087	679	20	resource	resource	NOUN
ejpam-6087	679	21	-	-	PUNCT
ejpam-6087	679	22	constrained	constrain	VERB
ejpam-6087	679	23	environments	environment	NOUN
ejpam-6087	679	24	such	such	ADJ
ejpam-6087	679	25	as	as	ADP
ejpam-6087	679	26	iot	iot	NOUN
ejpam-6087	679	27	or	or	CCONJ
ejpam-6087	679	28	edge	edge	NOUN
ejpam-6087	679	29	devices	device	NOUN
ejpam-6087	679	30	,	,	PUNCT
ejpam-6087	679	31	computational	computational	ADJ
ejpam-6087	679	32	efficiency	efficiency	NOUN
ejpam-6087	679	33	was	be	AUX
ejpam-6087	679	34	not	not	PART
ejpam-6087	679	35	emphasized	emphasize	VERB
ejpam-6087	679	36	.	.	PUNCT
ejpam-6087	680	1	however	however	ADV
ejpam-6087	680	2	,	,	PUNCT
ejpam-6087	680	3	such	such	ADJ
ejpam-6087	680	4	analysis	analysis	NOUN
ejpam-6087	680	5	would	would	AUX
ejpam-6087	680	6	be	be	AUX
ejpam-6087	680	7	highly	highly	ADV
ejpam-6087	680	8	valuable	valuable	ADJ
ejpam-6087	680	9	in	in	ADP
ejpam-6087	680	10	scenarios	scenario	NOUN
ejpam-6087	680	11	that	that	PRON
ejpam-6087	680	12	require	require	VERB
ejpam-6087	680	13	trade	trade	NOUN
ejpam-6087	680	14	-	-	PUNCT
ejpam-6087	680	15	offs	off	NOUN
ejpam-6087	680	16	between	between	ADP
ejpam-6087	680	17	accuracy	accuracy	NOUN
ejpam-6087	680	18	and	and	CCONJ
ejpam-6087	680	19	efficiency	efficiency	NOUN
ejpam-6087	680	20	,	,	PUNCT
ejpam-6087	680	21	and	and	CCONJ
ejpam-6087	680	22	could	could	AUX
ejpam-6087	680	23	inform	inform	VERB
ejpam-6087	680	24	deployment	deployment	NOUN
ejpam-6087	680	25	decisions	decision	NOUN
ejpam-6087	680	26	in	in	ADP
ejpam-6087	680	27	constrained	constrained	ADJ
ejpam-6087	680	28	environments	environment	NOUN
ejpam-6087	680	29	.	.	PUNCT
ejpam-6087	681	1	acknowledgements	acknowledgement	NOUN
ejpam-6087	681	2	machine	machine	NOUN
ejpam-6087	681	3	learning	learn	VERB
ejpam-6087	681	4	training	training	NOUN
ejpam-6087	681	5	and	and	CCONJ
ejpam-6087	681	6	evaluation	evaluation	NOUN
ejpam-6087	681	7	have	have	AUX
ejpam-6087	681	8	been	be	AUX
ejpam-6087	681	9	performed	perform	VERB
ejpam-6087	681	10	using	use	VERB
ejpam-6087	681	11	the	the	DET
ejpam-6087	681	12	phoenix	phoenix	PROPN
ejpam-6087	681	13	high	high	ADJ
ejpam-6087	681	14	performance	performance	NOUN
ejpam-6087	681	15	computing	computing	NOUN
ejpam-6087	681	16	facility	facility	NOUN
ejpam-6087	681	17	at	at	ADP
ejpam-6087	681	18	the	the	DET
ejpam-6087	681	19	american	american	PROPN
ejpam-6087	681	20	university	university	PROPN
ejpam-6087	681	21	of	of	ADP
ejpam-6087	681	22	the	the	DET
ejpam-6087	681	23	middle	middle	PROPN
ejpam-6087	681	24	east	east	PROPN
ejpam-6087	681	25	,	,	PUNCT
ejpam-6087	681	26	kuwait	kuwait	PROPN
ejpam-6087	681	27	.	.	PUNCT
ejpam-6087	682	1	references	reference	NOUN
ejpam-6087	682	2	[	[	X
ejpam-6087	682	3	1	1	NUM
ejpam-6087	682	4	]	]	PUNCT
ejpam-6087	682	5	femi	femi	PROPN
ejpam-6087	682	6	olan	olan	PROPN
ejpam-6087	682	7	,	,	PUNCT
ejpam-6087	682	8	uchitha	uchitha	PROPN
ejpam-6087	682	9	jayawickrama	jayawickrama	PROPN
ejpam-6087	682	10	,	,	PUNCT
ejpam-6087	682	11	emmanuel	emmanuel	PROPN
ejpam-6087	682	12	ogiemwonyi	ogiemwonyi	PROPN
ejpam-6087	682	13	arakpogun	arakpogun	PROPN
ejpam-6087	682	14	,	,	PUNCT
ejpam-6087	682	15	jana	jana	PROPN
ejpam-6087	682	16	suklan	suklan	PROPN
ejpam-6087	682	17	,	,	PUNCT
ejpam-6087	682	18	and	and	CCONJ
ejpam-6087	682	19	shaofeng	shaofeng	PROPN
ejpam-6087	682	20	liu	liu	PROPN
ejpam-6087	682	21	.	.	PUNCT
ejpam-6087	683	1	fake	fake	ADJ
ejpam-6087	683	2	news	news	NOUN
ejpam-6087	683	3	on	on	ADP
ejpam-6087	683	4	social	social	ADJ
ejpam-6087	683	5	media	medium	NOUN
ejpam-6087	683	6	:	:	PUNCT
ejpam-6087	683	7	the	the	DET
ejpam-6087	683	8	impact	impact	NOUN
ejpam-6087	683	9	on	on	ADP
ejpam-6087	683	10	society	society	NOUN
ejpam-6087	683	11	.	.	PUNCT
ejpam-6087	684	1	information	information	NOUN
ejpam-6087	684	2	systems	system	NOUN
ejpam-6087	684	3	frontiers	frontier	NOUN
ejpam-6087	684	4	,	,	PUNCT
ejpam-6087	684	5	26(2):443–458	26(2):443–458	NUM
ejpam-6087	684	6	,	,	PUNCT
ejpam-6087	684	7	apr	apr	NOUN
ejpam-6087	684	8	2024	2024	NUM
ejpam-6087	684	9	.	.	PUNCT
ejpam-6087	685	1	[	[	X
ejpam-6087	685	2	2	2	X
ejpam-6087	685	3	]	]	PUNCT
ejpam-6087	685	4	h.	h.	PROPN
ejpam-6087	685	5	allcott	allcott	PROPN
ejpam-6087	685	6	and	and	CCONJ
ejpam-6087	685	7	m.	m.	NOUN
ejpam-6087	685	8	gentzkow	gentzkow	PROPN
ejpam-6087	685	9	.	.	PUNCT
ejpam-6087	686	1	social	social	ADJ
ejpam-6087	686	2	media	medium	NOUN
ejpam-6087	686	3	and	and	CCONJ
ejpam-6087	686	4	fake	fake	ADJ
ejpam-6087	686	5	news	news	NOUN
ejpam-6087	686	6	in	in	ADP
ejpam-6087	686	7	the	the	DET
ejpam-6087	686	8	2016	2016	NUM
ejpam-6087	686	9	election	election	NOUN
ejpam-6087	686	10	.	.	PUNCT
ejpam-6087	687	1	journal	journal	NOUN
ejpam-6087	687	2	of	of	ADP
ejpam-6087	687	3	economic	economic	ADJ
ejpam-6087	687	4	perspectives	perspective	NOUN
ejpam-6087	687	5	,	,	PUNCT
ejpam-6087	687	6	31(2):211–236	31(2):211–236	PROPN
ejpam-6087	687	7	,	,	PUNCT
ejpam-6087	687	8	2017	2017	NUM
ejpam-6087	687	9	.	.	PUNCT
ejpam-6087	688	1	[	[	X
ejpam-6087	688	2	3	3	NUM
ejpam-6087	688	3	]	]	X
ejpam-6087	688	4	a.	a.	NOUN
ejpam-6087	688	5	gupta	gupta	PROPN
ejpam-6087	688	6	and	and	CCONJ
ejpam-6087	688	7	a.	a.	PROPN
ejpam-6087	688	8	roy	roy	PROPN
ejpam-6087	688	9	.	.	PROPN
ejpam-6087	688	10	manipulation	manipulation	NOUN
ejpam-6087	688	11	of	of	ADP
ejpam-6087	688	12	social	social	ADJ
ejpam-6087	688	13	media	medium	NOUN
ejpam-6087	688	14	during	during	ADP
ejpam-6087	688	15	the	the	DET
ejpam-6087	688	16	2019	2019	NUM
ejpam-6087	688	17	indian	indian	ADJ
ejpam-6087	688	18	general	general	ADJ
ejpam-6087	688	19	elections	election	NOUN
ejpam-6087	688	20	.	.	PUNCT
ejpam-6087	689	1	asian	asian	ADJ
ejpam-6087	689	2	journal	journal	NOUN
ejpam-6087	689	3	of	of	ADP
ejpam-6087	689	4	communication	communication	NOUN
ejpam-6087	689	5	,	,	PUNCT
ejpam-6087	689	6	29(5):537–556	29(5):537–556	NOUN
ejpam-6087	689	7	,	,	PUNCT
ejpam-6087	689	8	2019	2019	NUM
ejpam-6087	689	9	.	.	PUNCT
ejpam-6087	690	1	[	[	X
ejpam-6087	690	2	4	4	X
ejpam-6087	690	3	]	]	PUNCT
ejpam-6087	690	4	m.	m.	NOUN
ejpam-6087	690	5	cinelli	cinelli	PROPN
ejpam-6087	690	6	and	and	CCONJ
ejpam-6087	690	7	a.	a.	NOUN
ejpam-6087	690	8	galeazzi	galeazzi	NOUN
ejpam-6087	690	9	.	.	PUNCT
ejpam-6087	691	1	the	the	DET
ejpam-6087	691	2	covid-19	covid-19	PROPN
ejpam-6087	691	3	social	social	ADJ
ejpam-6087	691	4	media	medium	NOUN
ejpam-6087	691	5	infodemic	infodemic	ADJ
ejpam-6087	691	6	.	.	PUNCT
ejpam-6087	692	1	scientific	scientific	ADJ
ejpam-6087	692	2	reports	report	NOUN
ejpam-6087	692	3	,	,	PUNCT
ejpam-6087	692	4	10:1–10	10:1–10	NUM
ejpam-6087	692	5	,	,	PUNCT
ejpam-6087	692	6	2020	2020	NUM
ejpam-6087	692	7	.	.	PUNCT
ejpam-6087	693	1	[	[	X
ejpam-6087	693	2	5	5	NUM
ejpam-6087	693	3	]	]	X
ejpam-6087	693	4	omar	omar	PROPN
ejpam-6087	693	5	abdulwahabe	abdulwahabe	PROPN
ejpam-6087	693	6	mohamad	mohamad	PROPN
ejpam-6087	693	7	rasha	rasha	PROPN
ejpam-6087	693	8	talal	talal	PROPN
ejpam-6087	693	9	hameed	hameed	PROPN
ejpam-6087	693	10	.	.	PUNCT
ejpam-6087	694	1	federated	federated	ADJ
ejpam-6087	694	2	learning	learning	NOUN
ejpam-6087	694	3	in	in	ADP
ejpam-6087	694	4	iot	iot	NOUN
ejpam-6087	694	5	:	:	PUNCT
ejpam-6087	694	6	a	a	DET
ejpam-6087	694	7	survey	survey	NOUN
ejpam-6087	694	8	on	on	ADP
ejpam-6087	694	9	distributed	distribute	VERB
ejpam-6087	694	10	decision	decision	NOUN
ejpam-6087	694	11	making	making	NOUN
ejpam-6087	694	12	.	.	PUNCT
ejpam-6087	695	1	babylonian	babylonian	ADJ
ejpam-6087	695	2	journal	journal	PROPN
ejpam-6087	695	3	of	of	ADP
ejpam-6087	695	4	internet	internet	NOUN
ejpam-6087	695	5	of	of	ADP
ejpam-6087	695	6	things	thing	NOUN
ejpam-6087	695	7	,	,	PUNCT
ejpam-6087	695	8	2023:1–7	2023:1–7	PROPN
ejpam-6087	695	9	,	,	PUNCT
ejpam-6087	695	10	jan	jan	PROPN
ejpam-6087	695	11	.	.	PROPN
ejpam-6087	695	12	2023	2023	NUM
ejpam-6087	695	13	.	.	PUNCT
ejpam-6087	696	1	[	[	X
ejpam-6087	696	2	6	6	NUM
ejpam-6087	696	3	]	]	PUNCT
ejpam-6087	696	4	dipti	dipti	X
ejpam-6087	696	5	theng	theng	PROPN
ejpam-6087	696	6	and	and	CCONJ
ejpam-6087	696	7	kishor	kishor	PROPN
ejpam-6087	696	8	k.	k.	PROPN
ejpam-6087	696	9	bhoyar	bhoyar	PROPN
ejpam-6087	696	10	.	.	PUNCT
ejpam-6087	697	1	feature	feature	NOUN
ejpam-6087	697	2	selection	selection	NOUN
ejpam-6087	697	3	techniques	technique	NOUN
ejpam-6087	697	4	for	for	ADP
ejpam-6087	697	5	machine	machine	NOUN
ejpam-6087	697	6	learning	learning	NOUN
ejpam-6087	697	7	:	:	PUNCT
ejpam-6087	697	8	a	a	DET
ejpam-6087	697	9	survey	survey	NOUN
ejpam-6087	697	10	of	of	ADP
ejpam-6087	697	11	more	more	ADJ
ejpam-6087	697	12	than	than	ADP
ejpam-6087	697	13	two	two	NUM
ejpam-6087	697	14	decades	decade	NOUN
ejpam-6087	697	15	of	of	ADP
ejpam-6087	697	16	research	research	NOUN
ejpam-6087	697	17	.	.	PUNCT
ejpam-6087	698	1	knowledge	knowledge	NOUN
ejpam-6087	698	2	and	and	CCONJ
ejpam-6087	698	3	information	information	NOUN
ejpam-6087	698	4	systems	system	NOUN
ejpam-6087	698	5	,	,	PUNCT
ejpam-6087	698	6	66(3):1575–1637	66(3):1575–1637	PROPN
ejpam-6087	698	7	,	,	PUNCT
ejpam-6087	698	8	mar	mar	PROPN
ejpam-6087	698	9	2024	2024	NUM
ejpam-6087	698	10	.	.	PUNCT
ejpam-6087	699	1	[	[	X
ejpam-6087	699	2	7	7	X
ejpam-6087	699	3	]	]	X
ejpam-6087	699	4	i.	i.	NOUN
ejpam-6087	699	5	guyon	guyon	PROPN
ejpam-6087	699	6	and	and	CCONJ
ejpam-6087	699	7	a.	a.	PROPN
ejpam-6087	699	8	elisseeff	elisseeff	PROPN
ejpam-6087	699	9	.	.	PUNCT
ejpam-6087	700	1	an	an	DET
ejpam-6087	700	2	introduction	introduction	NOUN
ejpam-6087	700	3	to	to	ADP
ejpam-6087	700	4	variable	variable	ADJ
ejpam-6087	700	5	and	and	CCONJ
ejpam-6087	700	6	feature	feature	NOUN
ejpam-6087	700	7	selection	selection	NOUN
ejpam-6087	700	8	.	.	PUNCT
ejpam-6087	701	1	journal	journal	PROPN
ejpam-6087	701	2	of	of	ADP
ejpam-6087	701	3	machine	machine	NOUN
ejpam-6087	701	4	learning	learn	VERB
ejpam-6087	701	5	research	research	NOUN
ejpam-6087	701	6	,	,	PUNCT
ejpam-6087	701	7	3:1157–1182	3:1157–1182	PROPN
ejpam-6087	701	8	,	,	PUNCT
ejpam-6087	701	9	2003	2003	NUM
ejpam-6087	701	10	.	.	PUNCT
ejpam-6087	702	1	m.	m.	PROPN
ejpam-6087	702	2	al	al	PROPN
ejpam-6087	702	3	-	-	PUNCT
ejpam-6087	702	4	tarawneh	tarawneh	PROPN
ejpam-6087	702	5	et	et	PROPN
ejpam-6087	702	6	al	al	PROPN
ejpam-6087	702	7	.	.	PUNCT
ejpam-6087	702	8	/	/	SYM
ejpam-6087	702	9	eur	eur	PROPN
ejpam-6087	702	10	.	.	PUNCT
ejpam-6087	703	1	j.	j.	PROPN
ejpam-6087	703	2	pure	pure	PROPN
ejpam-6087	703	3	appl	appl	PROPN
ejpam-6087	703	4	.	.	PROPN
ejpam-6087	703	5	math	math	PROPN
ejpam-6087	703	6	,	,	PUNCT
ejpam-6087	703	7	18	18	NUM
ejpam-6087	703	8	(	(	PUNCT
ejpam-6087	703	9	2	2	NUM
ejpam-6087	703	10	)	)	PUNCT
ejpam-6087	703	11	(	(	PUNCT
ejpam-6087	703	12	2025	2025	NUM
ejpam-6087	703	13	)	)	PUNCT
ejpam-6087	703	14	,	,	PUNCT
ejpam-6087	703	15	6087	6087	NUM
ejpam-6087	703	16	37	37	NUM
ejpam-6087	703	17	of	of	ADP
ejpam-6087	703	18	39	39	NUM
ejpam-6087	703	19	[	[	SYM
ejpam-6087	703	20	8	8	NUM
ejpam-6087	703	21	]	]	X
ejpam-6087	703	22	girish	girish	ADJ
ejpam-6087	703	23	chandrashekar	chandrashekar	PROPN
ejpam-6087	703	24	and	and	CCONJ
ejpam-6087	703	25	ferat	ferat	PROPN
ejpam-6087	703	26	sahin	sahin	PROPN
ejpam-6087	703	27	.	.	PUNCT
ejpam-6087	704	1	a	a	DET
ejpam-6087	704	2	survey	survey	NOUN
ejpam-6087	704	3	on	on	ADP
ejpam-6087	704	4	feature	feature	NOUN
ejpam-6087	704	5	selection	selection	NOUN
ejpam-6087	704	6	methods	method	NOUN
ejpam-6087	704	7	.	.	PUNCT
ejpam-6087	705	1	computers	computer	NOUN
ejpam-6087	705	2	&	&	CCONJ
ejpam-6087	705	3	electrical	electrical	ADJ
ejpam-6087	705	4	engineering	engineering	NOUN
ejpam-6087	705	5	,	,	PUNCT
ejpam-6087	705	6	40(1):16–28	40(1):16–28	NUM
ejpam-6087	705	7	,	,	PUNCT
ejpam-6087	705	8	2014	2014	NUM
ejpam-6087	705	9	.	.	PUNCT
ejpam-6087	706	1	[	[	X
ejpam-6087	706	2	9	9	NUM
ejpam-6087	706	3	]	]	PUNCT
ejpam-6087	706	4	michael	michael	PROPN
ejpam-6087	706	5	greenacre	greenacre	PROPN
ejpam-6087	706	6	,	,	PUNCT
ejpam-6087	706	7	patrick	patrick	PROPN
ejpam-6087	706	8	j.	j.	PROPN
ejpam-6087	706	9	f.	f.	PROPN
ejpam-6087	706	10	groenen	groenen	PROPN
ejpam-6087	706	11	,	,	PUNCT
ejpam-6087	706	12	trevor	trevor	PROPN
ejpam-6087	706	13	hastie	hastie	PROPN
ejpam-6087	706	14	,	,	PUNCT
ejpam-6087	706	15	alfonso	alfonso	PROPN
ejpam-6087	706	16	iodice	iodice	PROPN
ejpam-6087	706	17	d’enza	d’enza	PROPN
ejpam-6087	706	18	,	,	PUNCT
ejpam-6087	706	19	angelos	angelos	PROPN
ejpam-6087	706	20	markos	markos	PROPN
ejpam-6087	706	21	,	,	PUNCT
ejpam-6087	706	22	and	and	CCONJ
ejpam-6087	706	23	elena	elena	PROPN
ejpam-6087	706	24	tuzhilina	tuzhilina	PROPN
ejpam-6087	706	25	.	.	PUNCT
ejpam-6087	706	26	principal	principal	ADJ
ejpam-6087	706	27	component	component	NOUN
ejpam-6087	706	28	analysis	analysis	NOUN
ejpam-6087	706	29	.	.	PUNCT
ejpam-6087	707	1	nature	nature	NOUN
ejpam-6087	707	2	reviews	review	VERB
ejpam-6087	707	3	methods	method	NOUN
ejpam-6087	707	4	primers	primer	NOUN
ejpam-6087	707	5	,	,	PUNCT
ejpam-6087	707	6	2(1):100	2(1):100	NUM
ejpam-6087	707	7	,	,	PUNCT
ejpam-6087	707	8	dec	dec	PROPN
ejpam-6087	707	9	2022	2022	NUM
ejpam-6087	707	10	.	.	PUNCT
ejpam-6087	708	1	[	[	X
ejpam-6087	708	2	10	10	NUM
ejpam-6087	708	3	]	]	X
ejpam-6087	708	4	rania	rania	PROPN
ejpam-6087	708	5	saidi	saidi	PROPN
ejpam-6087	708	6	,	,	PUNCT
ejpam-6087	708	7	waad	waad	PROPN
ejpam-6087	708	8	bouaguel	bouaguel	PROPN
ejpam-6087	708	9	ncir	ncir	PROPN
ejpam-6087	708	10	,	,	PUNCT
ejpam-6087	708	11	and	and	CCONJ
ejpam-6087	708	12	nadia	nadia	PROPN
ejpam-6087	708	13	essoussi	essoussi	PROPN
ejpam-6087	708	14	.	.	PUNCT
ejpam-6087	709	1	feature	feature	NOUN
ejpam-6087	709	2	selection	selection	NOUN
ejpam-6087	709	3	using	use	VERB
ejpam-6087	709	4	genetic	genetic	ADJ
ejpam-6087	709	5	algorithm	algorithm	NOUN
ejpam-6087	709	6	for	for	ADP
ejpam-6087	709	7	big	big	ADJ
ejpam-6087	709	8	data	datum	NOUN
ejpam-6087	709	9	.	.	PUNCT
ejpam-6087	710	1	in	in	ADP
ejpam-6087	710	2	the	the	DET
ejpam-6087	710	3	international	international	ADJ
ejpam-6087	710	4	conference	conference	NOUN
ejpam-6087	710	5	on	on	ADP
ejpam-6087	710	6	advanced	advanced	ADJ
ejpam-6087	710	7	machine	machine	NOUN
ejpam-6087	710	8	learning	learning	NOUN
ejpam-6087	710	9	technologies	technology	NOUN
ejpam-6087	710	10	and	and	CCONJ
ejpam-6087	710	11	applications	application	NOUN
ejpam-6087	710	12	(	(	PUNCT
ejpam-6087	710	13	amlta2018	amlta2018	NOUN
ejpam-6087	710	14	)	)	PUNCT
ejpam-6087	710	15	,	,	PUNCT
ejpam-6087	710	16	pages	page	NOUN
ejpam-6087	710	17	352–361	352–361	NUM
ejpam-6087	710	18	,	,	PUNCT
ejpam-6087	710	19	2018	2018	NUM
ejpam-6087	710	20	.	.	PUNCT
ejpam-6087	711	1	[	[	X
ejpam-6087	711	2	11	11	NUM
ejpam-6087	711	3	]	]	PUNCT
ejpam-6087	711	4	saad	saad	PROPN
ejpam-6087	711	5	munir	munir	PROPN
ejpam-6087	711	6	and	and	CCONJ
ejpam-6087	711	7	m.	m.	NOUN
ejpam-6087	711	8	asif	asif	PROPN
ejpam-6087	711	9	naeem	naeem	PROPN
ejpam-6087	711	10	.	.	PUNCT
ejpam-6087	712	1	bil	bil	VERB
ejpam-6087	712	2	-	-	PROPN
ejpam-6087	712	3	fand	fand	NOUN
ejpam-6087	712	4	:	:	PUNCT
ejpam-6087	712	5	leveraging	leverage	VERB
ejpam-6087	712	6	ensemble	ensemble	ADJ
ejpam-6087	712	7	technique	technique	NOUN
ejpam-6087	712	8	for	for	ADP
ejpam-6087	712	9	efficient	efficient	ADJ
ejpam-6087	712	10	bilingual	bilingual	ADJ
ejpam-6087	712	11	fake	fake	ADJ
ejpam-6087	712	12	news	news	NOUN
ejpam-6087	712	13	detection	detection	NOUN
ejpam-6087	712	14	.	.	PUNCT
ejpam-6087	713	1	international	international	ADJ
ejpam-6087	713	2	journal	journal	PROPN
ejpam-6087	713	3	of	of	ADP
ejpam-6087	713	4	machine	machine	NOUN
ejpam-6087	713	5	learning	learning	NOUN
ejpam-6087	713	6	and	and	CCONJ
ejpam-6087	713	7	cybernetics	cybernetic	NOUN
ejpam-6087	713	8	,	,	PUNCT
ejpam-6087	713	9	mar	mar	PROPN
ejpam-6087	713	10	2024	2024	NUM
ejpam-6087	713	11	.	.	PUNCT
ejpam-6087	714	1	[	[	X
ejpam-6087	714	2	12	12	NUM
ejpam-6087	714	3	]	]	X
ejpam-6087	714	4	ehtesham	ehtesham	PROPN
ejpam-6087	714	5	hashmi	hashmi	PROPN
ejpam-6087	714	6	,	,	PUNCT
ejpam-6087	714	7	sule	sule	PROPN
ejpam-6087	714	8	yildirim	yildirim	PROPN
ejpam-6087	714	9	yayilgan	yayilgan	PROPN
ejpam-6087	714	10	,	,	PUNCT
ejpam-6087	714	11	muhammad	muhammad	PROPN
ejpam-6087	714	12	mudassar	mudassar	PROPN
ejpam-6087	714	13	yamin	yamin	PROPN
ejpam-6087	714	14	,	,	PUNCT
ejpam-6087	714	15	subhan	subhan	PROPN
ejpam-6087	714	16	ali	ali	PROPN
ejpam-6087	714	17	,	,	PUNCT
ejpam-6087	714	18	and	and	CCONJ
ejpam-6087	714	19	mohamed	mohamed	PROPN
ejpam-6087	714	20	abomhara	abomhara	PROPN
ejpam-6087	714	21	.	.	PUNCT
ejpam-6087	715	1	advancing	advance	VERB
ejpam-6087	715	2	fake	fake	ADJ
ejpam-6087	715	3	news	news	NOUN
ejpam-6087	715	4	detection	detection	NOUN
ejpam-6087	715	5	:	:	PUNCT
ejpam-6087	715	6	hybrid	hybrid	ADJ
ejpam-6087	715	7	deep	deep	ADJ
ejpam-6087	715	8	learning	learning	NOUN
ejpam-6087	715	9	with	with	ADP
ejpam-6087	715	10	fasttext	fasttext	NOUN
ejpam-6087	715	11	and	and	CCONJ
ejpam-6087	715	12	explainable	explainable	ADJ
ejpam-6087	715	13	ai	ai	NOUN
ejpam-6087	715	14	.	.	PUNCT
ejpam-6087	715	15	ieee	ieee	NOUN
ejpam-6087	715	16	access	access	NOUN
ejpam-6087	715	17	,	,	PUNCT
ejpam-6087	715	18	12:44462–44480	12:44462–44480	NUM
ejpam-6087	715	19	,	,	PUNCT
ejpam-6087	715	20	2024	2024	NUM
ejpam-6087	715	21	.	.	PUNCT
ejpam-6087	716	1	[	[	X
ejpam-6087	716	2	13	13	NUM
ejpam-6087	716	3	]	]	X
ejpam-6087	716	4	mahabuba	mahabuba	PROPN
ejpam-6087	716	5	akhter	akhter	NOUN
ejpam-6087	716	6	,	,	PUNCT
ejpam-6087	716	7	syed	syed	PROPN
ejpam-6087	716	8	md	md	PROPN
ejpam-6087	716	9	.	.	PUNCT
ejpam-6087	716	10	minhaz	minhaz	PROPN
ejpam-6087	716	11	hossain	hossain	PROPN
ejpam-6087	716	12	,	,	PUNCT
ejpam-6087	716	13	rizma	rizma	PROPN
ejpam-6087	716	14	sijana	sijana	PROPN
ejpam-6087	716	15	nigar	nigar	PROPN
ejpam-6087	716	16	,	,	PUNCT
ejpam-6087	716	17	srabanti	srabanti	PROPN
ejpam-6087	716	18	paul	paul	PROPN
ejpam-6087	716	19	,	,	PUNCT
ejpam-6087	716	20	khaleque	khaleque	PROPN
ejpam-6087	716	21	md	md	PROPN
ejpam-6087	716	22	.	.	PROPN
ejpam-6087	716	23	aashiq	aashiq	PROPN
ejpam-6087	716	24	kamal	kamal	PROPN
ejpam-6087	716	25	,	,	PUNCT
ejpam-6087	716	26	anik	anik	PROPN
ejpam-6087	716	27	sen	sen	PROPN
ejpam-6087	716	28	,	,	PUNCT
ejpam-6087	716	29	and	and	CCONJ
ejpam-6087	716	30	iqbal	iqbal	PROPN
ejpam-6087	716	31	h.	h.	PROPN
ejpam-6087	716	32	sarker	sarker	PROPN
ejpam-6087	716	33	.	.	PUNCT
ejpam-6087	717	1	covid-19	covid-19	PROPN
ejpam-6087	717	2	fake	fake	ADJ
ejpam-6087	717	3	news	news	NOUN
ejpam-6087	717	4	detection	detection	NOUN
ejpam-6087	717	5	using	use	VERB
ejpam-6087	717	6	deep	deep	ADJ
ejpam-6087	717	7	learning	learning	NOUN
ejpam-6087	717	8	model	model	NOUN
ejpam-6087	717	9	.	.	PUNCT
ejpam-6087	718	1	annals	annal	NOUN
ejpam-6087	718	2	of	of	ADP
ejpam-6087	718	3	data	data	NOUN
ejpam-6087	718	4	science	science	NOUN
ejpam-6087	718	5	,	,	PUNCT
ejpam-6087	718	6	jan	jan	PROPN
ejpam-6087	718	7	2024	2024	NUM
ejpam-6087	718	8	.	.	PUNCT
ejpam-6087	719	1	[	[	X
ejpam-6087	719	2	14	14	NUM
ejpam-6087	719	3	]	]	X
ejpam-6087	719	4	chander	chander	NOUN
ejpam-6087	719	5	prabha	prabha	PROPN
ejpam-6087	719	6	,	,	PUNCT
ejpam-6087	719	7	meena	meena	PROPN
ejpam-6087	719	8	malik	malik	PROPN
ejpam-6087	719	9	,	,	PUNCT
ejpam-6087	719	10	shalini	shalini	PROPN
ejpam-6087	719	11	kumari	kumari	PROPN
ejpam-6087	719	12	,	,	PUNCT
ejpam-6087	719	13	neha	neha	NOUN
ejpam-6087	719	14	arya	arya	NOUN
ejpam-6087	719	15	,	,	PUNCT
ejpam-6087	719	16	parul	parul	PROPN
ejpam-6087	719	17	parihar	parihar	PROPN
ejpam-6087	719	18	,	,	PUNCT
ejpam-6087	719	19	and	and	CCONJ
ejpam-6087	719	20	jaspreet	jaspreet	PROPN
ejpam-6087	719	21	singh	singh	PROPN
ejpam-6087	719	22	.	.	PUNCT
ejpam-6087	720	1	detection	detection	NOUN
ejpam-6087	720	2	of	of	ADP
ejpam-6087	720	3	fake	fake	ADJ
ejpam-6087	720	4	news	news	NOUN
ejpam-6087	720	5	:	:	PUNCT
ejpam-6087	720	6	a	a	DET
ejpam-6087	720	7	comparative	comparative	ADJ
ejpam-6087	720	8	analysis	analysis	NOUN
ejpam-6087	720	9	using	use	VERB
ejpam-6087	720	10	machine	machine	NOUN
ejpam-6087	720	11	learning	learning	NOUN
ejpam-6087	720	12	.	.	PUNCT
ejpam-6087	721	1	aip	aip	PROPN
ejpam-6087	721	2	conference	conference	NOUN
ejpam-6087	721	3	proceedings	proceeding	NOUN
ejpam-6087	721	4	,	,	PUNCT
ejpam-6087	721	5	3072(1):040014	3072(1):040014	NUM
ejpam-6087	721	6	,	,	PUNCT
ejpam-6087	721	7	03	03	NUM
ejpam-6087	721	8	2024	2024	NUM
ejpam-6087	721	9	.	.	PUNCT
ejpam-6087	722	1	[	[	X
ejpam-6087	722	2	15	15	NUM
ejpam-6087	722	3	]	]	X
ejpam-6087	722	4	joy	joy	NOUN
ejpam-6087	722	5	gorai	gorai	PROPN
ejpam-6087	722	6	and	and	CCONJ
ejpam-6087	722	7	dilip	dilip	PROPN
ejpam-6087	722	8	kumar	kumar	PROPN
ejpam-6087	722	9	shaw	shaw	PROPN
ejpam-6087	722	10	.	.	PUNCT
ejpam-6087	723	1	semantic	semantic	ADJ
ejpam-6087	723	2	difference	difference	NOUN
ejpam-6087	723	3	-	-	PUNCT
ejpam-6087	723	4	based	base	VERB
ejpam-6087	723	5	feature	feature	NOUN
ejpam-6087	723	6	extraction	extraction	NOUN
ejpam-6087	723	7	technique	technique	NOUN
ejpam-6087	723	8	for	for	ADP
ejpam-6087	723	9	fake	fake	ADJ
ejpam-6087	723	10	news	news	NOUN
ejpam-6087	723	11	detection	detection	NOUN
ejpam-6087	723	12	.	.	PUNCT
ejpam-6087	724	1	the	the	DET
ejpam-6087	724	2	journal	journal	NOUN
ejpam-6087	724	3	of	of	ADP
ejpam-6087	724	4	supercomputing	supercomputing	PROPN
ejpam-6087	724	5	,	,	PUNCT
ejpam-6087	724	6	jun	jun	PROPN
ejpam-6087	724	7	2024	2024	NUM
ejpam-6087	724	8	.	.	PUNCT
ejpam-6087	725	1	[	[	X
ejpam-6087	725	2	16	16	NUM
ejpam-6087	725	3	]	]	PUNCT
ejpam-6087	725	4	muhammed	muhamme	VERB
ejpam-6087	725	5	baki	baki	PROPN
ejpam-6087	725	6	çakı	çakı	PROPN
ejpam-6087	725	7	and	and	CCONJ
ejpam-6087	725	8	muhammet	muhammet	VERB
ejpam-6087	725	9	sinan	sinan	PROPN
ejpam-6087	725	10	başarslan	başarslan	PROPN
ejpam-6087	725	11	.	.	PUNCT
ejpam-6087	726	1	classification	classification	NOUN
ejpam-6087	726	2	of	of	ADP
ejpam-6087	726	3	fake	fake	ADJ
ejpam-6087	726	4	news	news	NOUN
ejpam-6087	726	5	using	use	VERB
ejpam-6087	726	6	machine	machine	NOUN
ejpam-6087	726	7	learning	learning	NOUN
ejpam-6087	726	8	and	and	CCONJ
ejpam-6087	726	9	deep	deep	ADJ
ejpam-6087	726	10	learning	learning	NOUN
ejpam-6087	726	11	.	.	PUNCT
ejpam-6087	727	1	journal	journal	PROPN
ejpam-6087	727	2	of	of	ADP
ejpam-6087	727	3	artificial	artificial	ADJ
ejpam-6087	727	4	intelligence	intelligence	NOUN
ejpam-6087	727	5	and	and	CCONJ
ejpam-6087	727	6	data	datum	NOUN
ejpam-6087	727	7	science	science	NOUN
ejpam-6087	727	8	,	,	PUNCT
ejpam-6087	727	9	4(1):22–32	4(1):22–32	NUM
ejpam-6087	727	10	,	,	PUNCT
ejpam-6087	727	11	2024	2024	NUM
ejpam-6087	727	12	.	.	PUNCT
ejpam-6087	728	1	[	[	X
ejpam-6087	728	2	17	17	NUM
ejpam-6087	728	3	]	]	X
ejpam-6087	728	4	manvi	manvi	NOUN
ejpam-6087	728	5	bohra	bohra	PROPN
ejpam-6087	728	6	,	,	PUNCT
ejpam-6087	728	7	indrajeet	indrajeet	PROPN
ejpam-6087	728	8	kumar	kumar	PROPN
ejpam-6087	728	9	,	,	PUNCT
ejpam-6087	728	10	and	and	CCONJ
ejpam-6087	728	11	kamred	kamre	VERB
ejpam-6087	728	12	udham	udham	PROPN
ejpam-6087	728	13	singh	singh	PROPN
ejpam-6087	728	14	.	.	PUNCT
ejpam-6087	729	1	fake	fake	ADJ
ejpam-6087	729	2	news	news	NOUN
ejpam-6087	729	3	detection	detection	NOUN
ejpam-6087	729	4	using	use	VERB
ejpam-6087	729	5	different	different	ADJ
ejpam-6087	729	6	machine	machine	NOUN
ejpam-6087	729	7	learning	learn	VERB
ejpam-6087	729	8	algorithms	algorithm	NOUN
ejpam-6087	729	9	.	.	PUNCT
ejpam-6087	730	1	in	in	ADP
ejpam-6087	730	2	2024	2024	NUM
ejpam-6087	730	3	2nd	2nd	ADJ
ejpam-6087	730	4	international	international	ADJ
ejpam-6087	730	5	conference	conference	NOUN
ejpam-6087	730	6	on	on	ADP
ejpam-6087	730	7	device	device	NOUN
ejpam-6087	730	8	intelligence	intelligence	NOUN
ejpam-6087	730	9	,	,	PUNCT
ejpam-6087	730	10	computing	computing	NOUN
ejpam-6087	730	11	and	and	CCONJ
ejpam-6087	730	12	communication	communication	NOUN
ejpam-6087	730	13	technologies	technology	NOUN
ejpam-6087	730	14	(	(	PUNCT
ejpam-6087	730	15	dicct	dicct	NOUN
ejpam-6087	730	16	)	)	PUNCT
ejpam-6087	730	17	,	,	PUNCT
ejpam-6087	730	18	pages	page	NOUN
ejpam-6087	730	19	45–50	45–50	NUM
ejpam-6087	730	20	,	,	PUNCT
ejpam-6087	730	21	2024	2024	NUM
ejpam-6087	730	22	.	.	PUNCT
ejpam-6087	731	1	[	[	X
ejpam-6087	731	2	18	18	NUM
ejpam-6087	731	3	]	]	X
ejpam-6087	731	4	pooja	pooja	PROPN
ejpam-6087	731	5	malhotra	malhotra	PROPN
ejpam-6087	731	6	and	and	CCONJ
ejpam-6087	731	7	s.	s.	PROPN
ejpam-6087	731	8	k.	k.	PROPN
ejpam-6087	731	9	malik	malik	PROPN
ejpam-6087	731	10	.	.	PUNCT
ejpam-6087	732	1	fake	fake	ADJ
ejpam-6087	732	2	news	news	NOUN
ejpam-6087	732	3	detection	detection	NOUN
ejpam-6087	732	4	using	use	VERB
ejpam-6087	732	5	ensemble	ensemble	ADJ
ejpam-6087	732	6	techniques	technique	NOUN
ejpam-6087	732	7	.	.	PUNCT
ejpam-6087	733	1	multimedia	multimedia	NOUN
ejpam-6087	733	2	tools	tool	NOUN
ejpam-6087	733	3	and	and	CCONJ
ejpam-6087	733	4	applications	application	NOUN
ejpam-6087	733	5	,	,	PUNCT
ejpam-6087	733	6	83(14):42037–42062	83(14):42037–42062	NUM
ejpam-6087	733	7	,	,	PUNCT
ejpam-6087	733	8	apr	apr	NOUN
ejpam-6087	733	9	2024	2024	NUM
ejpam-6087	733	10	.	.	PUNCT
ejpam-6087	734	1	[	[	X
ejpam-6087	734	2	19	19	NUM
ejpam-6087	734	3	]	]	X
ejpam-6087	734	4	kirandeep	kirandeep	NOUN
ejpam-6087	734	5	kaur	kaur	PROPN
ejpam-6087	734	6	and	and	CCONJ
ejpam-6087	734	7	nirbhay	nirbhay	PROPN
ejpam-6087	734	8	kashyap	kashyap	PROPN
ejpam-6087	734	9	.	.	PUNCT
ejpam-6087	735	1	fake	fake	ADJ
ejpam-6087	735	2	news	news	NOUN
ejpam-6087	735	3	detection	detection	NOUN
ejpam-6087	735	4	using	use	VERB
ejpam-6087	735	5	boosting	boost	VERB
ejpam-6087	735	6	ensemble	ensemble	ADJ
ejpam-6087	735	7	method	method	NOUN
ejpam-6087	735	8	.	.	PUNCT
ejpam-6087	736	1	in	in	ADP
ejpam-6087	736	2	2024	2024	NUM
ejpam-6087	736	3	14th	14th	ADJ
ejpam-6087	736	4	international	international	ADJ
ejpam-6087	736	5	conference	conference	NOUN
ejpam-6087	736	6	on	on	ADP
ejpam-6087	736	7	cloud	cloud	NOUN
ejpam-6087	736	8	computing	computing	NOUN
ejpam-6087	736	9	,	,	PUNCT
ejpam-6087	736	10	data	data	NOUN
ejpam-6087	736	11	science	science	NOUN
ejpam-6087	736	12	engineering	engineering	NOUN
ejpam-6087	736	13	(	(	PUNCT
ejpam-6087	736	14	confluence	confluence	NOUN
ejpam-6087	736	15	)	)	PUNCT
ejpam-6087	736	16	,	,	PUNCT
ejpam-6087	736	17	pages	page	NOUN
ejpam-6087	736	18	532–537	532–537	NUM
ejpam-6087	736	19	,	,	PUNCT
ejpam-6087	736	20	2024	2024	NUM
ejpam-6087	736	21	.	.	PUNCT
ejpam-6087	737	1	[	[	X
ejpam-6087	737	2	20	20	NUM
ejpam-6087	737	3	]	]	PUNCT
ejpam-6087	737	4	mohammed	mohammed	PROPN
ejpam-6087	737	5	a.	a.	PROPN
ejpam-6087	737	6	taha	taha	PROPN
ejpam-6087	737	7	,	,	PUNCT
ejpam-6087	737	8	haider	haider	PROPN
ejpam-6087	737	9	d.	d.	PROPN
ejpam-6087	737	10	a.	a.	NOUN
ejpam-6087	737	11	jabar	jabar	PROPN
ejpam-6087	737	12	,	,	PUNCT
ejpam-6087	737	13	and	and	CCONJ
ejpam-6087	737	14	widad	widad	PROPN
ejpam-6087	737	15	k.	k.	PROPN
ejpam-6087	737	16	mohammed	mohammed	PROPN
ejpam-6087	737	17	.	.	PUNCT
ejpam-6087	738	1	fake	fake	ADJ
ejpam-6087	738	2	news	news	NOUN
ejpam-6087	738	3	detection	detection	NOUN
ejpam-6087	738	4	model	model	NOUN
ejpam-6087	738	5	basing	base	VERB
ejpam-6087	738	6	on	on	ADP
ejpam-6087	738	7	machine	machine	NOUN
ejpam-6087	738	8	learning	learning	NOUN
ejpam-6087	738	9	algorithms	algorithm	NOUN
ejpam-6087	738	10	.	.	PUNCT
ejpam-6087	739	1	baghdad	baghdad	PROPN
ejpam-6087	739	2	science	science	PROPN
ejpam-6087	739	3	journal	journal	PROPN
ejpam-6087	739	4	,	,	PUNCT
ejpam-6087	739	5	21(1):35–45	21(1):35–45	NUM
ejpam-6087	739	6	,	,	PUNCT
ejpam-6087	739	7	january	january	PROPN
ejpam-6087	739	8	2024	2024	NUM
ejpam-6087	739	9	.	.	PUNCT
ejpam-6087	740	1	published	publish	VERB
ejpam-6087	740	2	online	online	ADV
ejpam-6087	740	3	first	first	ADV
ejpam-6087	740	4	:	:	PUNCT
ejpam-6087	740	5	january	january	PROPN
ejpam-6087	740	6	20	20	NUM
ejpam-6087	740	7	,	,	PUNCT
ejpam-6087	740	8	2024	2024	NUM
ejpam-6087	740	9	.	.	PUNCT
ejpam-6087	741	1	received	receive	VERB
ejpam-6087	741	2	07/03/2023	07/03/2023	NUM
ejpam-6087	741	3	,	,	PUNCT
ejpam-6087	741	4	revised	revise	VERB
ejpam-6087	741	5	14/07/2023	14/07/2023	NUM
ejpam-6087	741	6	,	,	PUNCT
ejpam-6087	741	7	accepted	accept	VERB
ejpam-6087	741	8	16/07/2023	16/07/2023	NUM
ejpam-6087	741	9	.	.	PUNCT
ejpam-6087	742	1	[	[	X
ejpam-6087	742	2	21	21	NUM
ejpam-6087	742	3	]	]	X
ejpam-6087	742	4	gregorius	gregorius	ADJ
ejpam-6087	742	5	airlangga	airlangga	NOUN
ejpam-6087	742	6	.	.	PUNCT
ejpam-6087	743	1	comparative	comparative	ADJ
ejpam-6087	743	2	analysis	analysis	NOUN
ejpam-6087	743	3	of	of	ADP
ejpam-6087	743	4	machine	machine	NOUN
ejpam-6087	743	5	learning	learn	VERB
ejpam-6087	743	6	algorithms	algorithm	NOUN
ejpam-6087	743	7	for	for	ADP
ejpam-6087	743	8	detecting	detect	VERB
ejpam-6087	743	9	fake	fake	ADJ
ejpam-6087	743	10	news	news	NOUN
ejpam-6087	743	11	:	:	PUNCT
ejpam-6087	743	12	efficacy	efficacy	NOUN
ejpam-6087	743	13	and	and	CCONJ
ejpam-6087	743	14	accuracy	accuracy	NOUN
ejpam-6087	743	15	in	in	ADP
ejpam-6087	743	16	the	the	DET
ejpam-6087	743	17	modern	modern	ADJ
ejpam-6087	743	18	information	information	NOUN
ejpam-6087	743	19	ecosystem	ecosystem	NOUN
ejpam-6087	743	20	.	.	PUNCT
ejpam-6087	744	1	journal	journal	PROPN
ejpam-6087	744	2	of	of	ADP
ejpam-6087	744	3	computer	computer	NOUN
ejpam-6087	744	4	networks	network	NOUN
ejpam-6087	744	5	,	,	PUNCT
ejpam-6087	744	6	architecture	architecture	NOUN
ejpam-6087	744	7	and	and	CCONJ
ejpam-6087	744	8	high	high	ADJ
ejpam-6087	744	9	performance	performance	NOUN
ejpam-6087	744	10	computing	computing	NOUN
ejpam-6087	744	11	,	,	PUNCT
ejpam-6087	744	12	6(1):354–363	6(1):354–363	PROPN
ejpam-6087	744	13	,	,	PUNCT
ejpam-6087	744	14	jan	jan	PROPN
ejpam-6087	744	15	.	.	PROPN
ejpam-6087	744	16	2024	2024	NUM
ejpam-6087	744	17	.	.	PUNCT
ejpam-6087	745	1	[	[	X
ejpam-6087	745	2	22	22	NUM
ejpam-6087	745	3	]	]	X
ejpam-6087	745	4	sheik	sheik	PROPN
ejpam-6087	745	5	abdullah	abdullah	PROPN
ejpam-6087	745	6	a	a	PROPN
ejpam-6087	745	7	,	,	PUNCT
ejpam-6087	745	8	parkavi	parkavi	VERB
ejpam-6087	745	9	r	r	NOUN
ejpam-6087	745	10	,	,	PUNCT
ejpam-6087	745	11	p.je	p.je	NOUN
ejpam-6087	745	12	sai	sai	PROPN
ejpam-6087	745	13	kailash	kailash	PROPN
ejpam-6087	745	14	,	,	PUNCT
ejpam-6087	745	15	and	and	CCONJ
ejpam-6087	745	16	deepthi	deepthi	PROPN
ejpam-6087	745	17	ramesh	ramesh	PROPN
ejpam-6087	745	18	.	.	PUNCT
ejpam-6087	746	1	analysis	analysis	NOUN
ejpam-6087	746	2	of	of	ADP
ejpam-6087	746	3	onm	onm	NOUN
ejpam-6087	746	4	.	.	PUNCT
ejpam-6087	747	1	al	al	PROPN
ejpam-6087	747	2	-	-	PUNCT
ejpam-6087	747	3	tarawneh	tarawneh	PROPN
ejpam-6087	747	4	et	et	PROPN
ejpam-6087	747	5	al	al	PROPN
ejpam-6087	747	6	.	.	PUNCT
ejpam-6087	747	7	/	/	SYM
ejpam-6087	747	8	eur	eur	PROPN
ejpam-6087	747	9	.	.	PUNCT
ejpam-6087	748	1	j.	j.	PROPN
ejpam-6087	748	2	pure	pure	PROPN
ejpam-6087	748	3	appl	appl	PROPN
ejpam-6087	748	4	.	.	PROPN
ejpam-6087	748	5	math	math	PROPN
ejpam-6087	748	6	,	,	PUNCT
ejpam-6087	748	7	18	18	NUM
ejpam-6087	748	8	(	(	PUNCT
ejpam-6087	748	9	2	2	NUM
ejpam-6087	748	10	)	)	PUNCT
ejpam-6087	748	11	(	(	PUNCT
ejpam-6087	748	12	2025	2025	NUM
ejpam-6087	748	13	)	)	PUNCT
ejpam-6087	748	14	,	,	PUNCT
ejpam-6087	748	15	6087	6087	NUM
ejpam-6087	748	16	38	38	NUM
ejpam-6087	748	17	of	of	ADP
ejpam-6087	748	18	39	39	NUM
ejpam-6087	748	19	line	line	NOUN
ejpam-6087	748	20	fake	fake	ADJ
ejpam-6087	748	21	news	news	NOUN
ejpam-6087	748	22	using	use	VERB
ejpam-6087	748	23	machine	machine	NOUN
ejpam-6087	748	24	learning	learning	NOUN
ejpam-6087	748	25	techniques	technique	NOUN
ejpam-6087	748	26	.	.	PUNCT
ejpam-6087	749	1	in	in	ADP
ejpam-6087	749	2	2024	2024	NUM
ejpam-6087	749	3	second	second	ADJ
ejpam-6087	749	4	international	international	ADJ
ejpam-6087	749	5	conference	conference	NOUN
ejpam-6087	749	6	on	on	ADP
ejpam-6087	749	7	emerging	emerge	VERB
ejpam-6087	749	8	trends	trend	NOUN
ejpam-6087	749	9	in	in	ADP
ejpam-6087	749	10	information	information	NOUN
ejpam-6087	749	11	technology	technology	NOUN
ejpam-6087	749	12	and	and	CCONJ
ejpam-6087	749	13	engineering	engineering	NOUN
ejpam-6087	749	14	(	(	PUNCT
ejpam-6087	749	15	icetite	icetite	NOUN
ejpam-6087	749	16	)	)	PUNCT
ejpam-6087	749	17	,	,	PUNCT
ejpam-6087	749	18	pages	page	NOUN
ejpam-6087	749	19	1–5	1–5	NUM
ejpam-6087	749	20	,	,	PUNCT
ejpam-6087	749	21	2024	2024	NUM
ejpam-6087	749	22	.	.	PUNCT
ejpam-6087	750	1	[	[	X
ejpam-6087	750	2	23	23	NUM
ejpam-6087	750	3	]	]	X
ejpam-6087	750	4	azka	azka	PROPN
ejpam-6087	750	5	kishwar	kishwar	PROPN
ejpam-6087	750	6	and	and	CCONJ
ejpam-6087	750	7	adeel	adeel	PROPN
ejpam-6087	750	8	zafar	zafar	PROPN
ejpam-6087	750	9	.	.	PUNCT
ejpam-6087	751	1	fake	fake	ADJ
ejpam-6087	751	2	news	news	NOUN
ejpam-6087	751	3	detection	detection	NOUN
ejpam-6087	751	4	on	on	ADP
ejpam-6087	751	5	pakistani	pakistani	ADJ
ejpam-6087	751	6	news	news	PROPN
ejpam-6087	751	7	using	use	VERB
ejpam-6087	751	8	machine	machine	NOUN
ejpam-6087	751	9	learning	learning	NOUN
ejpam-6087	751	10	and	and	CCONJ
ejpam-6087	751	11	deep	deep	ADJ
ejpam-6087	751	12	learning	learning	NOUN
ejpam-6087	751	13	.	.	PUNCT
ejpam-6087	752	1	expert	expert	NOUN
ejpam-6087	752	2	systems	system	NOUN
ejpam-6087	752	3	with	with	ADP
ejpam-6087	752	4	applications	application	NOUN
ejpam-6087	752	5	,	,	PUNCT
ejpam-6087	752	6	211:118558	211:118558	NUM
ejpam-6087	752	7	,	,	PUNCT
ejpam-6087	752	8	2023	2023	NUM
ejpam-6087	752	9	.	.	PUNCT
ejpam-6087	753	1	[	[	X
ejpam-6087	753	2	24	24	NUM
ejpam-6087	753	3	]	]	X
ejpam-6087	753	4	minjung	minjung	PROPN
ejpam-6087	753	5	park	park	NOUN
ejpam-6087	753	6	and	and	CCONJ
ejpam-6087	753	7	sangmi	sangmi	VERB
ejpam-6087	753	8	chai	chai	NOUN
ejpam-6087	753	9	.	.	PUNCT
ejpam-6087	754	1	constructing	construct	VERB
ejpam-6087	754	2	a	a	DET
ejpam-6087	754	3	user	user	NOUN
ejpam-6087	754	4	-	-	PUNCT
ejpam-6087	754	5	centered	center	VERB
ejpam-6087	754	6	fake	fake	ADJ
ejpam-6087	754	7	news	news	NOUN
ejpam-6087	754	8	detection	detection	NOUN
ejpam-6087	754	9	model	model	NOUN
ejpam-6087	754	10	by	by	ADP
ejpam-6087	754	11	using	use	VERB
ejpam-6087	754	12	classification	classification	NOUN
ejpam-6087	754	13	algorithms	algorithm	NOUN
ejpam-6087	754	14	in	in	ADP
ejpam-6087	754	15	machine	machine	NOUN
ejpam-6087	754	16	learning	learn	VERB
ejpam-6087	754	17	techniques	technique	NOUN
ejpam-6087	754	18	.	.	PUNCT
ejpam-6087	755	1	ieee	ieee	NOUN
ejpam-6087	755	2	access	access	NOUN
ejpam-6087	755	3	,	,	PUNCT
ejpam-6087	755	4	11:71517–71527	11:71517–71527	NUM
ejpam-6087	755	5	,	,	PUNCT
ejpam-6087	755	6	2023	2023	NUM
ejpam-6087	755	7	.	.	PUNCT
ejpam-6087	756	1	[	[	X
ejpam-6087	756	2	25	25	NUM
ejpam-6087	756	3	]	]	X
ejpam-6087	756	4	mutaz	mutaz	PROPN
ejpam-6087	756	5	a.	a.	PROPN
ejpam-6087	756	6	b.	b.	PROPN
ejpam-6087	756	7	al	al	PROPN
ejpam-6087	756	8	-	-	PUNCT
ejpam-6087	756	9	tarawneh	tarawneh	PROPN
ejpam-6087	756	10	,	,	PUNCT
ejpam-6087	756	11	omar	omar	PROPN
ejpam-6087	756	12	al	al	PROPN
ejpam-6087	756	13	-	-	PUNCT
ejpam-6087	756	14	irr	irr	PROPN
ejpam-6087	756	15	,	,	PUNCT
ejpam-6087	756	16	khaled	khaled	PROPN
ejpam-6087	756	17	s.	s.	PROPN
ejpam-6087	756	18	al	al	PROPN
ejpam-6087	756	19	-	-	PUNCT
ejpam-6087	756	20	maaitah	maaitah	PROPN
ejpam-6087	756	21	,	,	PUNCT
ejpam-6087	756	22	hassan	hassan	PROPN
ejpam-6087	756	23	kanj	kanj	PROPN
ejpam-6087	756	24	,	,	PUNCT
ejpam-6087	756	25	and	and	CCONJ
ejpam-6087	756	26	wael	wael	PROPN
ejpam-6087	756	27	hosny	hosny	PROPN
ejpam-6087	756	28	fouad	fouad	PROPN
ejpam-6087	756	29	aly	aly	PROPN
ejpam-6087	756	30	.	.	PUNCT
ejpam-6087	757	1	enhancing	enhance	VERB
ejpam-6087	757	2	fake	fake	ADJ
ejpam-6087	757	3	news	news	NOUN
ejpam-6087	757	4	detection	detection	NOUN
ejpam-6087	757	5	with	with	ADP
ejpam-6087	757	6	word	word	NOUN
ejpam-6087	757	7	embedding	embed	VERB
ejpam-6087	757	8	:	:	PUNCT
ejpam-6087	757	9	a	a	DET
ejpam-6087	757	10	machine	machine	NOUN
ejpam-6087	757	11	learning	learning	NOUN
ejpam-6087	757	12	and	and	CCONJ
ejpam-6087	757	13	deep	deep	ADJ
ejpam-6087	757	14	learning	learning	NOUN
ejpam-6087	757	15	approach	approach	NOUN
ejpam-6087	757	16	.	.	PUNCT
ejpam-6087	758	1	computers	computer	NOUN
ejpam-6087	758	2	,	,	PUNCT
ejpam-6087	758	3	13(9	13(9	NUM
ejpam-6087	758	4	)	)	PUNCT
ejpam-6087	758	5	,	,	PUNCT
ejpam-6087	758	6	2024	2024	NUM
ejpam-6087	758	7	.	.	PUNCT
ejpam-6087	759	1	[	[	X
ejpam-6087	759	2	26	26	NUM
ejpam-6087	759	3	]	]	X
ejpam-6087	759	4	mutaz	mutaz	PROPN
ejpam-6087	759	5	a.	a.	PROPN
ejpam-6087	759	6	b.	b.	PROPN
ejpam-6087	759	7	al	al	PROPN
ejpam-6087	759	8	-	-	PUNCT
ejpam-6087	759	9	tarawneh	tarawneh	PROPN
ejpam-6087	759	10	,	,	PUNCT
ejpam-6087	759	11	omar	omar	PROPN
ejpam-6087	759	12	alirr	alirr	PROPN
ejpam-6087	759	13	,	,	PUNCT
ejpam-6087	759	14	and	and	CCONJ
ejpam-6087	759	15	hassan	hassan	PROPN
ejpam-6087	759	16	kanj	kanj	PROPN
ejpam-6087	759	17	.	.	PUNCT
ejpam-6087	759	18	performance	performance	NOUN
ejpam-6087	759	19	evaluation	evaluation	NOUN
ejpam-6087	759	20	of	of	ADP
ejpam-6087	759	21	machine	machine	NOUN
ejpam-6087	759	22	learning	learning	NOUN
ejpam-6087	759	23	-	-	PUNCT
ejpam-6087	759	24	based	base	VERB
ejpam-6087	759	25	cyber	cyber	NOUN
ejpam-6087	759	26	attack	attack	NOUN
ejpam-6087	759	27	detection	detection	NOUN
ejpam-6087	759	28	in	in	ADP
ejpam-6087	759	29	electric	electric	ADJ
ejpam-6087	759	30	vehicles	vehicle	NOUN
ejpam-6087	759	31	charging	charge	VERB
ejpam-6087	759	32	stations	station	NOUN
ejpam-6087	759	33	.	.	PUNCT
ejpam-6087	760	1	international	international	ADJ
ejpam-6087	760	2	journal	journal	NOUN
ejpam-6087	760	3	of	of	ADP
ejpam-6087	760	4	advanced	advanced	ADJ
ejpam-6087	760	5	computer	computer	NOUN
ejpam-6087	760	6	science	science	NOUN
ejpam-6087	760	7	and	and	CCONJ
ejpam-6087	760	8	applications	application	NOUN
ejpam-6087	760	9	,	,	PUNCT
ejpam-6087	760	10	16(3	16(3	NOUN
ejpam-6087	760	11	)	)	PUNCT
ejpam-6087	760	12	,	,	PUNCT
ejpam-6087	760	13	2025	2025	NUM
ejpam-6087	760	14	.	.	PUNCT
ejpam-6087	761	1	[	[	X
ejpam-6087	761	2	27	27	NUM
ejpam-6087	761	3	]	]	X
ejpam-6087	761	4	sudhansu	sudhansu	PROPN
ejpam-6087	761	5	r	r	NOUN
ejpam-6087	761	6	lenka	lenka	NOUN
ejpam-6087	761	7	,	,	PUNCT
ejpam-6087	761	8	sukant	sukant	ADJ
ejpam-6087	761	9	kishoro	kishoro	PROPN
ejpam-6087	761	10	bisoy	bisoy	PROPN
ejpam-6087	761	11	,	,	PUNCT
ejpam-6087	761	12	rojalina	rojalina	PROPN
ejpam-6087	761	13	priyadarshini	priyadarshini	PROPN
ejpam-6087	761	14	,	,	PUNCT
ejpam-6087	761	15	and	and	CCONJ
ejpam-6087	761	16	mangal	mangal	ADJ
ejpam-6087	761	17	sain	sain	NOUN
ejpam-6087	761	18	.	.	PUNCT
ejpam-6087	762	1	empirical	empirical	ADJ
ejpam-6087	762	2	analysis	analysis	NOUN
ejpam-6087	762	3	of	of	ADP
ejpam-6087	762	4	ensemble	ensemble	ADJ
ejpam-6087	762	5	learning	learning	NOUN
ejpam-6087	762	6	for	for	ADP
ejpam-6087	762	7	imbalanced	imbalanced	ADJ
ejpam-6087	762	8	credit	credit	NOUN
ejpam-6087	762	9	scoring	scoring	NOUN
ejpam-6087	762	10	datasets	dataset	NOUN
ejpam-6087	762	11	:	:	PUNCT
ejpam-6087	762	12	a	a	DET
ejpam-6087	762	13	systematic	systematic	ADJ
ejpam-6087	762	14	review	review	NOUN
ejpam-6087	762	15	.	.	PUNCT
ejpam-6087	763	1	wireless	wireless	ADJ
ejpam-6087	763	2	communications	communication	NOUN
ejpam-6087	763	3	and	and	CCONJ
ejpam-6087	763	4	mobile	mobile	NOUN
ejpam-6087	763	5	computing	computing	NOUN
ejpam-6087	763	6	,	,	PUNCT
ejpam-6087	763	7	2022(1):6584352	2022(1):6584352	NUM
ejpam-6087	763	8	,	,	PUNCT
ejpam-6087	763	9	2022	2022	NUM
ejpam-6087	763	10	.	.	PUNCT
ejpam-6087	764	1	[	[	X
ejpam-6087	764	2	28	28	NUM
ejpam-6087	764	3	]	]	X
ejpam-6087	764	4	adel	adel	PROPN
ejpam-6087	764	5	mellit	mellit	VERB
ejpam-6087	764	6	and	and	CCONJ
ejpam-6087	764	7	soteris	soteris	PROPN
ejpam-6087	764	8	kalogirou	kalogirou	PROPN
ejpam-6087	764	9	.	.	PUNCT
ejpam-6087	765	1	assessment	assessment	NOUN
ejpam-6087	765	2	of	of	ADP
ejpam-6087	765	3	machine	machine	NOUN
ejpam-6087	765	4	learning	learning	NOUN
ejpam-6087	765	5	and	and	CCONJ
ejpam-6087	765	6	ensemble	ensemble	ADJ
ejpam-6087	765	7	methods	method	NOUN
ejpam-6087	765	8	for	for	ADP
ejpam-6087	765	9	fault	fault	NOUN
ejpam-6087	765	10	diagnosis	diagnosis	NOUN
ejpam-6087	765	11	of	of	ADP
ejpam-6087	765	12	photovoltaic	photovoltaic	NOUN
ejpam-6087	765	13	systems	system	NOUN
ejpam-6087	765	14	.	.	PUNCT
ejpam-6087	766	1	renewable	renewable	ADJ
ejpam-6087	766	2	energy	energy	NOUN
ejpam-6087	766	3	,	,	PUNCT
ejpam-6087	766	4	184:1074	184:1074	NUM
ejpam-6087	766	5	–	–	PUNCT
ejpam-6087	766	6	1090	1090	NUM
ejpam-6087	766	7	,	,	PUNCT
ejpam-6087	766	8	2022	2022	NUM
ejpam-6087	766	9	.	.	PUNCT
ejpam-6087	767	1	[	[	X
ejpam-6087	767	2	29	29	NUM
ejpam-6087	767	3	]	]	X
ejpam-6087	767	4	mutaz	mutaz	PROPN
ejpam-6087	767	5	a.	a.	PROPN
ejpam-6087	767	6	b.	b.	PROPN
ejpam-6087	767	7	al	al	PROPN
ejpam-6087	767	8	-	-	PUNCT
ejpam-6087	767	9	tarawneh	tarawneh	PROPN
ejpam-6087	767	10	,	,	PUNCT
ejpam-6087	767	11	khaled	khaled	PROPN
ejpam-6087	767	12	s.	s.	PROPN
ejpam-6087	767	13	al	al	PROPN
ejpam-6087	767	14	-	-	PUNCT
ejpam-6087	767	15	maaitah	maaitah	PROPN
ejpam-6087	767	16	,	,	PUNCT
ejpam-6087	767	17	and	and	CCONJ
ejpam-6087	767	18	ashraf	ashraf	PROPN
ejpam-6087	767	19	alkhresheh	alkhresheh	NOUN
ejpam-6087	767	20	.	.	PUNCT
ejpam-6087	768	1	adaptive	adaptive	ADJ
ejpam-6087	768	2	ensemble	ensemble	ADJ
ejpam-6087	768	3	selection	selection	NOUN
ejpam-6087	768	4	for	for	ADP
ejpam-6087	768	5	personalized	personalize	VERB
ejpam-6087	768	6	cardiovascular	cardiovascular	ADJ
ejpam-6087	768	7	disease	disease	NOUN
ejpam-6087	768	8	prediction	prediction	NOUN
ejpam-6087	768	9	using	use	VERB
ejpam-6087	768	10	clustering	clustering	NOUN
ejpam-6087	768	11	and	and	CCONJ
ejpam-6087	768	12	feature	feature	NOUN
ejpam-6087	768	13	selection	selection	NOUN
ejpam-6087	768	14	.	.	PUNCT
ejpam-6087	769	1	international	international	ADJ
ejpam-6087	769	2	journal	journal	NOUN
ejpam-6087	769	3	of	of	ADP
ejpam-6087	769	4	advanced	advanced	ADJ
ejpam-6087	769	5	computer	computer	NOUN
ejpam-6087	769	6	science	science	NOUN
ejpam-6087	769	7	and	and	CCONJ
ejpam-6087	769	8	applications	application	NOUN
ejpam-6087	769	9	,	,	PUNCT
ejpam-6087	769	10	16(3	16(3	NOUN
ejpam-6087	769	11	)	)	PUNCT
ejpam-6087	769	12	,	,	PUNCT
ejpam-6087	769	13	2025	2025	NUM
ejpam-6087	769	14	.	.	PUNCT
ejpam-6087	770	1	[	[	X
ejpam-6087	770	2	30	30	NUM
ejpam-6087	770	3	]	]	X
ejpam-6087	770	4	richard	richard	PROPN
ejpam-6087	770	5	a.	a.	PROPN
ejpam-6087	770	6	berk	berk	PROPN
ejpam-6087	770	7	.	.	PUNCT
ejpam-6087	771	1	classification	classification	NOUN
ejpam-6087	771	2	and	and	CCONJ
ejpam-6087	771	3	regression	regression	NOUN
ejpam-6087	771	4	trees	tree	NOUN
ejpam-6087	771	5	(	(	PUNCT
ejpam-6087	771	6	cart	cart	NOUN
ejpam-6087	771	7	)	)	PUNCT
ejpam-6087	771	8	,	,	PUNCT
ejpam-6087	771	9	pages	page	NOUN
ejpam-6087	771	10	1–65	1–65	PROPN
ejpam-6087	771	11	.	.	PUNCT
ejpam-6087	772	1	springer	springer	PROPN
ejpam-6087	772	2	new	new	PROPN
ejpam-6087	772	3	york	york	PROPN
ejpam-6087	772	4	,	,	PUNCT
ejpam-6087	772	5	new	new	PROPN
ejpam-6087	772	6	york	york	PROPN
ejpam-6087	772	7	,	,	PUNCT
ejpam-6087	772	8	ny	ny	PROPN
ejpam-6087	772	9	,	,	PUNCT
ejpam-6087	772	10	2008	2008	NUM
ejpam-6087	772	11	.	.	PUNCT
ejpam-6087	773	1	[	[	X
ejpam-6087	773	2	31	31	NUM
ejpam-6087	773	3	]	]	PUNCT
ejpam-6087	773	4	ludmila	ludmila	PROPN
ejpam-6087	773	5	i	i	PRON
ejpam-6087	773	6	kuncheva	kuncheva	PROPN
ejpam-6087	773	7	.	.	PUNCT
ejpam-6087	774	1	combining	combine	VERB
ejpam-6087	774	2	pattern	pattern	NOUN
ejpam-6087	774	3	classifiers	classifier	NOUN
ejpam-6087	774	4	:	:	PUNCT
ejpam-6087	774	5	methods	method	NOUN
ejpam-6087	774	6	and	and	CCONJ
ejpam-6087	774	7	algorithms	algorithm	NOUN
ejpam-6087	774	8	.	.	PUNCT
ejpam-6087	775	1	john	john	PROPN
ejpam-6087	775	2	wiley	wiley	PROPN
ejpam-6087	775	3	&	&	CCONJ
ejpam-6087	775	4	sons	son	NOUN
ejpam-6087	775	5	,	,	PUNCT
ejpam-6087	775	6	2014	2014	NUM
ejpam-6087	775	7	.	.	PUNCT
ejpam-6087	776	1	[	[	X
ejpam-6087	776	2	32	32	NUM
ejpam-6087	776	3	]	]	PUNCT
ejpam-6087	776	4	bernhard	bernhard	PROPN
ejpam-6087	776	5	scholkopf	scholkopf	PROPN
ejpam-6087	776	6	.	.	PUNCT
ejpam-6087	777	1	support	support	NOUN
ejpam-6087	777	2	vector	vector	NOUN
ejpam-6087	777	3	machines	machine	NOUN
ejpam-6087	777	4	:	:	PUNCT
ejpam-6087	777	5	a	a	DET
ejpam-6087	777	6	practical	practical	ADJ
ejpam-6087	777	7	consequence	consequence	NOUN
ejpam-6087	777	8	of	of	ADP
ejpam-6087	777	9	learning	learn	VERB
ejpam-6087	777	10	theory	theory	NOUN
ejpam-6087	777	11	.	.	PUNCT
ejpam-6087	778	1	ieee	ieee	NOUN
ejpam-6087	778	2	intelligent	intelligent	ADJ
ejpam-6087	778	3	systems	system	NOUN
ejpam-6087	778	4	,	,	PUNCT
ejpam-6087	778	5	13	13	NUM
ejpam-6087	778	6	,	,	PUNCT
ejpam-6087	778	7	1998	1998	NUM
ejpam-6087	778	8	.	.	PUNCT
ejpam-6087	779	1	[	[	X
ejpam-6087	779	2	33	33	NUM
ejpam-6087	779	3	]	]	PUNCT
ejpam-6087	779	4	thi	thi	AUX
ejpam-6087	779	5	kha	kha	PROPN
ejpam-6087	779	6	nguyen	nguyen	NOUN
ejpam-6087	779	7	and	and	CCONJ
ejpam-6087	779	8	thi	thi	VERB
ejpam-6087	779	9	phuong	phuong	PROPN
ejpam-6087	779	10	trang	trang	PROPN
ejpam-6087	779	11	pham	pham	PROPN
ejpam-6087	779	12	.	.	PUNCT
ejpam-6087	780	1	predicting	predict	VERB
ejpam-6087	780	2	bankruptcy	bankruptcy	NOUN
ejpam-6087	780	3	using	use	VERB
ejpam-6087	780	4	machine	machine	NOUN
ejpam-6087	780	5	learning	learning	NOUN
ejpam-6087	780	6	algorithms	algorithm	NOUN
ejpam-6087	780	7	.	.	PUNCT
ejpam-6087	781	1	tp	tp	NOUN
ejpam-6087	781	2	ch́ı	ch́ı	PROPN
ejpam-6087	781	3	khoa	khoa	PROPN
ejpam-6087	781	4	hc	hc	ADP
ejpam-6087	781	5	và	và	PROPN
ejpam-6087	781	6	công	công	VERB
ejpam-6087	781	7	ngh	ngh	PROPN
ejpam-6087	781	8	-	-	PROPN
ejpam-6087	781	9	i	i	NOUN
ejpam-6087	781	10	hc	hc	PROPN
ejpam-6087	781	11	à	à	PROPN
ejpam-6087	781	12	nng	nng	PROPN
ejpam-6087	781	13	,	,	PUNCT
ejpam-6087	781	14	pages	page	NOUN
ejpam-6087	781	15	6–9	6–9	PROPN
ejpam-6087	781	16	,	,	PUNCT
ejpam-6087	781	17	2018	2018	NUM
ejpam-6087	781	18	.	.	PUNCT
ejpam-6087	782	1	[	[	X
ejpam-6087	782	2	34	34	NUM
ejpam-6087	782	3	]	]	X
ejpam-6087	782	4	thiago	thiago	PROPN
ejpam-6087	782	5	josé	josé	PROPN
ejpam-6087	782	6	lucas	lucas	PROPN
ejpam-6087	782	7	,	,	PUNCT
ejpam-6087	782	8	inaê	inaê	ADJ
ejpam-6087	782	9	soares	soare	NOUN
ejpam-6087	782	10	de	de	X
ejpam-6087	782	11	figueiredo	figueiredo	PROPN
ejpam-6087	782	12	,	,	PUNCT
ejpam-6087	782	13	carlos	carlos	PROPN
ejpam-6087	782	14	alexandre	alexandre	PROPN
ejpam-6087	782	15	carvalho	carvalho	PROPN
ejpam-6087	782	16	tojeiro	tojeiro	PROPN
ejpam-6087	782	17	,	,	PUNCT
ejpam-6087	782	18	alex	alex	PROPN
ejpam-6087	782	19	marino	marino	PROPN
ejpam-6087	782	20	g	g	PROPN
ejpam-6087	782	21	de	de	PROPN
ejpam-6087	782	22	almeida	almeida	PROPN
ejpam-6087	782	23	,	,	PUNCT
ejpam-6087	782	24	rafa	rafa	PROPN
ejpam-6087	782	25	l	l	PROPN
ejpam-6087	782	26	scherer	scherer	PROPN
ejpam-6087	782	27	,	,	PUNCT
ejpam-6087	782	28	josé	josé	ADJ
ejpam-6087	782	29	remo	remo	NOUN
ejpam-6087	782	30	f	f	PROPN
ejpam-6087	782	31	brega	brega	PROPN
ejpam-6087	782	32	,	,	PUNCT
ejpam-6087	782	33	joão	joão	PROPN
ejpam-6087	782	34	paulo	paulo	PROPN
ejpam-6087	782	35	papa	papa	PROPN
ejpam-6087	782	36	,	,	PUNCT
ejpam-6087	782	37	and	and	CCONJ
ejpam-6087	782	38	kelton	kelton	PROPN
ejpam-6087	782	39	augusto	augusto	PROPN
ejpam-6087	782	40	pontara	pontara	PROPN
ejpam-6087	782	41	da	da	PROPN
ejpam-6087	782	42	costa	costa	PROPN
ejpam-6087	782	43	.	.	PUNCT
ejpam-6087	783	1	a	a	DET
ejpam-6087	783	2	comprehensive	comprehensive	ADJ
ejpam-6087	783	3	survey	survey	NOUN
ejpam-6087	783	4	on	on	ADP
ejpam-6087	783	5	ensemble	ensemble	ADJ
ejpam-6087	783	6	learningbased	learningbase	VERB
ejpam-6087	783	7	intrusion	intrusion	NOUN
ejpam-6087	783	8	detection	detection	NOUN
ejpam-6087	783	9	approaches	approach	VERB
ejpam-6087	783	10	in	in	ADP
ejpam-6087	783	11	computer	computer	NOUN
ejpam-6087	783	12	networks	network	NOUN
ejpam-6087	783	13	.	.	PUNCT
ejpam-6087	784	1	ieee	ieee	NOUN
ejpam-6087	784	2	access	access	NOUN
ejpam-6087	784	3	,	,	PUNCT
ejpam-6087	784	4	2023	2023	NUM
ejpam-6087	784	5	.	.	PUNCT
ejpam-6087	785	1	[	[	X
ejpam-6087	785	2	35	35	NUM
ejpam-6087	785	3	]	]	PUNCT
ejpam-6087	785	4	pierre	pierre	NOUN
ejpam-6087	785	5	geurts	geurt	NOUN
ejpam-6087	785	6	,	,	PUNCT
ejpam-6087	785	7	damien	damien	PROPN
ejpam-6087	785	8	ernst	ernst	PROPN
ejpam-6087	785	9	,	,	PUNCT
ejpam-6087	785	10	and	and	CCONJ
ejpam-6087	785	11	louis	louis	PROPN
ejpam-6087	785	12	wehenkel	wehenkel	NOUN
ejpam-6087	785	13	.	.	PUNCT
ejpam-6087	786	1	extremely	extremely	ADV
ejpam-6087	786	2	randomized	randomize	VERB
ejpam-6087	786	3	trees	tree	NOUN
ejpam-6087	786	4	.	.	PUNCT
ejpam-6087	787	1	machine	machine	NOUN
ejpam-6087	787	2	learning	learning	PROPN
ejpam-6087	787	3	,	,	PUNCT
ejpam-6087	787	4	63:3–42	63:3–42	NUM
ejpam-6087	787	5	,	,	PUNCT
ejpam-6087	787	6	2006	2006	NUM
ejpam-6087	787	7	.	.	PUNCT
ejpam-6087	788	1	[	[	X
ejpam-6087	788	2	36	36	NUM
ejpam-6087	788	3	]	]	X
ejpam-6087	788	4	antonio	antonio	PROPN
ejpam-6087	788	5	criminisi	criminisi	PROPN
ejpam-6087	788	6	et	et	PROPN
ejpam-6087	788	7	al	al	PROPN
ejpam-6087	788	8	.	.	PUNCT
ejpam-6087	789	1	regression	regression	NOUN
ejpam-6087	789	2	forests	forest	NOUN
ejpam-6087	789	3	for	for	ADP
ejpam-6087	789	4	efficient	efficient	ADJ
ejpam-6087	789	5	anatomy	anatomy	NOUN
ejpam-6087	789	6	detection	detection	NOUN
ejpam-6087	789	7	and	and	CCONJ
ejpam-6087	789	8	localization	localization	NOUN
ejpam-6087	789	9	in	in	ADP
ejpam-6087	789	10	ct	ct	PROPN
ejpam-6087	789	11	studies	study	NOUN
ejpam-6087	789	12	,	,	PUNCT
ejpam-6087	789	13	sep	sep	PROPN
ejpam-6087	789	14	.	.	PROPN
ejpam-6087	789	15	20	20	NUM
ejpam-6087	789	16	,	,	PUNCT
ejpam-6087	789	17	2010	2010	NUM
ejpam-6087	789	18	,	,	PUNCT
ejpam-6087	789	19	medical	medical	ADJ
ejpam-6087	789	20	computer	computer	NOUN
ejpam-6087	789	21	visions	vision	NOUN
ejpam-6087	789	22	.	.	PUNCT
ejpam-6087	790	1	recognition	recognition	NOUN
ejpam-6087	790	2	techniques	technique	NOUN
ejpam-6087	790	3	and	and	CCONJ
ejpam-6087	790	4	applications	application	NOUN
ejpam-6087	790	5	in	in	ADP
ejpam-6087	790	6	medical	medical	ADJ
ejpam-6087	790	7	imaging	imaging	NOUN
ejpam-6087	790	8	.	.	PUNCT
ejpam-6087	791	1	[	[	X
ejpam-6087	791	2	37	37	NUM
ejpam-6087	791	3	]	]	X
ejpam-6087	791	4	earum	earum	PROPN
ejpam-6087	791	5	mushtaq	mushtaq	PROPN
ejpam-6087	791	6	,	,	PUNCT
ejpam-6087	791	7	aneela	aneela	PROPN
ejpam-6087	791	8	zameer	zameer	PROPN
ejpam-6087	791	9	,	,	PUNCT
ejpam-6087	791	10	and	and	CCONJ
ejpam-6087	791	11	asifullah	asifullah	PROPN
ejpam-6087	791	12	khan	khan	PROPN
ejpam-6087	791	13	.	.	PUNCT
ejpam-6087	792	1	a	a	DET
ejpam-6087	792	2	two	two	NUM
ejpam-6087	792	3	-	-	PUNCT
ejpam-6087	792	4	stage	stage	NOUN
ejpam-6087	792	5	stacked	stack	VERB
ejpam-6087	792	6	ensemble	ensemble	ADJ
ejpam-6087	792	7	m.	m.	NOUN
ejpam-6087	792	8	al	al	PROPN
ejpam-6087	792	9	-	-	PUNCT
ejpam-6087	792	10	tarawneh	tarawneh	PROPN
ejpam-6087	792	11	et	et	PROPN
ejpam-6087	792	12	al	al	PROPN
ejpam-6087	792	13	.	.	PUNCT
ejpam-6087	792	14	/	/	SYM
ejpam-6087	792	15	eur	eur	PROPN
ejpam-6087	792	16	.	.	PUNCT
ejpam-6087	793	1	j.	j.	PROPN
ejpam-6087	793	2	pure	pure	PROPN
ejpam-6087	793	3	appl	appl	PROPN
ejpam-6087	793	4	.	.	PROPN
ejpam-6087	793	5	math	math	PROPN
ejpam-6087	793	6	,	,	PUNCT
ejpam-6087	793	7	18	18	NUM
ejpam-6087	793	8	(	(	PUNCT
ejpam-6087	793	9	2	2	NUM
ejpam-6087	793	10	)	)	PUNCT
ejpam-6087	793	11	(	(	PUNCT
ejpam-6087	793	12	2025	2025	NUM
ejpam-6087	793	13	)	)	PUNCT
ejpam-6087	793	14	,	,	PUNCT
ejpam-6087	793	15	6087	6087	NUM
ejpam-6087	793	16	39	39	NUM
ejpam-6087	793	17	of	of	ADP
ejpam-6087	793	18	39	39	NUM
ejpam-6087	793	19	intrusion	intrusion	NOUN
ejpam-6087	793	20	detection	detection	NOUN
ejpam-6087	793	21	system	system	NOUN
ejpam-6087	793	22	using	use	VERB
ejpam-6087	793	23	five	five	NUM
ejpam-6087	793	24	base	base	NOUN
ejpam-6087	793	25	classifiers	classifier	NOUN
ejpam-6087	793	26	and	and	CCONJ
ejpam-6087	793	27	mlp	mlp	NOUN
ejpam-6087	793	28	with	with	ADP
ejpam-6087	793	29	optimal	optimal	ADJ
ejpam-6087	793	30	feature	feature	NOUN
ejpam-6087	793	31	selection	selection	NOUN
ejpam-6087	793	32	.	.	PUNCT
ejpam-6087	794	1	microprocessors	microprocessor	NOUN
ejpam-6087	794	2	and	and	CCONJ
ejpam-6087	794	3	microsystems	microsystem	NOUN
ejpam-6087	794	4	,	,	PUNCT
ejpam-6087	794	5	94:104660	94:104660	NUM
ejpam-6087	794	6	,	,	PUNCT
ejpam-6087	794	7	2022	2022	NUM
ejpam-6087	794	8	.	.	PUNCT
ejpam-6087	795	1	[	[	X
ejpam-6087	795	2	38	38	NUM
ejpam-6087	795	3	]	]	PUNCT
ejpam-6087	795	4	tianqi	tianqi	NOUN
ejpam-6087	795	5	chen	chen	PROPN
ejpam-6087	795	6	and	and	CCONJ
ejpam-6087	795	7	carlos	carlos	PROPN
ejpam-6087	795	8	guestrin	guestrin	PROPN
ejpam-6087	795	9	.	.	PUNCT
ejpam-6087	796	1	xgboost	xgboost	X
ejpam-6087	796	2	:	:	PUNCT
ejpam-6087	797	1	a	a	DET
ejpam-6087	797	2	scalable	scalable	ADJ
ejpam-6087	797	3	tree	tree	NOUN
ejpam-6087	797	4	boosting	boost	VERB
ejpam-6087	797	5	system	system	NOUN
ejpam-6087	797	6	.	.	PUNCT
ejpam-6087	798	1	in	in	ADP
ejpam-6087	798	2	proceedings	proceeding	NOUN
ejpam-6087	798	3	of	of	ADP
ejpam-6087	798	4	the	the	DET
ejpam-6087	798	5	22nd	22nd	PROPN
ejpam-6087	798	6	acm	acm	PROPN
ejpam-6087	798	7	sigkdd	sigkdd	PROPN
ejpam-6087	798	8	international	international	ADJ
ejpam-6087	798	9	conference	conference	NOUN
ejpam-6087	798	10	on	on	ADP
ejpam-6087	798	11	knowledge	knowledge	NOUN
ejpam-6087	798	12	discovery	discovery	PROPN
ejpam-6087	798	13	and	and	CCONJ
ejpam-6087	798	14	data	datum	NOUN
ejpam-6087	798	15	mining	mining	NOUN
ejpam-6087	798	16	,	,	PUNCT
ejpam-6087	798	17	pages	page	NOUN
ejpam-6087	798	18	785–794	785–794	NUM
ejpam-6087	798	19	,	,	PUNCT
ejpam-6087	798	20	2016	2016	NUM
ejpam-6087	798	21	.	.	PUNCT
ejpam-6087	799	1	[	[	X
ejpam-6087	799	2	39	39	NUM
ejpam-6087	799	3	]	]	PUNCT
ejpam-6087	799	4	hamoud	hamoud	NOUN
ejpam-6087	799	5	aljamaan	aljamaan	PROPN
ejpam-6087	799	6	and	and	CCONJ
ejpam-6087	799	7	amal	amal	PROPN
ejpam-6087	799	8	alazba	alazba	PROPN
ejpam-6087	799	9	.	.	PUNCT
ejpam-6087	800	1	software	software	NOUN
ejpam-6087	800	2	defect	defect	NOUN
ejpam-6087	800	3	prediction	prediction	NOUN
ejpam-6087	800	4	using	use	VERB
ejpam-6087	800	5	tree	tree	NOUN
ejpam-6087	800	6	-	-	PUNCT
ejpam-6087	800	7	based	base	VERB
ejpam-6087	800	8	ensembles	ensemble	NOUN
ejpam-6087	800	9	.	.	PUNCT
ejpam-6087	801	1	in	in	ADP
ejpam-6087	801	2	proceedings	proceeding	NOUN
ejpam-6087	801	3	of	of	ADP
ejpam-6087	801	4	the	the	DET
ejpam-6087	801	5	16th	16th	ADJ
ejpam-6087	801	6	acm	acm	PROPN
ejpam-6087	801	7	international	international	ADJ
ejpam-6087	801	8	conference	conference	NOUN
ejpam-6087	801	9	on	on	ADP
ejpam-6087	801	10	predictive	predictive	ADJ
ejpam-6087	801	11	models	model	NOUN
ejpam-6087	801	12	and	and	CCONJ
ejpam-6087	801	13	data	datum	NOUN
ejpam-6087	801	14	analytics	analytic	NOUN
ejpam-6087	801	15	in	in	ADP
ejpam-6087	801	16	software	software	NOUN
ejpam-6087	801	17	engineering	engineering	NOUN
ejpam-6087	801	18	,	,	PUNCT
ejpam-6087	801	19	pages	page	NOUN
ejpam-6087	801	20	1–10	1–10	NOUN
ejpam-6087	801	21	,	,	PUNCT
ejpam-6087	801	22	2020	2020	NUM
ejpam-6087	801	23	.	.	PUNCT
ejpam-6087	802	1	[	[	X
ejpam-6087	802	2	40	40	NUM
ejpam-6087	802	3	]	]	PUNCT
ejpam-6087	802	4	mimusa	mimusa	PROPN
ejpam-6087	802	5	azim	azim	PROPN
ejpam-6087	802	6	mim	mim	PROPN
ejpam-6087	802	7	,	,	PUNCT
ejpam-6087	802	8	nazia	nazia	PROPN
ejpam-6087	802	9	majadi	majadi	NOUN
ejpam-6087	802	10	,	,	PUNCT
ejpam-6087	802	11	and	and	CCONJ
ejpam-6087	802	12	peal	peal	ADJ
ejpam-6087	802	13	mazumder	mazumder	NOUN
ejpam-6087	802	14	.	.	PUNCT
ejpam-6087	803	1	a	a	DET
ejpam-6087	803	2	soft	soft	ADJ
ejpam-6087	803	3	voting	voting	NOUN
ejpam-6087	803	4	ensemble	ensemble	ADJ
ejpam-6087	803	5	learning	learning	NOUN
ejpam-6087	803	6	approach	approach	NOUN
ejpam-6087	803	7	for	for	ADP
ejpam-6087	803	8	credit	credit	NOUN
ejpam-6087	803	9	card	card	NOUN
ejpam-6087	803	10	fraud	fraud	NOUN
ejpam-6087	803	11	detection	detection	NOUN
ejpam-6087	803	12	.	.	PUNCT
ejpam-6087	804	1	heliyon	heliyon	NOUN
ejpam-6087	804	2	,	,	PUNCT
ejpam-6087	804	3	10(3	10(3	NUM
ejpam-6087	804	4	)	)	PUNCT
ejpam-6087	804	5	,	,	PUNCT
ejpam-6087	804	6	2024	2024	NUM
ejpam-6087	804	7	.	.	PUNCT
ejpam-6087	805	1	[	[	X
ejpam-6087	805	2	41	41	NUM
ejpam-6087	805	3	]	]	X
ejpam-6087	805	4	fei	fei	PROPN
ejpam-6087	805	5	tony	tony	PROPN
ejpam-6087	805	6	liu	liu	PROPN
ejpam-6087	805	7	,	,	PUNCT
ejpam-6087	805	8	kai	kai	PROPN
ejpam-6087	805	9	ming	ming	PROPN
ejpam-6087	805	10	ting	ting	PROPN
ejpam-6087	805	11	,	,	PUNCT
ejpam-6087	805	12	and	and	CCONJ
ejpam-6087	805	13	zhi	zhi	PROPN
ejpam-6087	805	14	-	-	PUNCT
ejpam-6087	805	15	hua	hua	PROPN
ejpam-6087	805	16	zhou	zhou	PROPN
ejpam-6087	805	17	.	.	PUNCT
ejpam-6087	806	1	isolation	isolation	NOUN
ejpam-6087	806	2	forest	forest	NOUN
ejpam-6087	806	3	.	.	PUNCT
ejpam-6087	807	1	in	in	ADP
ejpam-6087	807	2	2008	2008	NUM
ejpam-6087	807	3	eighth	eighth	ADJ
ejpam-6087	807	4	ieee	ieee	PROPN
ejpam-6087	807	5	international	international	ADJ
ejpam-6087	807	6	conference	conference	NOUN
ejpam-6087	807	7	on	on	ADP
ejpam-6087	807	8	data	datum	NOUN
ejpam-6087	807	9	mining	mining	NOUN
ejpam-6087	807	10	,	,	PUNCT
ejpam-6087	807	11	pages	page	NOUN
ejpam-6087	807	12	413–422	413–422	NUM
ejpam-6087	807	13	,	,	PUNCT
ejpam-6087	807	14	2008	2008	NUM
ejpam-6087	807	15	.	.	PUNCT
ejpam-6087	808	1	[	[	X
ejpam-6087	808	2	42	42	NUM
ejpam-6087	808	3	]	]	X
ejpam-6087	808	4	antonella	antonella	PROPN
ejpam-6087	808	5	mensi	mensi	PROPN
ejpam-6087	808	6	and	and	CCONJ
ejpam-6087	808	7	manuele	manuele	PROPN
ejpam-6087	808	8	bicego	bicego	NOUN
ejpam-6087	808	9	.	.	PUNCT
ejpam-6087	809	1	enhanced	enhance	VERB
ejpam-6087	809	2	anomaly	anomaly	NOUN
ejpam-6087	809	3	scores	score	NOUN
ejpam-6087	809	4	for	for	ADP
ejpam-6087	809	5	isolation	isolation	NOUN
ejpam-6087	809	6	forests	forest	NOUN
ejpam-6087	809	7	.	.	PUNCT
ejpam-6087	810	1	pattern	pattern	NOUN
ejpam-6087	810	2	recognition	recognition	NOUN
ejpam-6087	810	3	,	,	PUNCT
ejpam-6087	810	4	120:108115	120:108115	NUM
ejpam-6087	810	5	,	,	PUNCT
ejpam-6087	810	6	2021	2021	NUM
ejpam-6087	810	7	.	.	PUNCT