id	sid	tid	token	lemma	pos
easat-10684	1	1	edelweiss	edelweiss	PROPN
easat-10684	1	2	applied	apply	VERB
easat-10684	1	3	science	science	NOUN
easat-10684	1	4	and	and	CCONJ
easat-10684	1	5	technology	technology	NOUN
easat-10684	1	6	issn	issn	PROPN
easat-10684	1	7	:	:	PUNCT
easat-10684	1	8	2576	2576	NUM
easat-10684	1	9	-	-	SYM
easat-10684	1	10	8484	8484	NUM
easat-10684	1	11	vol	vol	NOUN
easat-10684	1	12	.	.	PROPN
easat-10684	2	1	9	9	NUM
easat-10684	2	2	,	,	PUNCT
easat-10684	2	3	no	no	INTJ
easat-10684	2	4	.	.	NOUN
easat-10684	2	5	10	10	NUM
easat-10684	2	6	,	,	PUNCT
easat-10684	2	7	1515	1515	NUM
easat-10684	2	8	-	-	SYM
easat-10684	2	9	1540	1540	NUM
easat-10684	2	10	2025	2025	NUM
easat-10684	2	11	publisher	publisher	NOUN
easat-10684	2	12	:	:	PUNCT
easat-10684	2	13	learning	learn	VERB
easat-10684	2	14	gate	gate	NOUN
easat-10684	2	15	doi	doi	PROPN
easat-10684	2	16	:	:	PUNCT
easat-10684	2	17	10.55214/2576	10.55214/2576	NUM
easat-10684	2	18	-	-	SYM
easat-10684	2	19	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	3	1	©	©	PROPN
easat-10684	3	2	2025	2025	NUM
easat-10684	3	3	by	by	ADP
easat-10684	3	4	the	the	DET
easat-10684	3	5	authors	author	NOUN
easat-10684	3	6	;	;	PUNCT
easat-10684	3	7	licensee	licensee	PROPN
easat-10684	3	8	learning	learning	NOUN
easat-10684	3	9	gate	gate	NOUN
easat-10684	3	10	©	©	PROPN
easat-10684	3	11	2025	2025	NUM
easat-10684	3	12	by	by	ADP
easat-10684	3	13	the	the	DET
easat-10684	3	14	authors	author	NOUN
easat-10684	3	15	;	;	PUNCT
easat-10684	3	16	licensee	licensee	PROPN
easat-10684	3	17	learning	learning	NOUN
easat-10684	3	18	gate	gate	NOUN
easat-10684	3	19	history	history	NOUN
easat-10684	3	20	:	:	PUNCT
easat-10684	3	21	received	receive	VERB
easat-10684	3	22	:	:	PUNCT
easat-10684	3	23	25	25	NUM
easat-10684	3	24	august	august	PROPN
easat-10684	3	25	2025	2025	NUM
easat-10684	3	26	;	;	PUNCT
easat-10684	3	27	revised	revise	VERB
easat-10684	3	28	:	:	PUNCT
easat-10684	3	29	6	6	NUM
easat-10684	3	30	october	october	NOUN
easat-10684	3	31	2025	2025	NUM
easat-10684	3	32	;	;	PUNCT
easat-10684	3	33	accepted	accept	VERB
easat-10684	3	34	:	:	PUNCT
easat-10684	3	35	10	10	NUM
easat-10684	3	36	october	october	NOUN
easat-10684	3	37	2025	2025	NUM
easat-10684	3	38	;	;	PUNCT
easat-10684	3	39	published	publish	VERB
easat-10684	3	40	:	:	PUNCT
easat-10684	4	1	24	24	NUM
easat-10684	4	2	october	october	PROPN
easat-10684	4	3	2025	2025	NUM
easat-10684	4	4	*	*	PUNCT
easat-10684	4	5	correspondence	correspondence	NOUN
easat-10684	4	6	:	:	PUNCT
easat-10684	4	7	abidemi.adeniyi@bowen.edu.ng	abidemi.adeniyi@bowen.edu.ng	ADP
easat-10684	4	8	a	a	DET
easat-10684	4	9	transfer	transfer	NOUN
easat-10684	4	10	learning	learning	NOUN
easat-10684	4	11	-	-	PUNCT
easat-10684	4	12	based	base	VERB
easat-10684	4	13	model	model	NOUN
easat-10684	4	14	for	for	ADP
easat-10684	4	15	fake	fake	ADJ
easat-10684	4	16	news	news	NOUN
easat-10684	4	17	detection	detection	NOUN
easat-10684	4	18	oluwakemi	oluwakemi	NOUN
easat-10684	4	19	christaina	christaina	NOUN
easat-10684	4	20	abikoye1	abikoye1	PROPN
easat-10684	4	21	,	,	PUNCT
easat-10684	4	22	taofeek	taofeek	ADJ
easat-10684	4	23	akintunde	akintunde	ADJ
easat-10684	4	24	abdulsalam2	abdulsalam2	NOUN
easat-10684	4	25	,	,	PUNCT
easat-10684	4	26	enesi	enesi	PROPN
easat-10684	4	27	femi	femi	PROPN
easat-10684	4	28	aminu3	aminu3	PROPN
easat-10684	4	29	,	,	PUNCT
easat-10684	4	30	joseph	joseph	PROPN
easat-10684	4	31	bamidele	bamidele	PROPN
easat-10684	4	32	awotunde4	awotunde4	PROPN
easat-10684	4	33	,	,	PUNCT
easat-10684	4	34	abidemi	abidemi	PROPN
easat-10684	4	35	emmanuel	emmanuel	PROPN
easat-10684	4	36	adeniyi5	adeniyi5	PROPN
easat-10684	5	1	1department	1department	NUM
easat-10684	5	2	of	of	ADP
easat-10684	5	3	computer	computer	NOUN
easat-10684	5	4	science	science	NOUN
easat-10684	5	5	,	,	PUNCT
easat-10684	5	6	university	university	NOUN
easat-10684	5	7	of	of	ADP
easat-10684	5	8	ilorin	ilorin	NOUN
easat-10684	5	9	,	,	PUNCT
easat-10684	5	10	240003	240003	NUM
easat-10684	5	11	,	,	PUNCT
easat-10684	5	12	nigeria	nigeria	PRON
easat-10684	5	13	;	;	PUNCT
easat-10684	5	14	abikoye.o@unilorin.edu.ng	abikoye.o@unilorin.edu.ng	PROPN
easat-10684	5	15	(	(	PUNCT
easat-10684	5	16	o.c.a	o.c.a	NOUN
easat-10684	5	17	.	.	PUNCT
easat-10684	5	18	)	)	PUNCT
easat-10684	5	19	.	.	PUNCT
easat-10684	6	1	2department	2department	NUM
easat-10684	6	2	of	of	ADP
easat-10684	6	3	computer	computer	NOUN
easat-10684	6	4	science	science	NOUN
easat-10684	6	5	,	,	PUNCT
easat-10684	6	6	lens	lens	NOUN
easat-10684	6	7	polytechnic	polytechnic	ADJ
easat-10684	6	8	offa	offa	NOUN
easat-10684	6	9	,	,	PUNCT
easat-10684	6	10	kwara	kwara	PROPN
easat-10684	6	11	,	,	PUNCT
easat-10684	6	12	nigeria	nigeria	PROPN
easat-10684	6	13	;	;	PUNCT
easat-10684	6	14	binarycode1995@gmail.com	binarycode1995@gmail.com	X
easat-10684	6	15	(	(	PUNCT
easat-10684	6	16	t.a.a	t.a.a	PROPN
easat-10684	6	17	.	.	PUNCT
easat-10684	6	18	)	)	PUNCT
easat-10684	6	19	.	.	PUNCT
easat-10684	7	1	3department	3department	NUM
easat-10684	7	2	of	of	ADP
easat-10684	7	3	computer	computer	NOUN
easat-10684	7	4	science	science	NOUN
easat-10684	7	5	,	,	PUNCT
easat-10684	7	6	federal	federal	ADJ
easat-10684	7	7	university	university	PROPN
easat-10684	7	8	of	of	ADP
easat-10684	7	9	technology	technology	NOUN
easat-10684	7	10	,	,	PUNCT
easat-10684	7	11	minna	minna	NOUN
easat-10684	7	12	,	,	PUNCT
easat-10684	7	13	niger	niger	NOUN
easat-10684	7	14	,	,	PUNCT
easat-10684	7	15	nigeria	nigeria	PRON
easat-10684	7	16	;	;	PUNCT
easat-10684	7	17	binarycode1995@gmail.com	binarycode1995@gmail.com	X
easat-10684	7	18	(	(	PUNCT
easat-10684	7	19	e.f.a	e.f.a	PROPN
easat-10684	7	20	.	.	PUNCT
easat-10684	7	21	)	)	PUNCT
easat-10684	7	22	.	.	PUNCT
easat-10684	8	1	4department	4department	NUM
easat-10684	8	2	of	of	ADP
easat-10684	8	3	computer	computer	NOUN
easat-10684	8	4	science	science	NOUN
easat-10684	8	5	,	,	PUNCT
easat-10684	8	6	university	university	NOUN
easat-10684	8	7	of	of	ADP
easat-10684	8	8	ilorin	ilorin	NOUN
easat-10684	8	9	,	,	PUNCT
easat-10684	8	10	24003	24003	NUM
easat-10684	8	11	,	,	PUNCT
easat-10684	8	12	nigeria	nigeria	PRON
easat-10684	8	13	;	;	PUNCT
easat-10684	8	14	awotunde.jb@unilorin.edu.ng	awotunde.jb@unilorin.edu.ng	PUNCT
easat-10684	8	15	(	(	PUNCT
easat-10684	8	16	j.b.a	j.b.a	NOUN
easat-10684	8	17	.	.	PUNCT
easat-10684	8	18	)	)	PUNCT
easat-10684	8	19	.	.	PUNCT
easat-10684	9	1	5department	5department	NUM
easat-10684	9	2	of	of	ADP
easat-10684	9	3	computer	computer	NOUN
easat-10684	9	4	science	science	NOUN
easat-10684	9	5	,	,	PUNCT
easat-10684	9	6	college	college	NOUN
easat-10684	9	7	of	of	ADP
easat-10684	9	8	computing	computing	NOUN
easat-10684	9	9	and	and	CCONJ
easat-10684	9	10	communication	communication	NOUN
easat-10684	9	11	studies	study	NOUN
easat-10684	9	12	,	,	PUNCT
easat-10684	9	13	bowen	bowen	PROPN
easat-10684	9	14	university	university	PROPN
easat-10684	9	15	,	,	PUNCT
easat-10684	9	16	iwo	iwo	PROPN
easat-10684	9	17	,	,	PUNCT
easat-10684	9	18	nigeria	nigeria	PROPN
easat-10684	9	19	;	;	PUNCT
easat-10684	9	20	abidemi.adeniyi@bowen.edu.ng	abidemi.adeniyi@bowen.edu.ng	ADP
easat-10684	9	21	(	(	PUNCT
easat-10684	9	22	a.e.a	a.e.a	ADJ
easat-10684	9	23	.	.	PUNCT
easat-10684	9	24	)	)	PUNCT
easat-10684	9	25	.	.	PUNCT
easat-10684	10	1	5chitkara	5chitkara	NUM
easat-10684	10	2	university	university	NOUN
easat-10684	10	3	institute	institute	NOUN
easat-10684	10	4	of	of	ADP
easat-10684	10	5	engineering	engineering	NOUN
easat-10684	10	6	and	and	CCONJ
easat-10684	10	7	technology	technology	NOUN
easat-10684	10	8	,	,	PUNCT
easat-10684	10	9	chitkara	chitkara	ADJ
easat-10684	10	10	university	university	NOUN
easat-10684	10	11	,	,	PUNCT
easat-10684	10	12	punjab	punjab	PROPN
easat-10684	10	13	,	,	PUNCT
easat-10684	10	14	india	india	PROPN
easat-10684	10	15	.	.	PUNCT
easat-10684	11	1	abstract	abstract	PROPN
easat-10684	11	2	:	:	PUNCT
easat-10684	11	3	the	the	DET
easat-10684	11	4	world	world	NOUN
easat-10684	11	5	we	we	PRON
easat-10684	11	6	know	know	VERB
easat-10684	11	7	today	today	NOUN
easat-10684	11	8	has	have	VERB
easat-10684	11	9	vast	vast	ADJ
easat-10684	11	10	information	information	NOUN
easat-10684	11	11	,	,	PUNCT
easat-10684	11	12	events	event	NOUN
easat-10684	11	13	,	,	PUNCT
easat-10684	11	14	data	datum	NOUN
easat-10684	11	15	,	,	PUNCT
easat-10684	11	16	and	and	CCONJ
easat-10684	11	17	news	news	NOUN
easat-10684	11	18	spreading	spread	VERB
easat-10684	11	19	rapidly	rapidly	ADV
easat-10684	11	20	across	across	ADP
easat-10684	11	21	the	the	DET
easat-10684	11	22	globe	globe	NOUN
easat-10684	11	23	at	at	ADP
easat-10684	11	24	exponential	exponential	ADJ
easat-10684	11	25	rates	rate	NOUN
easat-10684	11	26	.	.	PUNCT
easat-10684	12	1	the	the	DET
easat-10684	12	2	invention	invention	NOUN
easat-10684	12	3	of	of	ADP
easat-10684	12	4	modern	modern	ADJ
easat-10684	12	5	telecommunication	telecommunication	NOUN
easat-10684	12	6	technology	technology	NOUN
easat-10684	12	7	facilitates	facilitate	VERB
easat-10684	12	8	this	this	DET
easat-10684	12	9	exponential	exponential	ADJ
easat-10684	12	10	growth	growth	NOUN
easat-10684	12	11	of	of	ADP
easat-10684	12	12	fake	fake	ADJ
easat-10684	12	13	news	news	NOUN
easat-10684	12	14	.	.	PUNCT
easat-10684	13	1	with	with	ADP
easat-10684	13	2	the	the	DET
easat-10684	13	3	exponential	exponential	ADJ
easat-10684	13	4	dissemination	dissemination	NOUN
easat-10684	13	5	of	of	ADP
easat-10684	13	6	news	news	NOUN
easat-10684	13	7	,	,	PUNCT
easat-10684	13	8	fake	fake	ADJ
easat-10684	13	9	news	news	NOUN
easat-10684	13	10	has	have	AUX
easat-10684	13	11	also	also	ADV
easat-10684	13	12	gained	gain	VERB
easat-10684	13	13	the	the	DET
easat-10684	13	14	potential	potential	NOUN
easat-10684	13	15	to	to	PART
easat-10684	13	16	spread	spread	VERB
easat-10684	13	17	fast	fast	ADV
easat-10684	13	18	across	across	ADP
easat-10684	13	19	social	social	ADJ
easat-10684	13	20	media	medium	NOUN
easat-10684	13	21	and	and	CCONJ
easat-10684	13	22	blog	blog	NOUN
easat-10684	13	23	post	post	NOUN
easat-10684	13	24	spaces	space	NOUN
easat-10684	13	25	.	.	PUNCT
easat-10684	14	1	fake	fake	ADJ
easat-10684	14	2	news	news	NOUN
easat-10684	14	3	is	be	AUX
easat-10684	14	4	primarily	primarily	ADV
easat-10684	14	5	aimed	aim	VERB
easat-10684	14	6	at	at	ADP
easat-10684	14	7	spreading	spread	VERB
easat-10684	14	8	false	false	ADJ
easat-10684	14	9	information	information	NOUN
easat-10684	14	10	that	that	PRON
easat-10684	14	11	is	be	AUX
easat-10684	14	12	capable	capable	ADJ
easat-10684	14	13	of	of	ADP
easat-10684	14	14	influencing	influence	VERB
easat-10684	14	15	the	the	DET
easat-10684	14	16	opinion	opinion	NOUN
easat-10684	14	17	of	of	ADP
easat-10684	14	18	the	the	DET
easat-10684	14	19	masses	masse	NOUN
easat-10684	14	20	negatively	negatively	ADV
easat-10684	14	21	.	.	PUNCT
easat-10684	15	1	previous	previous	ADJ
easat-10684	15	2	researchers	researcher	NOUN
easat-10684	15	3	have	have	AUX
easat-10684	15	4	proposed	propose	VERB
easat-10684	15	5	many	many	ADJ
easat-10684	15	6	techniques	technique	NOUN
easat-10684	15	7	,	,	PUNCT
easat-10684	15	8	including	include	VERB
easat-10684	15	9	machine	machine	NOUN
easat-10684	15	10	learning	learning	NOUN
easat-10684	15	11	and	and	CCONJ
easat-10684	15	12	deep	deep	ADJ
easat-10684	15	13	learning	learning	NOUN
easat-10684	15	14	,	,	PUNCT
easat-10684	15	15	for	for	ADP
easat-10684	15	16	fake	fake	ADJ
easat-10684	15	17	news	news	NOUN
easat-10684	15	18	detection	detection	NOUN
easat-10684	15	19	.	.	PUNCT
easat-10684	16	1	extensive	extensive	ADJ
easat-10684	16	2	studies	study	NOUN
easat-10684	16	3	show	show	VERB
easat-10684	16	4	that	that	SCONJ
easat-10684	16	5	past	past	ADJ
easat-10684	16	6	researchers	researcher	NOUN
easat-10684	16	7	have	have	AUX
easat-10684	16	8	not	not	PART
easat-10684	16	9	yet	yet	ADV
easat-10684	16	10	established	establish	VERB
easat-10684	16	11	satisfactory	satisfactory	ADJ
easat-10684	16	12	performance	performance	NOUN
easat-10684	16	13	,	,	PUNCT
easat-10684	16	14	while	while	SCONJ
easat-10684	16	15	some	some	PRON
easat-10684	16	16	suffer	suffer	VERB
easat-10684	16	17	from	from	ADP
easat-10684	16	18	model	model	NOUN
easat-10684	16	19	overfitting	overfitte	VERB
easat-10684	16	20	on	on	ADP
easat-10684	16	21	training	training	NOUN
easat-10684	16	22	sets	set	NOUN
easat-10684	16	23	with	with	ADP
easat-10684	16	24	poor	poor	ADJ
easat-10684	16	25	performance	performance	NOUN
easat-10684	16	26	on	on	ADP
easat-10684	16	27	testing	test	VERB
easat-10684	16	28	data	datum	NOUN
easat-10684	16	29	.	.	PUNCT
easat-10684	17	1	achieving	achieve	VERB
easat-10684	17	2	optimum	optimum	ADJ
easat-10684	17	3	performance	performance	NOUN
easat-10684	17	4	using	use	VERB
easat-10684	17	5	a	a	DET
easat-10684	17	6	machine	machine	NOUN
easat-10684	17	7	learning	learn	VERB
easat-10684	17	8	approach	approach	NOUN
easat-10684	17	9	is	be	AUX
easat-10684	17	10	highly	highly	ADV
easat-10684	17	11	difficult	difficult	ADJ
easat-10684	17	12	,	,	PUNCT
easat-10684	17	13	especially	especially	ADV
easat-10684	17	14	when	when	SCONJ
easat-10684	17	15	a	a	DET
easat-10684	17	16	statistical	statistical	ADJ
easat-10684	17	17	approach	approach	NOUN
easat-10684	17	18	is	be	AUX
easat-10684	17	19	adopted	adopt	VERB
easat-10684	17	20	for	for	ADP
easat-10684	17	21	the	the	DET
easat-10684	17	22	natural	natural	ADJ
easat-10684	17	23	language	language	NOUN
easat-10684	17	24	processing	processing	NOUN
easat-10684	17	25	task	task	NOUN
easat-10684	17	26	.	.	PUNCT
easat-10684	18	1	hence	hence	ADV
easat-10684	18	2	,	,	PUNCT
easat-10684	18	3	this	this	DET
easat-10684	18	4	study	study	NOUN
easat-10684	18	5	proposed	propose	VERB
easat-10684	18	6	a	a	DET
easat-10684	18	7	transfer	transfer	NOUN
easat-10684	18	8	learning	learning	NOUN
easat-10684	18	9	-	-	PUNCT
easat-10684	18	10	based	base	VERB
easat-10684	18	11	bidirectional	bidirectional	ADJ
easat-10684	18	12	long	long	ADJ
easat-10684	18	13	short	short	ADJ
easat-10684	18	14	-	-	PUNCT
easat-10684	18	15	term	term	NOUN
easat-10684	18	16	memory	memory	NOUN
easat-10684	18	17	to	to	ADP
easat-10684	18	18	machine	machine	NOUN
easat-10684	18	19	learning	learning	NOUN
easat-10684	18	20	(	(	PUNCT
easat-10684	18	21	bi	bi	NOUN
easat-10684	18	22	-	-	ADJ
easat-10684	18	23	lstm-2	lstm-2	NOUN
easat-10684	18	24	-	-	PUNCT
easat-10684	18	25	ml	ml	NOUN
easat-10684	18	26	)	)	PUNCT
easat-10684	18	27	model	model	NOUN
easat-10684	18	28	for	for	ADP
easat-10684	18	29	the	the	DET
easat-10684	18	30	detection	detection	NOUN
easat-10684	18	31	of	of	ADP
easat-10684	18	32	fake	fake	ADJ
easat-10684	18	33	news	news	NOUN
easat-10684	18	34	.	.	PUNCT
easat-10684	19	1	the	the	DET
easat-10684	19	2	bi	bi	ADJ
easat-10684	19	3	-	-	ADJ
easat-10684	19	4	lstm	lstm	ADJ
easat-10684	19	5	architecture	architecture	NOUN
easat-10684	19	6	is	be	AUX
easat-10684	19	7	pretrained	pretraine	VERB
easat-10684	19	8	on	on	ADP
easat-10684	19	9	the	the	DET
easat-10684	19	10	phoney	phoney	ADJ
easat-10684	19	11	news	news	NOUN
easat-10684	19	12	dataset	dataset	VERB
easat-10684	19	13	for	for	ADP
easat-10684	19	14	extracting	extract	VERB
easat-10684	19	15	highly	highly	ADV
easat-10684	19	16	contextualized	contextualized	ADJ
easat-10684	19	17	feature	feature	NOUN
easat-10684	19	18	embedding	embed	VERB
easat-10684	19	19	,	,	PUNCT
easat-10684	19	20	which	which	PRON
easat-10684	19	21	is	be	AUX
easat-10684	19	22	further	far	ADV
easat-10684	19	23	used	use	VERB
easat-10684	19	24	to	to	PART
easat-10684	19	25	enhance	enhance	VERB
easat-10684	19	26	the	the	DET
easat-10684	19	27	training	training	NOUN
easat-10684	19	28	process	process	NOUN
easat-10684	19	29	of	of	ADP
easat-10684	19	30	various	various	ADJ
easat-10684	19	31	machine	machine	NOUN
easat-10684	19	32	learning	learn	VERB
easat-10684	19	33	algorithms	algorithm	NOUN
easat-10684	19	34	,	,	PUNCT
easat-10684	19	35	including	include	VERB
easat-10684	19	36	svm	svm	PROPN
easat-10684	19	37	,	,	PUNCT
easat-10684	19	38	lr	lr	PROPN
easat-10684	19	39	,	,	PUNCT
easat-10684	19	40	nb	nb	PROPN
easat-10684	19	41	,	,	PUNCT
easat-10684	19	42	dt	dt	X
easat-10684	19	43	,	,	PUNCT
easat-10684	19	44	and	and	CCONJ
easat-10684	19	45	knn	knn	PROPN
easat-10684	19	46	.	.	PUNCT
easat-10684	20	1	the	the	DET
easat-10684	20	2	experimental	experimental	ADJ
easat-10684	20	3	results	result	NOUN
easat-10684	20	4	show	show	VERB
easat-10684	20	5	that	that	SCONJ
easat-10684	20	6	the	the	DET
easat-10684	20	7	bi	bi	NOUN
easat-10684	20	8	-	-	PUNCT
easat-10684	20	9	lstm-2	lstm-2	NOUN
easat-10684	20	10	-	-	PUNCT
easat-10684	20	11	ml	ml	ADV
easat-10684	20	12	fake	fake	ADJ
easat-10684	20	13	news	news	NOUN
easat-10684	20	14	detection	detection	NOUN
easat-10684	20	15	model	model	NOUN
easat-10684	20	16	yielded	yield	VERB
easat-10684	20	17	better	well	ADJ
easat-10684	20	18	performance	performance	NOUN
easat-10684	20	19	than	than	ADP
easat-10684	20	20	previous	previous	ADJ
easat-10684	20	21	approaches	approach	NOUN
easat-10684	20	22	,	,	PUNCT
easat-10684	20	23	and	and	CCONJ
easat-10684	20	24	the	the	DET
easat-10684	20	25	bi	bi	NOUN
easat-10684	20	26	-	-	PUNCT
easat-10684	20	27	lstm-2	lstm-2	NOUN
easat-10684	20	28	-	-	PUNCT
easat-10684	20	29	ml	ml	NOUN
easat-10684	20	30	model	model	NOUN
easat-10684	20	31	enhanced	enhance	VERB
easat-10684	20	32	the	the	DET
easat-10684	20	33	performance	performance	NOUN
easat-10684	20	34	of	of	ADP
easat-10684	20	35	machine	machine	NOUN
easat-10684	20	36	learning	learn	VERB
easat-10684	20	37	algorithms	algorithm	NOUN
easat-10684	20	38	with	with	ADP
easat-10684	20	39	an	an	DET
easat-10684	20	40	8.54	8.54	NUM
easat-10684	20	41	%	%	NOUN
easat-10684	20	42	average	average	ADJ
easat-10684	20	43	increase	increase	NOUN
easat-10684	20	44	in	in	ADP
easat-10684	20	45	accuracy	accuracy	NOUN
easat-10684	20	46	.	.	PUNCT
easat-10684	21	1	in	in	ADP
easat-10684	21	2	conclusion	conclusion	NOUN
easat-10684	21	3	,	,	PUNCT
easat-10684	21	4	the	the	DET
easat-10684	21	5	bi	bi	NOUN
easat-10684	21	6	-	-	PUNCT
easat-10684	21	7	lstm-2	lstm-2	NOUN
easat-10684	21	8	-	-	PUNCT
easat-10684	21	9	ml	ml	NOUN
easat-10684	21	10	model	model	NOUN
easat-10684	21	11	has	have	AUX
easat-10684	21	12	demonstrated	demonstrate	VERB
easat-10684	21	13	better	well	ADJ
easat-10684	21	14	performance	performance	NOUN
easat-10684	21	15	by	by	ADP
easat-10684	21	16	enhancing	enhance	VERB
easat-10684	21	17	machine	machine	NOUN
easat-10684	21	18	learning	learn	VERB
easat-10684	21	19	algorithms	algorithm	NOUN
easat-10684	21	20	for	for	ADP
easat-10684	21	21	detecting	detect	VERB
easat-10684	21	22	fake	fake	ADJ
easat-10684	21	23	news	news	NOUN
easat-10684	21	24	.	.	PUNCT
easat-10684	22	1	keywords	keyword	NOUN
easat-10684	22	2	:	:	PUNCT
easat-10684	22	3	bi	bi	NOUN
easat-10684	22	4	-	-	ADJ
easat-10684	22	5	lstm	lstm	ADJ
easat-10684	22	6	,	,	PUNCT
easat-10684	22	7	deep	deep	ADJ
easat-10684	22	8	learning	learning	NOUN
easat-10684	22	9	,	,	PUNCT
easat-10684	22	10	internet	internet	NOUN
easat-10684	22	11	,	,	PUNCT
easat-10684	22	12	fake	fake	ADJ
easat-10684	22	13	news	news	NOUN
easat-10684	22	14	,	,	PUNCT
easat-10684	22	15	machine	machine	NOUN
easat-10684	22	16	learning	learning	NOUN
easat-10684	22	17	,	,	PUNCT
easat-10684	22	18	social	social	ADJ
easat-10684	22	19	media	medium	NOUN
easat-10684	22	20	,	,	PUNCT
easat-10684	22	21	transfer	transfer	NOUN
easat-10684	22	22	learning	learning	NOUN
easat-10684	22	23	.	.	PUNCT
easat-10684	23	1	1	1	X
easat-10684	23	2	.	.	X
easat-10684	23	3	introduction	introduction	NOUN
easat-10684	23	4	in	in	ADP
easat-10684	23	5	recent	recent	ADJ
easat-10684	23	6	years	year	NOUN
easat-10684	23	7	,	,	PUNCT
easat-10684	23	8	social	social	ADJ
easat-10684	23	9	media	medium	NOUN
easat-10684	23	10	has	have	AUX
easat-10684	23	11	emerged	emerge	VERB
easat-10684	23	12	as	as	ADP
easat-10684	23	13	the	the	DET
easat-10684	23	14	primary	primary	ADJ
easat-10684	23	15	platform	platform	NOUN
easat-10684	23	16	for	for	ADP
easat-10684	23	17	news	news	NOUN
easat-10684	23	18	and	and	CCONJ
easat-10684	23	19	information	information	NOUN
easat-10684	23	20	dissemination	dissemination	NOUN
easat-10684	23	21	,	,	PUNCT
easat-10684	23	22	facilitated	facilitate	VERB
easat-10684	23	23	by	by	ADP
easat-10684	23	24	the	the	DET
easat-10684	23	25	internet	internet	NOUN
easat-10684	23	26	,	,	PUNCT
easat-10684	23	27	enabling	enable	VERB
easat-10684	23	28	an	an	DET
easat-10684	23	29	environment	environment	NOUN
easat-10684	23	30	for	for	ADP
easat-10684	23	31	the	the	DET
easat-10684	23	32	easy	easy	ADJ
easat-10684	23	33	posting	posting	NOUN
easat-10684	23	34	of	of	ADP
easat-10684	23	35	media	medium	NOUN
easat-10684	23	36	content	content	NOUN
easat-10684	23	37	[	[	X
easat-10684	23	38	1	1	NUM
easat-10684	23	39	]	]	PUNCT
easat-10684	23	40	.	.	PUNCT
easat-10684	24	1	recent	recent	ADJ
easat-10684	24	2	research	research	NOUN
easat-10684	24	3	indicates	indicate	VERB
easat-10684	24	4	a	a	DET
easat-10684	24	5	shift	shift	NOUN
easat-10684	24	6	in	in	ADP
easat-10684	24	7	news	news	NOUN
easat-10684	24	8	consumption	consumption	NOUN
easat-10684	24	9	,	,	PUNCT
easat-10684	24	10	with	with	ADP
easat-10684	24	11	many	many	ADJ
easat-10684	24	12	users	user	NOUN
easat-10684	24	13	preferring	prefer	VERB
easat-10684	24	14	digital	digital	NOUN
easat-10684	24	15	-	-	PUNCT
easat-10684	24	16	based	base	VERB
easat-10684	24	17	news	news	NOUN
easat-10684	24	18	over	over	ADP
easat-10684	24	19	traditional	traditional	ADJ
easat-10684	24	20	newspapers	newspaper	NOUN
easat-10684	24	21	.	.	PUNCT
easat-10684	25	1	however	however	ADV
easat-10684	25	2	,	,	PUNCT
easat-10684	25	3	the	the	DET
easat-10684	25	4	authenticity	authenticity	NOUN
easat-10684	25	5	of	of	ADP
easat-10684	25	6	news	news	NOUN
easat-10684	25	7	on	on	ADP
easat-10684	25	8	social	social	ADJ
easat-10684	25	9	media	medium	NOUN
easat-10684	25	10	(	(	PUNCT
easat-10684	25	11	digital	digital	ADJ
easat-10684	25	12	news	news	NOUN
easat-10684	25	13	)	)	PUNCT
easat-10684	25	14	is	be	AUX
easat-10684	25	15	a	a	DET
easat-10684	25	16	challenging	challenging	ADJ
easat-10684	25	17	process	process	NOUN
easat-10684	25	18	,	,	PUNCT
easat-10684	25	19	unlike	unlike	ADP
easat-10684	25	20	content	content	NOUN
easat-10684	25	21	from	from	ADP
easat-10684	25	22	radio	radio	NOUN
easat-10684	25	23	and	and	CCONJ
easat-10684	25	24	television	television	NOUN
easat-10684	25	25	,	,	PUNCT
easat-10684	25	26	which	which	PRON
easat-10684	25	27	are	be	AUX
easat-10684	25	28	subjected	subject	VERB
easat-10684	25	29	to	to	ADP
easat-10684	25	30	critical	critical	ADJ
easat-10684	25	31	review	review	NOUN
easat-10684	25	32	and	and	CCONJ
easat-10684	25	33	supervision	supervision	NOUN
easat-10684	25	34	before	before	ADP
easat-10684	25	35	broadcasting	broadcast	VERB
easat-10684	25	36	[	[	X
easat-10684	25	37	2	2	NUM
easat-10684	25	38	]	]	PUNCT
easat-10684	25	39	.	.	PUNCT
easat-10684	26	1	the	the	DET
easat-10684	26	2	detection	detection	NOUN
easat-10684	26	3	of	of	ADP
easat-10684	26	4	fake	fake	ADJ
easat-10684	26	5	news	news	NOUN
easat-10684	26	6	remains	remain	VERB
easat-10684	26	7	essential	essential	ADJ
easat-10684	26	8	due	due	ADP
easat-10684	26	9	to	to	ADP
easat-10684	26	10	the	the	DET
easat-10684	26	11	widespread	widespread	ADJ
easat-10684	26	12	dissemination	dissemination	NOUN
easat-10684	26	13	of	of	ADP
easat-10684	26	14	misinformation	misinformation	NOUN
easat-10684	26	15	,	,	PUNCT
easat-10684	26	16	rumors	rumor	NOUN
easat-10684	26	17	,	,	PUNCT
easat-10684	26	18	and	and	CCONJ
easat-10684	26	19	fake	fake	ADJ
easat-10684	26	20	news	news	NOUN
easat-10684	26	21	on	on	ADP
easat-10684	26	22	platforms	platform	NOUN
easat-10684	26	23	such	such	ADJ
easat-10684	26	24	as	as	ADP
easat-10684	26	25	facebook	facebook	PROPN
easat-10684	26	26	,	,	PUNCT
easat-10684	26	27	instagram	instagram	PROPN
easat-10684	26	28	,	,	PUNCT
easat-10684	26	29	twitter	twitter	NOUN
easat-10684	26	30	,	,	PUNCT
easat-10684	26	31	and	and	CCONJ
easat-10684	26	32	whatsapp	whatsapp	NOUN
easat-10684	26	33	[	[	X
easat-10684	26	34	3	3	NUM
easat-10684	26	35	]	]	PUNCT
easat-10684	26	36	.	.	PUNCT
easat-10684	27	1	training	train	VERB
easat-10684	27	2	intelligent	intelligent	ADJ
easat-10684	27	3	systems	system	NOUN
easat-10684	27	4	to	to	PART
easat-10684	27	5	detect	detect	VERB
easat-10684	27	6	fake	fake	ADJ
easat-10684	27	7	news	news	NOUN
easat-10684	27	8	online	online	NOUN
easat-10684	27	9	is	be	AUX
easat-10684	27	10	challenging	challenge	VERB
easat-10684	27	11	due	due	ADJ
easat-10684	27	12	to	to	ADP
easat-10684	27	13	the	the	DET
easat-10684	27	14	diverse	diverse	ADJ
easat-10684	27	15	forms	form	NOUN
easat-10684	27	16	and	and	CCONJ
easat-10684	27	17	sources	source	NOUN
easat-10684	27	18	of	of	ADP
easat-10684	27	19	interdisciplinary	interdisciplinary	ADJ
easat-10684	27	20	news	news	NOUN
easat-10684	27	21	.	.	PUNCT
easat-10684	28	1	to	to	PART
easat-10684	28	2	effectively	effectively	ADV
easat-10684	28	3	detect	detect	VERB
easat-10684	28	4	phony	phony	ADJ
easat-10684	28	5	news	news	NOUN
easat-10684	28	6	,	,	PUNCT
easat-10684	28	7	various	various	ADJ
easat-10684	28	8	1516	1516	NUM
easat-10684	28	9	edelweiss	edelweiss	PROPN
easat-10684	28	10	applied	apply	VERB
easat-10684	28	11	science	science	NOUN
easat-10684	28	12	and	and	CCONJ
easat-10684	28	13	technology	technology	NOUN
easat-10684	28	14	issn	issn	PROPN
easat-10684	28	15	:	:	PUNCT
easat-10684	28	16	2576	2576	NUM
easat-10684	28	17	-	-	SYM
easat-10684	28	18	8484	8484	NUM
easat-10684	28	19	vol	vol	NOUN
easat-10684	28	20	.	.	PROPN
easat-10684	29	1	9	9	NUM
easat-10684	29	2	,	,	PUNCT
easat-10684	29	3	no	no	INTJ
easat-10684	29	4	.	.	NOUN
easat-10684	29	5	10	10	NUM
easat-10684	29	6	:	:	PUNCT
easat-10684	29	7	1515	1515	NUM
easat-10684	29	8	-	-	SYM
easat-10684	29	9	1540	1540	NUM
easat-10684	29	10	,	,	PUNCT
easat-10684	29	11	2025	2025	NUM
easat-10684	29	12	doi	doi	NOUN
easat-10684	29	13	:	:	PUNCT
easat-10684	29	14	10.55214/2576	10.55214/2576	NUM
easat-10684	29	15	-	-	SYM
easat-10684	29	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	30	1	©	©	PROPN
easat-10684	30	2	2025	2025	NUM
easat-10684	30	3	by	by	ADP
easat-10684	30	4	the	the	DET
easat-10684	30	5	authors	author	NOUN
easat-10684	30	6	;	;	PUNCT
easat-10684	30	7	licensee	licensee	PROPN
easat-10684	30	8	learning	learn	VERB
easat-10684	30	9	gate	gate	NOUN
easat-10684	30	10	natural	natural	ADJ
easat-10684	30	11	language	language	NOUN
easat-10684	30	12	processing	processing	NOUN
easat-10684	30	13	(	(	PUNCT
easat-10684	30	14	nlp	nlp	ADJ
easat-10684	30	15	)	)	PUNCT
easat-10684	30	16	tools	tool	NOUN
easat-10684	30	17	,	,	PUNCT
easat-10684	30	18	such	such	ADJ
easat-10684	30	19	as	as	ADP
easat-10684	30	20	part	part	NOUN
easat-10684	30	21	-	-	PUNCT
easat-10684	30	22	of	of	ADP
easat-10684	30	23	-	-	PUNCT
easat-10684	30	24	speech	speech	NOUN
easat-10684	30	25	tagging	tagging	NOUN
easat-10684	30	26	,	,	PUNCT
easat-10684	30	27	sentiment	sentiment	NOUN
easat-10684	30	28	analysis	analysis	NOUN
easat-10684	30	29	,	,	PUNCT
easat-10684	30	30	and	and	CCONJ
easat-10684	30	31	vectorization	vectorization	NOUN
easat-10684	30	32	techniques	technique	NOUN
easat-10684	30	33	,	,	PUNCT
easat-10684	30	34	need	need	VERB
easat-10684	30	35	to	to	PART
easat-10684	30	36	be	be	AUX
easat-10684	30	37	adopted	adopt	VERB
easat-10684	30	38	[	[	PUNCT
easat-10684	30	39	4	4	NUM
easat-10684	30	40	]	]	PUNCT
easat-10684	30	41	.	.	PUNCT
easat-10684	31	1	additionally	additionally	ADV
easat-10684	31	2	,	,	PUNCT
easat-10684	31	3	other	other	ADJ
easat-10684	31	4	resources	resource	NOUN
easat-10684	31	5	and	and	CCONJ
easat-10684	31	6	tools	tool	NOUN
easat-10684	31	7	are	be	AUX
easat-10684	31	8	available	available	ADJ
easat-10684	31	9	to	to	PART
easat-10684	31	10	help	help	VERB
easat-10684	31	11	individuals	individual	NOUN
easat-10684	31	12	identify	identify	VERB
easat-10684	31	13	and	and	CCONJ
easat-10684	31	14	spot	spot	VERB
easat-10684	31	15	fake	fake	ADJ
easat-10684	31	16	news	news	NOUN
easat-10684	31	17	on	on	ADP
easat-10684	31	18	their	their	PRON
easat-10684	31	19	social	social	ADJ
easat-10684	31	20	media	medium	NOUN
easat-10684	31	21	feeds	feed	VERB
easat-10684	31	22	by	by	ADP
easat-10684	31	23	analyzing	analyze	VERB
easat-10684	31	24	essential	essential	ADJ
easat-10684	31	25	facts	fact	NOUN
easat-10684	31	26	and	and	CCONJ
easat-10684	31	27	critically	critically	ADV
easat-10684	31	28	evaluating	evaluate	VERB
easat-10684	31	29	news	news	NOUN
easat-10684	31	30	content	content	NOUN
easat-10684	31	31	.	.	PUNCT
easat-10684	32	1	the	the	DET
easat-10684	32	2	primary	primary	ADJ
easat-10684	32	3	objective	objective	NOUN
easat-10684	32	4	of	of	ADP
easat-10684	32	5	fake	fake	ADJ
easat-10684	32	6	news	news	NOUN
easat-10684	32	7	detection	detection	NOUN
easat-10684	32	8	is	be	AUX
easat-10684	32	9	to	to	PART
easat-10684	32	10	prevent	prevent	VERB
easat-10684	32	11	the	the	DET
easat-10684	32	12	quick	quick	ADJ
easat-10684	32	13	dissemination	dissemination	NOUN
easat-10684	32	14	of	of	ADP
easat-10684	32	15	false	false	ADJ
easat-10684	32	16	information	information	NOUN
easat-10684	32	17	across	across	ADP
easat-10684	32	18	various	various	ADJ
easat-10684	32	19	platforms	platform	NOUN
easat-10684	32	20	,	,	PUNCT
easat-10684	32	21	including	include	VERB
easat-10684	32	22	social	social	ADJ
easat-10684	32	23	media	medium	NOUN
easat-10684	32	24	and	and	CCONJ
easat-10684	32	25	messaging	message	VERB
easat-10684	32	26	applications	application	NOUN
easat-10684	32	27	.	.	PUNCT
easat-10684	33	1	fake	fake	ADJ
easat-10684	33	2	news	news	NOUN
easat-10684	33	3	detection	detection	NOUN
easat-10684	33	4	systems	system	NOUN
easat-10684	33	5	can	can	AUX
easat-10684	33	6	result	result	VERB
easat-10684	33	7	in	in	ADP
easat-10684	33	8	harmful	harmful	ADJ
easat-10684	33	9	consequences	consequence	NOUN
easat-10684	33	10	,	,	PUNCT
easat-10684	33	11	such	such	ADJ
easat-10684	33	12	as	as	ADP
easat-10684	33	13	mob	mob	NOUN
easat-10684	33	14	violence	violence	NOUN
easat-10684	33	15	,	,	PUNCT
easat-10684	33	16	if	if	SCONJ
easat-10684	33	17	not	not	PART
easat-10684	33	18	properly	properly	ADV
easat-10684	33	19	addressed	address	VERB
easat-10684	33	20	[	[	X
easat-10684	33	21	3	3	NUM
easat-10684	33	22	]	]	PUNCT
easat-10684	33	23	.	.	PUNCT
easat-10684	34	1	additionally	additionally	ADV
easat-10684	34	2	,	,	PUNCT
easat-10684	34	3	hossain	hossain	PROPN
easat-10684	34	4	et	et	PROPN
easat-10684	34	5	al	al	PROPN
easat-10684	34	6	.	.	PUNCT
easat-10684	35	1	[	[	X
easat-10684	35	2	5	5	NUM
easat-10684	35	3	]	]	PUNCT
easat-10684	35	4	revealed	reveal	VERB
easat-10684	35	5	that	that	SCONJ
easat-10684	35	6	fake	fake	ADJ
easat-10684	35	7	news	news	NOUN
easat-10684	35	8	detection	detection	NOUN
easat-10684	35	9	aims	aim	VERB
easat-10684	35	10	to	to	PART
easat-10684	35	11	identify	identify	VERB
easat-10684	35	12	misleading	mislead	VERB
easat-10684	35	13	information	information	NOUN
easat-10684	35	14	and	and	CCONJ
easat-10684	35	15	prevent	prevent	VERB
easat-10684	35	16	unwarranted	unwarranted	ADJ
easat-10684	35	17	acts	act	NOUN
easat-10684	35	18	of	of	ADP
easat-10684	35	19	violence	violence	NOUN
easat-10684	35	20	,	,	PUNCT
easat-10684	35	21	which	which	PRON
easat-10684	35	22	could	could	AUX
easat-10684	35	23	ultimately	ultimately	ADV
easat-10684	35	24	safeguard	safeguard	VERB
easat-10684	35	25	societal	societal	ADJ
easat-10684	35	26	peace	peace	NOUN
easat-10684	35	27	.	.	PUNCT
easat-10684	36	1	researchers	researcher	NOUN
easat-10684	36	2	have	have	AUX
easat-10684	36	3	proposed	propose	VERB
easat-10684	36	4	machine	machine	NOUN
easat-10684	36	5	learning	learn	VERB
easat-10684	36	6	algorithms	algorithm	NOUN
easat-10684	36	7	and	and	CCONJ
easat-10684	36	8	deep	deep	ADJ
easat-10684	36	9	learning	learning	NOUN
easat-10684	36	10	techniques	technique	NOUN
easat-10684	36	11	in	in	ADP
easat-10684	36	12	the	the	DET
easat-10684	36	13	past	past	NOUN
easat-10684	36	14	for	for	ADP
easat-10684	36	15	fake	fake	ADJ
easat-10684	36	16	news	news	NOUN
easat-10684	36	17	detection	detection	NOUN
easat-10684	36	18	,	,	PUNCT
easat-10684	36	19	including	include	VERB
easat-10684	36	20	support	support	NOUN
easat-10684	36	21	vector	vector	NOUN
easat-10684	36	22	machine	machine	NOUN
easat-10684	36	23	(	(	PUNCT
easat-10684	36	24	svm	svm	PROPN
easat-10684	36	25	)	)	PUNCT
easat-10684	36	26	,	,	PUNCT
easat-10684	36	27	naïve	naïve	ADJ
easat-10684	36	28	bayes	bayes	PROPN
easat-10684	36	29	(	(	PUNCT
easat-10684	36	30	nb	nb	PROPN
easat-10684	36	31	)	)	PUNCT
easat-10684	36	32	,	,	PUNCT
easat-10684	36	33	k	k	X
easat-10684	36	34	-	-	PUNCT
easat-10684	36	35	nearest	near	ADJ
easat-10684	36	36	neighbour	neighbour	NOUN
easat-10684	36	37	(	(	PUNCT
easat-10684	36	38	knn	knn	PROPN
easat-10684	36	39	)	)	PUNCT
easat-10684	36	40	,	,	PUNCT
easat-10684	36	41	logistic	logistic	ADJ
easat-10684	36	42	regression	regression	NOUN
easat-10684	36	43	(	(	PUNCT
easat-10684	36	44	lr	lr	NOUN
easat-10684	36	45	)	)	PUNCT
easat-10684	36	46	,	,	PUNCT
easat-10684	36	47	decision	decision	NOUN
easat-10684	36	48	tree	tree	NOUN
easat-10684	36	49	(	(	PUNCT
easat-10684	36	50	dt	dt	PROPN
easat-10684	36	51	)	)	PUNCT
easat-10684	36	52	,	,	PUNCT
easat-10684	36	53	recurrent	recurrent	ADJ
easat-10684	36	54	neural	neural	ADJ
easat-10684	36	55	network	network	NOUN
easat-10684	36	56	(	(	PUNCT
easat-10684	36	57	rnn	rnn	PROPN
easat-10684	36	58	)	)	PUNCT
easat-10684	36	59	,	,	PUNCT
easat-10684	36	60	gated	gate	VERB
easat-10684	36	61	recurrent	recurrent	ADJ
easat-10684	36	62	unit	unit	NOUN
easat-10684	36	63	(	(	PUNCT
easat-10684	36	64	gru	gru	PROPN
easat-10684	36	65	)	)	PUNCT
easat-10684	36	66	,	,	PUNCT
easat-10684	36	67	long	long	ADJ
easat-10684	36	68	short	short	ADJ
easat-10684	36	69	-	-	PUNCT
easat-10684	36	70	term	term	NOUN
easat-10684	36	71	memory	memory	NOUN
easat-10684	36	72	(	(	PUNCT
easat-10684	36	73	lstm	lstm	NOUN
easat-10684	36	74	)	)	PUNCT
easat-10684	36	75	,	,	PUNCT
easat-10684	36	76	and	and	CCONJ
easat-10684	36	77	bidirectional	bidirectional	ADJ
easat-10684	36	78	long	long	ADJ
easat-10684	36	79	short	short	ADJ
easat-10684	36	80	-	-	PUNCT
easat-10684	36	81	term	term	NOUN
easat-10684	36	82	memory	memory	NOUN
easat-10684	36	83	(	(	PUNCT
easat-10684	36	84	bi	bi	NOUN
easat-10684	36	85	-	-	ADJ
easat-10684	36	86	lstm	lstm	ADJ
easat-10684	36	87	)	)	PUNCT
easat-10684	37	1	[	[	X
easat-10684	37	2	6	6	NUM
easat-10684	37	3	]	]	PUNCT
easat-10684	37	4	.	.	PUNCT
easat-10684	38	1	artificial	artificial	ADJ
easat-10684	38	2	intelligence	intelligence	NOUN
easat-10684	38	3	(	(	PUNCT
easat-10684	38	4	ai	ai	PROPN
easat-10684	38	5	)	)	PUNCT
easat-10684	38	6	is	be	AUX
easat-10684	38	7	a	a	DET
easat-10684	38	8	broad	broad	ADJ
easat-10684	38	9	term	term	NOUN
easat-10684	38	10	denoting	denote	VERB
easat-10684	38	11	the	the	DET
easat-10684	38	12	field	field	NOUN
easat-10684	38	13	of	of	ADP
easat-10684	38	14	study	study	NOUN
easat-10684	38	15	where	where	SCONJ
easat-10684	38	16	machines	machine	NOUN
easat-10684	38	17	are	be	AUX
easat-10684	38	18	trained	train	VERB
easat-10684	38	19	to	to	PART
easat-10684	38	20	perform	perform	VERB
easat-10684	38	21	intelligent	intelligent	ADJ
easat-10684	38	22	tasks	task	NOUN
easat-10684	38	23	without	without	ADP
easat-10684	38	24	human	human	ADJ
easat-10684	38	25	intervention	intervention	NOUN
easat-10684	38	26	[	[	X
easat-10684	38	27	7	7	NUM
easat-10684	38	28	]	]	PUNCT
easat-10684	38	29	.	.	PUNCT
easat-10684	39	1	the	the	DET
easat-10684	39	2	ai	ai	PROPN
easat-10684	39	3	domain	domain	NOUN
easat-10684	39	4	is	be	AUX
easat-10684	39	5	divided	divide	VERB
easat-10684	39	6	into	into	ADP
easat-10684	39	7	subfields	subfield	NOUN
easat-10684	39	8	such	such	ADJ
easat-10684	39	9	as	as	ADP
easat-10684	39	10	knowledge	knowledge	NOUN
easat-10684	39	11	processing	processing	NOUN
easat-10684	39	12	,	,	PUNCT
easat-10684	39	13	pattern	pattern	NOUN
easat-10684	39	14	recognition	recognition	NOUN
easat-10684	39	15	,	,	PUNCT
easat-10684	39	16	natural	natural	ADJ
easat-10684	39	17	language	language	NOUN
easat-10684	39	18	processing	processing	NOUN
easat-10684	39	19	,	,	PUNCT
easat-10684	39	20	deep	deep	ADJ
easat-10684	39	21	learning	learning	NOUN
easat-10684	39	22	,	,	PUNCT
easat-10684	39	23	and	and	CCONJ
easat-10684	39	24	machine	machine	NOUN
easat-10684	39	25	learning	learn	VERB
easat-10684	39	26	for	for	ADP
easat-10684	39	27	building	build	VERB
easat-10684	39	28	intelligent	intelligent	ADJ
easat-10684	39	29	systems	system	NOUN
easat-10684	39	30	.	.	PUNCT
easat-10684	40	1	2021	2021	NUM
easat-10684	40	2	reveals	reveal	VERB
easat-10684	40	3	that	that	SCONJ
easat-10684	40	4	machine	machine	NOUN
easat-10684	40	5	learning	learn	VERB
easat-10684	40	6	technology	technology	NOUN
easat-10684	40	7	greatly	greatly	ADV
easat-10684	40	8	reduces	reduce	VERB
easat-10684	40	9	human	human	ADJ
easat-10684	40	10	work	work	NOUN
easat-10684	40	11	by	by	ADP
easat-10684	40	12	using	use	VERB
easat-10684	40	13	statistical	statistical	ADJ
easat-10684	40	14	and	and	CCONJ
easat-10684	40	15	mathematical	mathematical	ADJ
easat-10684	40	16	algorithms	algorithms	NOUN
easat-10684	40	17	to	to	PART
easat-10684	40	18	achieve	achieve	VERB
easat-10684	40	19	a	a	DET
easat-10684	40	20	task	task	NOUN
easat-10684	40	21	more	more	ADV
easat-10684	40	22	efficiently	efficiently	ADV
easat-10684	40	23	.	.	PUNCT
easat-10684	41	1	natural	natural	ADJ
easat-10684	41	2	language	language	NOUN
easat-10684	41	3	processing	processing	NOUN
easat-10684	41	4	(	(	PUNCT
easat-10684	41	5	nlp	nlp	NOUN
easat-10684	41	6	)	)	PUNCT
easat-10684	41	7	provides	provide	VERB
easat-10684	41	8	the	the	DET
easat-10684	41	9	tools	tool	NOUN
easat-10684	41	10	and	and	CCONJ
easat-10684	41	11	technology	technology	NOUN
easat-10684	41	12	to	to	PART
easat-10684	41	13	build	build	VERB
easat-10684	41	14	communication	communication	NOUN
easat-10684	41	15	interfaces	interface	NOUN
easat-10684	41	16	between	between	ADP
easat-10684	41	17	humans	human	NOUN
easat-10684	41	18	and	and	CCONJ
easat-10684	41	19	computers	computer	NOUN
easat-10684	41	20	using	use	VERB
easat-10684	41	21	natural	natural	ADJ
easat-10684	41	22	languages	language	NOUN
easat-10684	41	23	and	and	CCONJ
easat-10684	41	24	computational	computational	ADJ
easat-10684	41	25	algorithms	algorithm	NOUN
easat-10684	41	26	[	[	X
easat-10684	41	27	8	8	NUM
easat-10684	41	28	]	]	PUNCT
easat-10684	41	29	.	.	PUNCT
easat-10684	42	1	nlp	nlp	PROPN
easat-10684	42	2	has	have	VERB
easat-10684	42	3	wide	wide	ADJ
easat-10684	42	4	applications	application	NOUN
easat-10684	42	5	in	in	ADP
easat-10684	42	6	medicine	medicine	NOUN
easat-10684	42	7	,	,	PUNCT
easat-10684	42	8	language	language	NOUN
easat-10684	42	9	translation	translation	NOUN
easat-10684	42	10	,	,	PUNCT
easat-10684	42	11	sentiment	sentiment	NOUN
easat-10684	42	12	analysis	analysis	NOUN
easat-10684	42	13	,	,	PUNCT
easat-10684	42	14	and	and	CCONJ
easat-10684	42	15	false	false	ADJ
easat-10684	42	16	news	news	NOUN
easat-10684	42	17	detection	detection	NOUN
easat-10684	42	18	khurana	khurana	PROPN
easat-10684	42	19	et	et	PROPN
easat-10684	42	20	al	al	PROPN
easat-10684	42	21	.	.	PUNCT
easat-10684	43	1	[	[	X
easat-10684	43	2	9	9	NUM
easat-10684	43	3	]	]	PUNCT
easat-10684	43	4	.	.	PUNCT
easat-10684	44	1	patel	patel	PROPN
easat-10684	44	2	and	and	CCONJ
easat-10684	44	3	patel	patel	PROPN
easat-10684	45	1	[	[	X
easat-10684	45	2	10	10	NUM
easat-10684	45	3	]	]	PUNCT
easat-10684	45	4	confirmed	confirm	VERB
easat-10684	45	5	that	that	SCONJ
easat-10684	45	6	deep	deep	ADJ
easat-10684	45	7	learning	learning	NOUN
easat-10684	45	8	models	model	NOUN
easat-10684	45	9	have	have	AUX
easat-10684	45	10	been	be	AUX
easat-10684	45	11	successfully	successfully	ADV
easat-10684	45	12	deployed	deploy	VERB
easat-10684	45	13	to	to	PART
easat-10684	45	14	address	address	VERB
easat-10684	45	15	natural	natural	ADJ
easat-10684	45	16	language	language	NOUN
easat-10684	45	17	processing	processing	NOUN
easat-10684	45	18	tasks	task	NOUN
easat-10684	45	19	with	with	ADP
easat-10684	45	20	high	high	ADJ
easat-10684	45	21	-	-	PUNCT
easat-10684	45	22	performance	performance	NOUN
easat-10684	45	23	records	record	NOUN
easat-10684	45	24	;	;	PUNCT
easat-10684	45	25	commonly	commonly	ADV
easat-10684	45	26	used	use	VERB
easat-10684	45	27	algorithms	algorithm	NOUN
easat-10684	45	28	include	include	VERB
easat-10684	45	29	recurrent	recurrent	ADJ
easat-10684	45	30	neural	neural	ADJ
easat-10684	45	31	networks	network	NOUN
easat-10684	45	32	(	(	PUNCT
easat-10684	45	33	rnn	rnn	PROPN
easat-10684	45	34	)	)	PUNCT
easat-10684	45	35	,	,	PUNCT
easat-10684	45	36	long	long	ADJ
easat-10684	45	37	short	short	ADJ
easat-10684	45	38	-	-	PUNCT
easat-10684	45	39	term	term	NOUN
easat-10684	45	40	memory	memory	NOUN
easat-10684	45	41	(	(	PUNCT
easat-10684	45	42	lstm	lstm	NOUN
easat-10684	45	43	)	)	PUNCT
easat-10684	45	44	,	,	PUNCT
easat-10684	45	45	gated	gate	VERB
easat-10684	45	46	recurrent	recurrent	ADJ
easat-10684	45	47	unit	unit	NOUN
easat-10684	45	48	(	(	PUNCT
easat-10684	45	49	gru	gru	PROPN
easat-10684	45	50	)	)	PUNCT
easat-10684	45	51	,	,	PUNCT
easat-10684	45	52	convolutional	convolutional	ADJ
easat-10684	45	53	neural	neural	ADJ
easat-10684	45	54	network	network	NOUN
easat-10684	45	55	(	(	PUNCT
easat-10684	45	56	cnn	cnn	PROPN
easat-10684	45	57	)	)	PUNCT
easat-10684	45	58	,	,	PUNCT
easat-10684	45	59	and	and	CCONJ
easat-10684	45	60	bi	bi	NOUN
easat-10684	45	61	-	-	NOUN
easat-10684	45	62	lstm	lstm	ADJ
easat-10684	45	63	.	.	PUNCT
easat-10684	46	1	the	the	DET
easat-10684	46	2	rnn	rnn	NOUN
easat-10684	46	3	architecture	architecture	NOUN
easat-10684	46	4	is	be	AUX
easat-10684	46	5	the	the	DET
easat-10684	46	6	basis	basis	NOUN
easat-10684	46	7	of	of	ADP
easat-10684	46	8	all	all	DET
easat-10684	46	9	recurrent	recurrent	NOUN
easat-10684	46	10	-	-	PUNCT
easat-10684	46	11	based	base	VERB
easat-10684	46	12	network	network	NOUN
easat-10684	46	13	architectures	architecture	NOUN
easat-10684	46	14	,	,	PUNCT
easat-10684	46	15	which	which	PRON
easat-10684	46	16	are	be	AUX
easat-10684	46	17	primarily	primarily	ADV
easat-10684	46	18	used	use	VERB
easat-10684	46	19	for	for	ADP
easat-10684	46	20	natural	natural	ADJ
easat-10684	46	21	language	language	NOUN
easat-10684	46	22	processing	processing	NOUN
easat-10684	46	23	tasks	task	NOUN
easat-10684	46	24	and	and	CCONJ
easat-10684	46	25	processing	processing	NOUN
easat-10684	46	26	of	of	ADP
easat-10684	46	27	sequential	sequential	ADJ
easat-10684	46	28	-	-	PUNCT
easat-10684	46	29	based	base	VERB
easat-10684	46	30	data	datum	NOUN
easat-10684	46	31	(	(	PUNCT
easat-10684	46	32	text	text	NOUN
easat-10684	46	33	data	datum	NOUN
easat-10684	46	34	)	)	PUNCT
easat-10684	46	35	in	in	ADP
easat-10684	46	36	a	a	DET
easat-10684	46	37	feed	feed	NOUN
easat-10684	46	38	-	-	PUNCT
easat-10684	46	39	forward	forward	NOUN
easat-10684	46	40	manner	manner	NOUN
easat-10684	46	41	[	[	X
easat-10684	46	42	10	10	NUM
easat-10684	46	43	]	]	PUNCT
easat-10684	46	44	.	.	PUNCT
easat-10684	47	1	long	long	ADJ
easat-10684	47	2	short	short	ADJ
easat-10684	47	3	-	-	PUNCT
easat-10684	47	4	term	term	NOUN
easat-10684	47	5	memory	memory	NOUN
easat-10684	47	6	(	(	PUNCT
easat-10684	47	7	lstm	lstm	NOUN
easat-10684	47	8	)	)	PUNCT
easat-10684	47	9	architecture	architecture	NOUN
easat-10684	47	10	comprehensively	comprehensively	ADV
easat-10684	47	11	addresses	address	VERB
easat-10684	47	12	short	short	ADJ
easat-10684	47	13	-	-	PUNCT
easat-10684	47	14	term	term	NOUN
easat-10684	47	15	memory	memory	NOUN
easat-10684	47	16	limitations	limitation	NOUN
easat-10684	47	17	in	in	ADP
easat-10684	47	18	a	a	DET
easat-10684	47	19	recurrent	recurrent	ADJ
easat-10684	47	20	neural	neural	ADJ
easat-10684	47	21	network	network	NOUN
easat-10684	47	22	.	.	PUNCT
easat-10684	48	1	the	the	DET
easat-10684	48	2	architecture	architecture	NOUN
easat-10684	48	3	incorporates	incorporate	VERB
easat-10684	48	4	internal	internal	ADJ
easat-10684	48	5	mechanisms	mechanism	NOUN
easat-10684	48	6	known	know	VERB
easat-10684	48	7	as	as	ADP
easat-10684	48	8	gates	gate	NOUN
easat-10684	48	9	,	,	PUNCT
easat-10684	48	10	which	which	PRON
easat-10684	48	11	can	can	AUX
easat-10684	48	12	control	control	VERB
easat-10684	48	13	the	the	DET
easat-10684	48	14	flow	flow	NOUN
easat-10684	48	15	of	of	ADP
easat-10684	48	16	information	information	NOUN
easat-10684	48	17	[	[	X
easat-10684	48	18	5	5	NUM
easat-10684	48	19	]	]	PUNCT
easat-10684	48	20	.	.	PUNCT
easat-10684	49	1	lstms	lstms	PROPN
easat-10684	49	2	are	be	AUX
easat-10684	49	3	widely	widely	ADV
easat-10684	49	4	adopted	adopt	VERB
easat-10684	49	5	by	by	ADP
easat-10684	49	6	researchers	researcher	NOUN
easat-10684	49	7	to	to	PART
easat-10684	49	8	address	address	VERB
easat-10684	49	9	text	text	NOUN
easat-10684	49	10	-	-	PUNCT
easat-10684	49	11	based	base	VERB
easat-10684	49	12	classification	classification	NOUN
easat-10684	49	13	problems	problem	NOUN
easat-10684	49	14	[	[	X
easat-10684	49	15	11	11	NUM
easat-10684	49	16	]	]	PUNCT
easat-10684	49	17	and	and	CCONJ
easat-10684	49	18	bidirectional	bidirectional	ADJ
easat-10684	49	19	lstm	lstm	NOUN
easat-10684	49	20	(	(	PUNCT
easat-10684	49	21	bi	bi	NOUN
easat-10684	49	22	-	-	ADJ
easat-10684	49	23	lstm	lstm	ADJ
easat-10684	49	24	)	)	PUNCT
easat-10684	49	25	models	model	NOUN
easat-10684	49	26	have	have	AUX
easat-10684	49	27	demonstrated	demonstrate	VERB
easat-10684	49	28	high	high	ADJ
easat-10684	49	29	effectiveness	effectiveness	NOUN
easat-10684	49	30	in	in	ADP
easat-10684	49	31	gathering	gather	VERB
easat-10684	49	32	sequential	sequential	ADJ
easat-10684	49	33	information	information	NOUN
easat-10684	49	34	from	from	ADP
easat-10684	49	35	both	both	CCONJ
easat-10684	49	36	forward	forward	ADJ
easat-10684	49	37	and	and	CCONJ
easat-10684	49	38	backward	backward	ADJ
easat-10684	49	39	directions	direction	NOUN
easat-10684	49	40	of	of	ADP
easat-10684	49	41	the	the	DET
easat-10684	49	42	text	text	NOUN
easat-10684	49	43	data	datum	NOUN
easat-10684	50	1	[	[	X
easat-10684	50	2	12	12	NUM
easat-10684	50	3	]	]	PUNCT
easat-10684	50	4	.	.	PUNCT
easat-10684	51	1	1.1	1.1	NUM
easat-10684	51	2	.	.	PUNCT
easat-10684	52	1	motivation	motivation	NOUN
easat-10684	52	2	pandey	pandey	PROPN
easat-10684	52	3	et	et	PROPN
easat-10684	52	4	al	al	PROPN
easat-10684	52	5	.	.	PUNCT
easat-10684	53	1	[	[	X
easat-10684	53	2	6	6	NUM
easat-10684	53	3	]	]	PUNCT
easat-10684	53	4	revealed	reveal	VERB
easat-10684	53	5	that	that	SCONJ
easat-10684	53	6	the	the	DET
easat-10684	53	7	drastic	drastic	ADJ
easat-10684	53	8	change	change	NOUN
easat-10684	53	9	in	in	ADP
easat-10684	53	10	news	news	NOUN
easat-10684	53	11	dissemination	dissemination	NOUN
easat-10684	53	12	from	from	ADP
easat-10684	53	13	traditional	traditional	ADJ
easat-10684	53	14	newspapers	newspaper	NOUN
easat-10684	53	15	to	to	ADP
easat-10684	53	16	electronic	electronic	ADJ
easat-10684	53	17	-	-	PUNCT
easat-10684	53	18	based	base	VERB
easat-10684	53	19	news	news	NOUN
easat-10684	53	20	forecasting	forecasting	NOUN
easat-10684	53	21	has	have	AUX
easat-10684	53	22	contributed	contribute	VERB
easat-10684	53	23	to	to	ADP
easat-10684	53	24	the	the	DET
easat-10684	53	25	widespread	widespread	ADJ
easat-10684	53	26	issues	issue	NOUN
easat-10684	53	27	of	of	ADP
easat-10684	53	28	fake	fake	ADJ
easat-10684	53	29	news	news	NOUN
easat-10684	53	30	in	in	ADP
easat-10684	53	31	today	today	NOUN
easat-10684	53	32	’s	’s	PART
easat-10684	53	33	world	world	NOUN
easat-10684	53	34	.	.	PUNCT
easat-10684	54	1	this	this	PRON
easat-10684	54	2	resulted	result	VERB
easat-10684	54	3	from	from	ADP
easat-10684	54	4	a	a	DET
easat-10684	54	5	limited	limited	ADJ
easat-10684	54	6	or	or	CCONJ
easat-10684	54	7	no	no	PRON
easat-10684	54	8	supervision	supervision	NOUN
easat-10684	54	9	policy	policy	NOUN
easat-10684	54	10	scheme	scheme	NOUN
easat-10684	54	11	for	for	ADP
easat-10684	54	12	monitoring	monitor	VERB
easat-10684	54	13	news	news	NOUN
easat-10684	54	14	being	be	AUX
easat-10684	54	15	published	publish	VERB
easat-10684	54	16	online	online	ADV
easat-10684	54	17	[	[	X
easat-10684	54	18	5	5	NUM
easat-10684	54	19	]	]	PUNCT
easat-10684	54	20	.	.	PUNCT
easat-10684	55	1	many	many	ADJ
easat-10684	55	2	researchers	researcher	NOUN
easat-10684	55	3	in	in	ADP
easat-10684	55	4	the	the	DET
easat-10684	55	5	past	past	NOUN
easat-10684	55	6	have	have	AUX
easat-10684	55	7	proposed	propose	VERB
easat-10684	55	8	various	various	ADJ
easat-10684	55	9	techniques	technique	NOUN
easat-10684	55	10	to	to	PART
easat-10684	55	11	address	address	VERB
easat-10684	55	12	this	this	DET
easat-10684	55	13	issue	issue	NOUN
easat-10684	55	14	,	,	PUNCT
easat-10684	55	15	but	but	CCONJ
easat-10684	55	16	with	with	ADP
easat-10684	55	17	less	less	ADJ
easat-10684	55	18	satisfactory	satisfactory	ADJ
easat-10684	55	19	performance	performance	NOUN
easat-10684	55	20	[	[	X
easat-10684	55	21	6	6	NUM
easat-10684	55	22	,	,	PUNCT
easat-10684	55	23	13	13	NUM
easat-10684	55	24	]	]	PUNCT
easat-10684	55	25	.	.	PUNCT
easat-10684	56	1	achieving	achieve	VERB
easat-10684	56	2	optimal	optimal	ADJ
easat-10684	56	3	performance	performance	NOUN
easat-10684	56	4	in	in	ADP
easat-10684	56	5	the	the	DET
easat-10684	56	6	domain	domain	NOUN
easat-10684	56	7	of	of	ADP
easat-10684	56	8	nlp	nlp	NOUN
easat-10684	56	9	is	be	AUX
easat-10684	56	10	highly	highly	ADV
easat-10684	56	11	challenging	challenging	ADJ
easat-10684	56	12	.	.	PUNCT
easat-10684	57	1	hence	hence	ADV
easat-10684	57	2	,	,	PUNCT
easat-10684	57	3	concepts	concept	NOUN
easat-10684	57	4	such	such	ADJ
easat-10684	57	5	as	as	ADP
easat-10684	57	6	transfer	transfer	NOUN
easat-10684	57	7	learning	learning	NOUN
easat-10684	57	8	and	and	CCONJ
easat-10684	57	9	hybridization	hybridization	NOUN
easat-10684	57	10	techniques	technique	NOUN
easat-10684	57	11	are	be	AUX
easat-10684	57	12	suggested	suggest	VERB
easat-10684	57	13	by	by	ADP
easat-10684	57	14	researchers	researcher	NOUN
easat-10684	57	15	in	in	ADP
easat-10684	57	16	this	this	DET
easat-10684	57	17	field	field	NOUN
easat-10684	57	18	as	as	ADP
easat-10684	57	19	kaliyar	kaliyar	PROPN
easat-10684	57	20	et	et	PROPN
easat-10684	57	21	al	al	PROPN
easat-10684	57	22	.	.	PUNCT
easat-10684	58	1	[	[	X
easat-10684	58	2	14	14	NUM
easat-10684	58	3	]	]	PUNCT
easat-10684	58	4	,	,	PUNCT
easat-10684	58	5	jain	jain	PROPN
easat-10684	58	6	et	et	PROPN
easat-10684	58	7	al	al	PROPN
easat-10684	58	8	.	.	PUNCT
easat-10684	59	1	[	[	X
easat-10684	59	2	3	3	NUM
easat-10684	59	3	]	]	PUNCT
easat-10684	59	4	,	,	PUNCT
easat-10684	59	5	kumar	kumar	PROPN
easat-10684	59	6	et	et	PROPN
easat-10684	59	7	al	al	PROPN
easat-10684	59	8	.	.	PUNCT
easat-10684	60	1	[	[	X
easat-10684	60	2	11	11	NUM
easat-10684	60	3	]	]	PUNCT
easat-10684	60	4	,	,	PUNCT
easat-10684	60	5	and	and	CCONJ
easat-10684	60	6	jaybhaye	jaybhaye	NOUN
easat-10684	60	7	[	[	X
easat-10684	60	8	13	13	NUM
easat-10684	60	9	]	]	PUNCT
easat-10684	60	10	to	to	PART
easat-10684	60	11	improve	improve	VERB
easat-10684	60	12	the	the	DET
easat-10684	60	13	performance	performance	NOUN
easat-10684	60	14	of	of	ADP
easat-10684	60	15	a	a	DET
easat-10684	60	16	real	real	ADJ
easat-10684	60	17	-	-	PUNCT
easat-10684	60	18	time	time	NOUN
easat-10684	60	19	fake	fake	ADJ
easat-10684	60	20	news	news	NOUN
easat-10684	60	21	detection	detection	NOUN
easat-10684	60	22	model	model	NOUN
easat-10684	60	23	.	.	PUNCT
easat-10684	61	1	additionally	additionally	ADV
easat-10684	61	2	,	,	PUNCT
easat-10684	61	3	pandey	pandey	PROPN
easat-10684	61	4	et	et	PROPN
easat-10684	61	5	al	al	PROPN
easat-10684	61	6	.	.	PUNCT
easat-10684	62	1	[	[	X
easat-10684	62	2	6	6	NUM
easat-10684	62	3	]	]	PUNCT
easat-10684	62	4	and	and	CCONJ
easat-10684	62	5	jaybhaye	jaybhaye	NOUN
easat-10684	63	1	[	[	X
easat-10684	63	2	13	13	NUM
easat-10684	63	3	]	]	PUNCT
easat-10684	63	4	utilize	utilize	VERB
easat-10684	63	5	various	various	ADJ
easat-10684	63	6	machine	machine	NOUN
easat-10684	63	7	learning	learn	VERB
easat-10684	63	8	algorithms	algorithm	NOUN
easat-10684	63	9	for	for	ADP
easat-10684	63	10	building	build	VERB
easat-10684	63	11	fake	fake	ADJ
easat-10684	63	12	news	news	NOUN
easat-10684	63	13	detection	detection	NOUN
easat-10684	63	14	models	model	NOUN
easat-10684	63	15	,	,	PUNCT
easat-10684	63	16	including	include	VERB
easat-10684	63	17	support	support	NOUN
easat-10684	63	18	vector	vector	NOUN
easat-10684	63	19	machine	machine	NOUN
easat-10684	63	20	(	(	PUNCT
easat-10684	63	21	svm	svm	PROPN
easat-10684	63	22	)	)	PUNCT
easat-10684	63	23	,	,	PUNCT
easat-10684	63	24	k	k	X
easat-10684	63	25	-	-	PUNCT
easat-10684	63	26	nearest	near	ADJ
easat-10684	63	27	neighbour	neighbour	NOUN
easat-10684	63	28	(	(	PUNCT
easat-10684	63	29	knn	knn	PROPN
easat-10684	63	30	)	)	PUNCT
easat-10684	63	31	,	,	PUNCT
easat-10684	63	32	logistic	logistic	ADJ
easat-10684	63	33	regression	regression	NOUN
easat-10684	63	34	(	(	PUNCT
easat-10684	63	35	lr	lr	NOUN
easat-10684	63	36	)	)	PUNCT
easat-10684	63	37	,	,	PUNCT
easat-10684	63	38	decision	decision	NOUN
easat-10684	63	39	tree	tree	NOUN
easat-10684	63	40	(	(	PUNCT
easat-10684	63	41	dt	dt	NOUN
easat-10684	63	42	)	)	PUNCT
easat-10684	63	43	,	,	PUNCT
easat-10684	63	44	and	and	CCONJ
easat-10684	63	45	naïve	naïve	ADJ
easat-10684	63	46	bayes	bayes	NOUN
easat-10684	63	47	(	(	PUNCT
easat-10684	63	48	nb	nb	NOUN
easat-10684	63	49	)	)	PUNCT
easat-10684	63	50	algorithms	algorithm	NOUN
easat-10684	63	51	.	.	PUNCT
easat-10684	64	1	the	the	DET
easat-10684	64	2	result	result	NOUN
easat-10684	64	3	presentation	presentation	NOUN
easat-10684	64	4	of	of	ADP
easat-10684	64	5	pandey	pandey	PROPN
easat-10684	64	6	,	,	PUNCT
easat-10684	64	7	et	et	PROPN
easat-10684	64	8	al	al	PROPN
easat-10684	64	9	.	.	PUNCT
easat-10684	65	1	[	[	X
easat-10684	65	2	6	6	X
easat-10684	65	3	]	]	PUNCT
easat-10684	65	4	shows	show	VERB
easat-10684	65	5	that	that	SCONJ
easat-10684	65	6	the	the	DET
easat-10684	65	7	highest	high	ADJ
easat-10684	65	8	accuracy	accuracy	NOUN
easat-10684	65	9	attained	attain	VERB
easat-10684	65	10	in	in	ADP
easat-10684	65	11	their	their	PRON
easat-10684	65	12	study	study	NOUN
easat-10684	65	13	is	be	AUX
easat-10684	65	14	90.04	90.04	NUM
easat-10684	65	15	%	%	NOUN
easat-10684	65	16	,	,	PUNCT
easat-10684	65	17	which	which	PRON
easat-10684	65	18	creates	create	VERB
easat-10684	65	19	room	room	NOUN
easat-10684	65	20	for	for	ADP
easat-10684	65	21	further	further	ADJ
easat-10684	65	22	improvement	improvement	NOUN
easat-10684	65	23	,	,	PUNCT
easat-10684	65	24	and	and	CCONJ
easat-10684	65	25	jaybhaye	jaybhaye	NOUN
easat-10684	65	26	[	[	X
easat-10684	65	27	13	13	NUM
easat-10684	65	28	]	]	PUNCT
easat-10684	65	29	suffers	suffer	VERB
easat-10684	65	30	from	from	ADP
easat-10684	65	31	overfitting	overfitte	VERB
easat-10684	65	32	during	during	ADP
easat-10684	65	33	training	training	NOUN
easat-10684	65	34	with	with	ADP
easat-10684	65	35	low	low	ADJ
easat-10684	65	36	performance	performance	NOUN
easat-10684	65	37	on	on	ADP
easat-10684	65	38	testing	test	VERB
easat-10684	65	39	data	datum	NOUN
easat-10684	65	40	.	.	PUNCT
easat-10684	66	1	hence	hence	ADV
easat-10684	66	2	,	,	PUNCT
easat-10684	66	3	this	this	DET
easat-10684	66	4	study	study	NOUN
easat-10684	66	5	proposes	propose	VERB
easat-10684	66	6	an	an	DET
easat-10684	66	7	enhanced	enhanced	ADJ
easat-10684	66	8	machine	machine	NOUN
easat-10684	66	9	learning	learning	NOUN
easat-10684	66	10	-	-	PUNCT
easat-10684	66	11	based	base	VERB
easat-10684	66	12	algorithm	algorithm	NOUN
easat-10684	66	13	using	use	VERB
easat-10684	66	14	transfer	transfer	NOUN
easat-10684	66	15	learning	learning	NOUN
easat-10684	66	16	and	and	CCONJ
easat-10684	66	17	hybridization	hybridization	NOUN
easat-10684	66	18	techniques	technique	NOUN
easat-10684	66	19	.	.	PUNCT
easat-10684	67	1	1517	1517	NUM
easat-10684	67	2	edelweiss	edelweiss	PROPN
easat-10684	67	3	applied	apply	VERB
easat-10684	67	4	science	science	NOUN
easat-10684	67	5	and	and	CCONJ
easat-10684	67	6	technology	technology	NOUN
easat-10684	67	7	issn	issn	PROPN
easat-10684	67	8	:	:	PUNCT
easat-10684	67	9	2576	2576	NUM
easat-10684	67	10	-	-	SYM
easat-10684	67	11	8484	8484	NUM
easat-10684	67	12	vol	vol	NOUN
easat-10684	67	13	.	.	PROPN
easat-10684	68	1	9	9	NUM
easat-10684	68	2	,	,	PUNCT
easat-10684	68	3	no	no	INTJ
easat-10684	68	4	.	.	NOUN
easat-10684	68	5	10	10	NUM
easat-10684	68	6	:	:	PUNCT
easat-10684	68	7	1515	1515	NUM
easat-10684	68	8	-	-	SYM
easat-10684	68	9	1540	1540	NUM
easat-10684	68	10	,	,	PUNCT
easat-10684	68	11	2025	2025	NUM
easat-10684	68	12	doi	doi	NOUN
easat-10684	68	13	:	:	PUNCT
easat-10684	68	14	10.55214/2576	10.55214/2576	NUM
easat-10684	68	15	-	-	SYM
easat-10684	68	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	69	1	©	©	PROPN
easat-10684	69	2	2025	2025	NUM
easat-10684	69	3	by	by	ADP
easat-10684	69	4	the	the	DET
easat-10684	69	5	authors	author	NOUN
easat-10684	69	6	;	;	PUNCT
easat-10684	69	7	licensee	licensee	PROPN
easat-10684	69	8	learning	learning	NOUN
easat-10684	69	9	gate	gate	VERB
easat-10684	69	10	the	the	DET
easat-10684	69	11	primary	primary	ADJ
easat-10684	69	12	aim	aim	NOUN
easat-10684	69	13	is	be	AUX
easat-10684	69	14	to	to	PART
easat-10684	69	15	enhance	enhance	VERB
easat-10684	69	16	the	the	DET
easat-10684	69	17	performance	performance	NOUN
easat-10684	69	18	of	of	ADP
easat-10684	69	19	a	a	DET
easat-10684	69	20	machine	machine	NOUN
easat-10684	69	21	learning	learning	NOUN
easat-10684	69	22	-	-	PUNCT
easat-10684	69	23	based	base	VERB
easat-10684	69	24	fake	fake	ADJ
easat-10684	69	25	news	news	NOUN
easat-10684	69	26	detection	detection	NOUN
easat-10684	69	27	model	model	NOUN
easat-10684	69	28	using	use	VERB
easat-10684	69	29	the	the	DET
easat-10684	69	30	transfer	transfer	NOUN
easat-10684	69	31	knowledge	knowledge	NOUN
easat-10684	69	32	approach	approach	NOUN
easat-10684	69	33	from	from	ADP
easat-10684	69	34	bi	bi	ADJ
easat-10684	69	35	-	-	ADJ
easat-10684	69	36	directional	directional	ADJ
easat-10684	69	37	long	long	ADJ
easat-10684	69	38	short	short	ADJ
easat-10684	69	39	-	-	PUNCT
easat-10684	69	40	term	term	NOUN
easat-10684	69	41	memory	memory	NOUN
easat-10684	69	42	.	.	PUNCT
easat-10684	70	1	furthermore	furthermore	ADV
easat-10684	70	2	,	,	PUNCT
easat-10684	70	3	an	an	DET
easat-10684	70	4	extensive	extensive	ADJ
easat-10684	70	5	experimental	experimental	ADJ
easat-10684	70	6	study	study	NOUN
easat-10684	70	7	will	will	AUX
easat-10684	70	8	be	be	AUX
easat-10684	70	9	conducted	conduct	VERB
easat-10684	70	10	on	on	ADP
easat-10684	70	11	the	the	DET
easat-10684	70	12	effect	effect	NOUN
easat-10684	70	13	of	of	ADP
easat-10684	70	14	pre	pre	ADJ
easat-10684	70	15	-	-	ADJ
easat-10684	70	16	trained	train	VERB
easat-10684	70	17	bi	bi	NOUN
easat-10684	70	18	-	-	ADJ
easat-10684	70	19	lstm	lstm	ADJ
easat-10684	70	20	weights	weight	NOUN
easat-10684	70	21	on	on	ADP
easat-10684	70	22	the	the	DET
easat-10684	70	23	performance	performance	NOUN
easat-10684	70	24	of	of	ADP
easat-10684	70	25	various	various	ADJ
easat-10684	70	26	machine	machine	NOUN
easat-10684	70	27	learning	learn	VERB
easat-10684	70	28	algorithms	algorithm	NOUN
easat-10684	70	29	(	(	PUNCT
easat-10684	70	30	lr	lr	INTJ
easat-10684	70	31	,	,	PUNCT
easat-10684	70	32	svm	svm	ADJ
easat-10684	70	33	,	,	PUNCT
easat-10684	70	34	dt	dt	PROPN
easat-10684	70	35	,	,	PUNCT
easat-10684	70	36	nb	nb	PROPN
easat-10684	70	37	,	,	PUNCT
easat-10684	70	38	and	and	CCONJ
easat-10684	70	39	dt	dt	PROPN
easat-10684	70	40	)	)	PUNCT
easat-10684	70	41	.	.	PUNCT
easat-10684	71	1	finally	finally	ADV
easat-10684	71	2	,	,	PUNCT
easat-10684	71	3	the	the	DET
easat-10684	71	4	proposed	propose	VERB
easat-10684	71	5	bi	bi	NOUN
easat-10684	71	6	-	-	PUNCT
easat-10684	71	7	lstm-2	lstm-2	NOUN
easat-10684	71	8	-	-	PUNCT
easat-10684	71	9	ml	ml	NOUN
easat-10684	71	10	model	model	NOUN
easat-10684	71	11	will	will	AUX
easat-10684	71	12	be	be	AUX
easat-10684	71	13	evaluated	evaluate	VERB
easat-10684	71	14	against	against	ADP
easat-10684	71	15	existing	exist	VERB
easat-10684	71	16	machine	machine	NOUN
easat-10684	71	17	learning	learning	NOUN
easat-10684	71	18	models	model	NOUN
easat-10684	71	19	.	.	PUNCT
easat-10684	72	1	the	the	DET
easat-10684	72	2	research	research	NOUN
easat-10684	72	3	work	work	NOUN
easat-10684	72	4	intends	intend	VERB
easat-10684	72	5	to	to	PART
easat-10684	72	6	solidify	solidify	VERB
easat-10684	72	7	the	the	DET
easat-10684	72	8	hypothetical	hypothetical	ADJ
easat-10684	72	9	claim	claim	NOUN
easat-10684	72	10	and	and	CCONJ
easat-10684	72	11	suggestion	suggestion	NOUN
easat-10684	72	12	made	make	VERB
easat-10684	72	13	by	by	ADP
easat-10684	72	14	previous	previous	ADJ
easat-10684	72	15	researchers	researcher	NOUN
easat-10684	72	16	on	on	ADP
easat-10684	72	17	the	the	DET
easat-10684	72	18	possibilities	possibility	NOUN
easat-10684	72	19	of	of	ADP
easat-10684	72	20	improving	improve	VERB
easat-10684	72	21	the	the	DET
easat-10684	72	22	performance	performance	NOUN
easat-10684	72	23	of	of	ADP
easat-10684	72	24	fake	fake	ADJ
easat-10684	72	25	news	news	NOUN
easat-10684	72	26	detection	detection	NOUN
easat-10684	72	27	models	model	NOUN
easat-10684	72	28	,	,	PUNCT
easat-10684	72	29	along	along	ADP
easat-10684	72	30	with	with	ADP
easat-10684	72	31	their	their	PRON
easat-10684	72	32	accuracy	accuracy	NOUN
easat-10684	72	33	,	,	PUNCT
easat-10684	72	34	using	use	VERB
easat-10684	72	35	transfer	transfer	NOUN
easat-10684	72	36	learning	learning	NOUN
easat-10684	72	37	techniques	technique	NOUN
easat-10684	72	38	.	.	PUNCT
easat-10684	73	1	hence	hence	ADV
easat-10684	73	2	,	,	PUNCT
easat-10684	73	3	proper	proper	ADJ
easat-10684	73	4	study	study	NOUN
easat-10684	73	5	and	and	CCONJ
easat-10684	73	6	comparative	comparative	ADJ
easat-10684	73	7	experiments	experiment	NOUN
easat-10684	73	8	will	will	AUX
easat-10684	73	9	be	be	AUX
easat-10684	73	10	conducted	conduct	VERB
easat-10684	73	11	between	between	ADP
easat-10684	73	12	benchmark	benchmark	NOUN
easat-10684	73	13	studies	study	NOUN
easat-10684	73	14	,	,	PUNCT
easat-10684	73	15	pandey	pandey	PROPN
easat-10684	73	16	et	et	PROPN
easat-10684	73	17	al	al	PROPN
easat-10684	73	18	.	.	PUNCT
easat-10684	74	1	[	[	X
easat-10684	74	2	6	6	NUM
easat-10684	74	3	]	]	PUNCT
easat-10684	74	4	and	and	CCONJ
easat-10684	74	5	jaybhaye	jaybhaye	NOUN
easat-10684	74	6	[	[	X
easat-10684	74	7	13	13	NUM
easat-10684	74	8	]	]	PUNCT
easat-10684	74	9	the	the	DET
easat-10684	74	10	proposed	propose	VERB
easat-10684	74	11	bilsmt-2	bilsmt-2	NUM
easat-10684	74	12	-	-	PUNCT
easat-10684	74	13	ml	ml	NOUN
easat-10684	74	14	transfer	transfer	NOUN
easat-10684	74	15	learning	learn	VERB
easat-10684	74	16	fake	fake	ADJ
easat-10684	74	17	news	news	NOUN
easat-10684	74	18	detection	detection	NOUN
easat-10684	74	19	model	model	NOUN
easat-10684	74	20	for	for	ADP
easat-10684	74	21	proper	proper	ADJ
easat-10684	74	22	verification	verification	NOUN
easat-10684	74	23	.	.	PUNCT
easat-10684	75	1	the	the	DET
easat-10684	75	2	following	follow	VERB
easat-10684	75	3	are	be	AUX
easat-10684	75	4	the	the	DET
easat-10684	75	5	key	key	ADJ
easat-10684	75	6	contributions	contribution	NOUN
easat-10684	75	7	of	of	ADP
easat-10684	75	8	the	the	DET
easat-10684	75	9	proposed	propose	VERB
easat-10684	75	10	model	model	NOUN
easat-10684	75	11	:	:	PUNCT
easat-10684	75	12	•	•	ADP
easat-10684	75	13	the	the	DET
easat-10684	75	14	study	study	NOUN
easat-10684	75	15	enhances	enhance	VERB
easat-10684	75	16	the	the	DET
easat-10684	75	17	performance	performance	NOUN
easat-10684	75	18	of	of	ADP
easat-10684	75	19	an	an	DET
easat-10684	75	20	ml	ml	ADV
easat-10684	75	21	-	-	PUNCT
easat-10684	75	22	based	base	VERB
easat-10684	75	23	fake	fake	ADJ
easat-10684	75	24	news	news	NOUN
easat-10684	75	25	detection	detection	NOUN
easat-10684	75	26	model	model	NOUN
easat-10684	75	27	using	use	VERB
easat-10684	75	28	the	the	DET
easat-10684	75	29	transfer	transfer	NOUN
easat-10684	75	30	knowledge	knowledge	NOUN
easat-10684	75	31	approach	approach	NOUN
easat-10684	75	32	called	call	VERB
easat-10684	75	33	bi	bi	NOUN
easat-10684	75	34	-	-	NOUN
easat-10684	75	35	lstm	lstm	NOUN
easat-10684	75	36	.	.	PUNCT
easat-10684	76	1	•	•	NUM
easat-10684	76	2	an	an	DET
easat-10684	76	3	extensive	extensive	ADJ
easat-10684	76	4	experimental	experimental	ADJ
easat-10684	76	5	study	study	NOUN
easat-10684	76	6	will	will	AUX
easat-10684	76	7	analyze	analyze	VERB
easat-10684	76	8	the	the	DET
easat-10684	76	9	impact	impact	NOUN
easat-10684	76	10	of	of	ADP
easat-10684	76	11	pre	pre	ADJ
easat-10684	76	12	-	-	ADJ
easat-10684	76	13	trained	train	VERB
easat-10684	76	14	bi	bi	NOUN
easat-10684	76	15	-	-	ADJ
easat-10684	76	16	lstm	lstm	ADJ
easat-10684	76	17	weights	weight	NOUN
easat-10684	76	18	on	on	ADP
easat-10684	76	19	the	the	DET
easat-10684	76	20	performance	performance	NOUN
easat-10684	76	21	of	of	ADP
easat-10684	76	22	different	different	ADJ
easat-10684	76	23	machine	machine	NOUN
easat-10684	76	24	learning	learn	VERB
easat-10684	76	25	algorithms	algorithm	NOUN
easat-10684	76	26	,	,	PUNCT
easat-10684	76	27	including	include	VERB
easat-10684	76	28	lr	lr	NOUN
easat-10684	76	29	,	,	PUNCT
easat-10684	76	30	svm	svm	ADJ
easat-10684	76	31	,	,	PUNCT
easat-10684	76	32	dt	dt	X
easat-10684	76	33	,	,	PUNCT
easat-10684	76	34	and	and	CCONJ
easat-10684	76	35	nb	nb	INTJ
easat-10684	76	36	.	.	PROPN
easat-10684	76	37	•	•	NUM
easat-10684	76	38	the	the	DET
easat-10684	76	39	bi	bi	NOUN
easat-10684	76	40	-	-	PUNCT
easat-10684	76	41	lstm-2	lstm-2	NOUN
easat-10684	76	42	-	-	PUNCT
easat-10684	76	43	ml	ml	NOUN
easat-10684	76	44	model	model	NOUN
easat-10684	76	45	was	be	AUX
easat-10684	76	46	evaluated	evaluate	VERB
easat-10684	76	47	and	and	CCONJ
easat-10684	76	48	compared	compare	VERB
easat-10684	76	49	with	with	ADP
easat-10684	76	50	existing	exist	VERB
easat-10684	76	51	ml	ml	NOUN
easat-10684	76	52	-	-	PUNCT
easat-10684	76	53	based	base	VERB
easat-10684	76	54	models	model	NOUN
easat-10684	76	55	to	to	PART
easat-10684	76	56	establish	establish	VERB
easat-10684	76	57	its	its	PRON
easat-10684	76	58	effectiveness	effectiveness	NOUN
easat-10684	76	59	.	.	PUNCT
easat-10684	77	1	•	•	NUM
easat-10684	77	2	a	a	DET
easat-10684	77	3	comprehensive	comprehensive	ADJ
easat-10684	77	4	comparative	comparative	ADJ
easat-10684	77	5	study	study	NOUN
easat-10684	77	6	was	be	AUX
easat-10684	77	7	conducted	conduct	VERB
easat-10684	77	8	with	with	ADP
easat-10684	77	9	the	the	DET
easat-10684	77	10	benchmark	benchmark	ADJ
easat-10684	77	11	studies	study	NOUN
easat-10684	77	12	to	to	PART
easat-10684	77	13	ensure	ensure	VERB
easat-10684	77	14	proper	proper	ADJ
easat-10684	77	15	evaluation	evaluation	NOUN
easat-10684	77	16	of	of	ADP
easat-10684	77	17	the	the	DET
easat-10684	77	18	proposed	propose	VERB
easat-10684	77	19	model	model	NOUN
easat-10684	77	20	.	.	PUNCT
easat-10684	78	1	the	the	DET
easat-10684	78	2	remainder	remainder	NOUN
easat-10684	78	3	of	of	ADP
easat-10684	78	4	the	the	DET
easat-10684	78	5	study	study	NOUN
easat-10684	78	6	is	be	AUX
easat-10684	78	7	as	as	SCONJ
easat-10684	78	8	follows	follow	VERB
easat-10684	78	9	:	:	PUNCT
easat-10684	78	10	section	section	NOUN
easat-10684	78	11	2	2	NUM
easat-10684	78	12	discusses	discuss	VERB
easat-10684	78	13	related	related	ADJ
easat-10684	78	14	work	work	NOUN
easat-10684	78	15	,	,	PUNCT
easat-10684	78	16	section	section	NOUN
easat-10684	78	17	3	3	NUM
easat-10684	78	18	discusses	discuss	VERB
easat-10684	78	19	the	the	DET
easat-10684	78	20	methodology	methodology	NOUN
easat-10684	78	21	used	use	VERB
easat-10684	78	22	in	in	ADP
easat-10684	78	23	this	this	DET
easat-10684	78	24	study	study	NOUN
easat-10684	78	25	,	,	PUNCT
easat-10684	78	26	section	section	NOUN
easat-10684	78	27	4	4	NUM
easat-10684	78	28	discusses	discuss	VERB
easat-10684	78	29	the	the	DET
easat-10684	78	30	results	result	NOUN
easat-10684	78	31	and	and	CCONJ
easat-10684	78	32	discussion	discussion	NOUN
easat-10684	78	33	,	,	PUNCT
easat-10684	78	34	and	and	CCONJ
easat-10684	78	35	section	section	NOUN
easat-10684	78	36	5	5	NUM
easat-10684	78	37	concludes	conclude	VERB
easat-10684	78	38	the	the	DET
easat-10684	78	39	study	study	NOUN
easat-10684	78	40	.	.	PUNCT
easat-10684	79	1	2	2	X
easat-10684	79	2	.	.	X
easat-10684	79	3	related	relate	VERB
easat-10684	79	4	work	work	NOUN
easat-10684	79	5	the	the	DET
easat-10684	79	6	rapid	rapid	ADJ
easat-10684	79	7	proliferation	proliferation	NOUN
easat-10684	79	8	of	of	ADP
easat-10684	79	9	misinformation	misinformation	NOUN
easat-10684	79	10	on	on	ADP
easat-10684	79	11	digital	digital	ADJ
easat-10684	79	12	platforms	platform	NOUN
easat-10684	79	13	has	have	AUX
easat-10684	79	14	raised	raise	VERB
easat-10684	79	15	significant	significant	ADJ
easat-10684	79	16	concerns	concern	NOUN
easat-10684	79	17	about	about	ADP
easat-10684	79	18	its	its	PRON
easat-10684	79	19	impact	impact	NOUN
easat-10684	79	20	on	on	ADP
easat-10684	79	21	society	society	NOUN
easat-10684	79	22	,	,	PUNCT
easat-10684	79	23	necessitating	necessitate	VERB
easat-10684	79	24	effective	effective	ADJ
easat-10684	79	25	methods	method	NOUN
easat-10684	79	26	for	for	ADP
easat-10684	79	27	identifying	identify	VERB
easat-10684	79	28	and	and	CCONJ
easat-10684	79	29	mitigating	mitigate	VERB
easat-10684	79	30	fake	fake	ADJ
easat-10684	79	31	news	news	NOUN
easat-10684	79	32	[	[	X
easat-10684	79	33	15	15	NUM
easat-10684	79	34	,	,	PUNCT
easat-10684	79	35	16	16	NUM
easat-10684	79	36	]	]	PUNCT
easat-10684	79	37	.	.	PUNCT
easat-10684	80	1	recent	recent	ADJ
easat-10684	80	2	studies	study	NOUN
easat-10684	80	3	indicate	indicate	VERB
easat-10684	80	4	that	that	SCONJ
easat-10684	80	5	ml	ml	NOUN
easat-10684	80	6	and	and	CCONJ
easat-10684	80	7	dl	dl	PROPN
easat-10684	80	8	models	model	NOUN
easat-10684	80	9	significantly	significantly	ADV
easat-10684	80	10	outperform	outperform	VERB
easat-10684	80	11	traditional	traditional	ADJ
easat-10684	80	12	methods	method	NOUN
easat-10684	80	13	in	in	ADP
easat-10684	80	14	detecting	detect	VERB
easat-10684	80	15	fake	fake	ADJ
easat-10684	80	16	news	news	NOUN
easat-10684	80	17	,	,	PUNCT
easat-10684	80	18	showcasing	showcase	VERB
easat-10684	80	19	their	their	PRON
easat-10684	80	20	robustness	robustness	NOUN
easat-10684	80	21	across	across	ADP
easat-10684	80	22	various	various	ADJ
easat-10684	80	23	datasets	dataset	NOUN
easat-10684	80	24	.	.	PUNCT
easat-10684	81	1	these	these	DET
easat-10684	81	2	advanced	advanced	ADJ
easat-10684	81	3	models	model	NOUN
easat-10684	81	4	leverage	leverage	VERB
easat-10684	81	5	sophisticated	sophisticated	ADJ
easat-10684	81	6	algorithms	algorithm	NOUN
easat-10684	81	7	and	and	CCONJ
easat-10684	81	8	embeddings	embedding	NOUN
easat-10684	81	9	to	to	PART
easat-10684	81	10	enhance	enhance	VERB
easat-10684	81	11	classification	classification	NOUN
easat-10684	81	12	accuracy	accuracy	NOUN
easat-10684	81	13	,	,	PUNCT
easat-10684	81	14	making	make	VERB
easat-10684	81	15	them	they	PRON
easat-10684	81	16	essential	essential	ADJ
easat-10684	81	17	tools	tool	NOUN
easat-10684	81	18	in	in	ADP
easat-10684	81	19	combating	combat	VERB
easat-10684	81	20	misinformation	misinformation	NOUN
easat-10684	81	21	.	.	PUNCT
easat-10684	82	1	models	model	NOUN
easat-10684	82	2	like	like	ADP
easat-10684	82	3	cnn	cnn	PROPN
easat-10684	82	4	,	,	PUNCT
easat-10684	82	5	bert	bert	PROPN
easat-10684	82	6	,	,	PUNCT
easat-10684	82	7	and	and	CCONJ
easat-10684	82	8	gpt	gpt	NOUN
easat-10684	82	9	have	have	AUX
easat-10684	82	10	shown	show	VERB
easat-10684	82	11	exceptional	exceptional	ADJ
easat-10684	82	12	capabilities	capability	NOUN
easat-10684	82	13	in	in	ADP
easat-10684	82	14	identifying	identify	VERB
easat-10684	82	15	fake	fake	ADJ
easat-10684	82	16	news	news	NOUN
easat-10684	82	17	by	by	ADP
easat-10684	82	18	understanding	understand	VERB
easat-10684	82	19	contextual	contextual	ADJ
easat-10684	82	20	language	language	NOUN
easat-10684	82	21	and	and	CCONJ
easat-10684	82	22	detecting	detect	VERB
easat-10684	82	23	patterns	pattern	NOUN
easat-10684	82	24	in	in	ADP
easat-10684	82	25	text	text	NOUN
easat-10684	82	26	[	[	X
easat-10684	82	27	17	17	NUM
easat-10684	82	28	]	]	PUNCT
easat-10684	82	29	.	.	PUNCT
easat-10684	83	1	combining	combine	VERB
easat-10684	83	2	dl	dl	PROPN
easat-10684	83	3	with	with	ADP
easat-10684	83	4	traditional	traditional	ADJ
easat-10684	83	5	ml	ml	NOUN
easat-10684	83	6	methods	method	NOUN
easat-10684	83	7	has	have	AUX
easat-10684	83	8	led	lead	VERB
easat-10684	83	9	to	to	ADP
easat-10684	83	10	improved	improve	VERB
easat-10684	83	11	performance	performance	NOUN
easat-10684	83	12	metrics	metric	NOUN
easat-10684	83	13	,	,	PUNCT
easat-10684	83	14	including	include	VERB
easat-10684	83	15	accuracy	accuracy	NOUN
easat-10684	83	16	and	and	CCONJ
easat-10684	83	17	f1	f1	NOUN
easat-10684	83	18	scores	score	NOUN
easat-10684	83	19	[	[	X
easat-10684	83	20	18	18	NUM
easat-10684	83	21	]	]	PUNCT
easat-10684	83	22	.	.	PUNCT
easat-10684	84	1	techniques	technique	NOUN
easat-10684	84	2	such	such	ADJ
easat-10684	84	3	as	as	ADP
easat-10684	84	4	bi	bi	NOUN
easat-10684	84	5	-	-	NOUN
easat-10684	84	6	gru	gru	NOUN
easat-10684	84	7	and	and	CCONJ
easat-10684	84	8	bi	bi	NOUN
easat-10684	84	9	-	-	ADJ
easat-10684	84	10	lstm	lstm	ADJ
easat-10684	84	11	have	have	AUX
easat-10684	84	12	demonstrated	demonstrate	VERB
easat-10684	84	13	superior	superior	ADJ
easat-10684	84	14	results	result	NOUN
easat-10684	84	15	in	in	ADP
easat-10684	84	16	arabic	arabic	ADJ
easat-10684	84	17	fake	fake	ADJ
easat-10684	84	18	news	news	NOUN
easat-10684	84	19	detection	detection	NOUN
easat-10684	84	20	,	,	PUNCT
easat-10684	84	21	achieving	achieve	VERB
easat-10684	84	22	f1	f1	NOUN
easat-10684	84	23	scores	score	NOUN
easat-10684	84	24	above	above	ADP
easat-10684	84	25	0.98	0.98	NUM
easat-10684	84	26	[	[	X
easat-10684	84	27	19	19	NUM
easat-10684	84	28	]	]	PUNCT
easat-10684	84	29	.	.	PUNCT
easat-10684	85	1	research	research	NOUN
easat-10684	85	2	has	have	AUX
easat-10684	85	3	explored	explore	VERB
easat-10684	85	4	multimodal	multimodal	NOUN
easat-10684	85	5	approaches	approach	NOUN
easat-10684	85	6	that	that	PRON
easat-10684	85	7	analyze	analyze	VERB
easat-10684	85	8	text	text	NOUN
easat-10684	85	9	,	,	PUNCT
easat-10684	85	10	images	image	NOUN
easat-10684	85	11	,	,	PUNCT
easat-10684	85	12	and	and	CCONJ
easat-10684	85	13	videos	video	NOUN
easat-10684	85	14	,	,	PUNCT
easat-10684	85	15	enhancing	enhance	VERB
easat-10684	85	16	detection	detection	NOUN
easat-10684	85	17	capabilities	capability	NOUN
easat-10684	85	18	beyond	beyond	ADP
easat-10684	85	19	text	text	NOUN
easat-10684	85	20	alone	alone	ADV
easat-10684	85	21	[	[	X
easat-10684	85	22	20	20	NUM
easat-10684	85	23	]	]	PUNCT
easat-10684	85	24	.	.	PUNCT
easat-10684	86	1	the	the	DET
easat-10684	86	2	use	use	NOUN
easat-10684	86	3	of	of	ADP
easat-10684	86	4	embeddings	embedding	NOUN
easat-10684	86	5	such	such	ADJ
easat-10684	86	6	as	as	ADP
easat-10684	86	7	tf	tf	PROPN
easat-10684	86	8	-	-	PUNCT
easat-10684	86	9	idf	idf	PROPN
easat-10684	86	10	and	and	CCONJ
easat-10684	86	11	fasttext	fasttext	PROPN
easat-10684	86	12	has	have	AUX
easat-10684	86	13	been	be	AUX
easat-10684	86	14	crucial	crucial	ADJ
easat-10684	86	15	in	in	ADP
easat-10684	86	16	improving	improve	VERB
easat-10684	86	17	model	model	NOUN
easat-10684	86	18	performance	performance	NOUN
easat-10684	86	19	,	,	PUNCT
easat-10684	86	20	with	with	ADP
easat-10684	86	21	tf	tf	PROPN
easat-10684	86	22	-	-	PUNCT
easat-10684	86	23	idf	idf	PROPN
easat-10684	86	24	yielding	yield	VERB
easat-10684	86	25	the	the	DET
easat-10684	86	26	best	good	ADJ
easat-10684	86	27	results	result	NOUN
easat-10684	86	28	for	for	ADP
easat-10684	86	29	certain	certain	ADJ
easat-10684	86	30	classifiers	classifier	NOUN
easat-10684	86	31	[	[	X
easat-10684	86	32	21	21	NUM
easat-10684	86	33	]	]	PUNCT
easat-10684	86	34	.	.	PUNCT
easat-10684	87	1	while	while	SCONJ
easat-10684	87	2	these	these	DET
easat-10684	87	3	advanced	advanced	ADJ
easat-10684	87	4	models	model	NOUN
easat-10684	87	5	show	show	VERB
easat-10684	87	6	promise	promise	NOUN
easat-10684	87	7	,	,	PUNCT
easat-10684	87	8	challenges	challenge	NOUN
easat-10684	87	9	remain	remain	VERB
easat-10684	87	10	in	in	ADP
easat-10684	87	11	ensuring	ensure	VERB
easat-10684	87	12	their	their	PRON
easat-10684	87	13	adaptability	adaptability	NOUN
easat-10684	87	14	across	across	ADP
easat-10684	87	15	different	different	ADJ
easat-10684	87	16	languages	language	NOUN
easat-10684	87	17	and	and	CCONJ
easat-10684	87	18	media	medium	NOUN
easat-10684	87	19	formats	format	NOUN
easat-10684	87	20	,	,	PUNCT
easat-10684	87	21	highlighting	highlight	VERB
easat-10684	87	22	the	the	DET
easat-10684	87	23	need	need	NOUN
easat-10684	87	24	for	for	ADP
easat-10684	87	25	ongoing	ongoing	ADJ
easat-10684	87	26	research	research	NOUN
easat-10684	87	27	in	in	ADP
easat-10684	87	28	this	this	DET
easat-10684	87	29	area	area	NOUN
easat-10684	87	30	.	.	PUNCT
easat-10684	88	1	apostol	apostol	NOUN
easat-10684	88	2	and	and	CCONJ
easat-10684	88	3	truică	truică	NOUN
easat-10684	89	1	[	[	X
easat-10684	89	2	22	22	NUM
easat-10684	89	3	]	]	PUNCT
easat-10684	89	4	discovered	discover	VERB
easat-10684	89	5	that	that	SCONJ
easat-10684	89	6	the	the	DET
easat-10684	89	7	transformation	transformation	NOUN
easat-10684	89	8	and	and	CCONJ
easat-10684	89	9	digitization	digitization	NOUN
easat-10684	89	10	of	of	ADP
easat-10684	89	11	social	social	ADJ
easat-10684	89	12	media	medium	NOUN
easat-10684	89	13	come	come	VERB
easat-10684	89	14	with	with	ADP
easat-10684	89	15	many	many	ADJ
easat-10684	89	16	advantages	advantage	NOUN
easat-10684	89	17	but	but	CCONJ
easat-10684	89	18	also	also	ADV
easat-10684	89	19	open	open	VERB
easat-10684	89	20	the	the	DET
easat-10684	89	21	doors	door	NOUN
easat-10684	89	22	to	to	ADP
easat-10684	89	23	widespread	widespread	ADJ
easat-10684	89	24	disinformation	disinformation	NOUN
easat-10684	89	25	and	and	CCONJ
easat-10684	89	26	malformation	malformation	NOUN
easat-10684	89	27	via	via	ADP
easat-10684	89	28	fake	fake	ADJ
easat-10684	89	29	news	news	NOUN
easat-10684	89	30	pandey	pandey	PROPN
easat-10684	89	31	et	et	PROPN
easat-10684	89	32	al	al	PROPN
easat-10684	89	33	.	.	PUNCT
easat-10684	90	1	[	[	X
easat-10684	90	2	6	6	NUM
easat-10684	90	3	]	]	PUNCT
easat-10684	90	4	and	and	CCONJ
easat-10684	90	5	kaliyar	kaliyar	PROPN
easat-10684	90	6	et	et	PROPN
easat-10684	90	7	al	al	PROPN
easat-10684	90	8	.	.	PUNCT
easat-10684	91	1	[	[	X
easat-10684	91	2	14	14	NUM
easat-10684	91	3	]	]	PUNCT
easat-10684	91	4	.	.	PUNCT
easat-10684	92	1	apostol	apostol	PROPN
easat-10684	92	2	and	and	CCONJ
easat-10684	92	3	truică	truică	NOUN
easat-10684	92	4	[	[	X
easat-10684	92	5	22	22	NUM
easat-10684	92	6	]	]	PUNCT
easat-10684	92	7	used	use	VERB
easat-10684	92	8	a	a	DET
easat-10684	92	9	document	document	NOUN
easat-10684	92	10	embedding	embed	VERB
easat-10684	92	11	approach	approach	NOUN
easat-10684	92	12	to	to	PART
easat-10684	92	13	build	build	VERB
easat-10684	92	14	multiple	multiple	ADJ
easat-10684	92	15	models	model	NOUN
easat-10684	92	16	and	and	CCONJ
easat-10684	92	17	present	present	VERB
easat-10684	92	18	an	an	DET
easat-10684	92	19	architecture	architecture	NOUN
easat-10684	92	20	based	base	VERB
easat-10684	92	21	on	on	ADP
easat-10684	92	22	multi	multi	ADJ
easat-10684	92	23	-	-	ADJ
easat-10684	92	24	labelled	label	VERB
easat-10684	92	25	classification	classification	NOUN
easat-10684	92	26	.	.	PUNCT
easat-10684	93	1	the	the	DET
easat-10684	93	2	results	result	NOUN
easat-10684	93	3	obtained	obtain	VERB
easat-10684	93	4	show	show	NOUN
easat-10684	93	5	that	that	SCONJ
easat-10684	93	6	the	the	DET
easat-10684	93	7	approach	approach	NOUN
easat-10684	93	8	performs	perform	VERB
easat-10684	93	9	better	well	ADJ
easat-10684	93	10	than	than	ADP
easat-10684	93	11	the	the	DET
easat-10684	93	12	complex	complex	ADJ
easat-10684	93	13	state	state	NOUN
easat-10684	93	14	-	-	PUNCT
easat-10684	93	15	of	of	ADP
easat-10684	93	16	-	-	PUNCT
easat-10684	93	17	the	the	DET
easat-10684	93	18	-	-	PUNCT
easat-10684	93	19	art	art	NOUN
easat-10684	93	20	deep	deep	ADJ
easat-10684	93	21	neural	neural	ADJ
easat-10684	93	22	network	network	NOUN
easat-10684	93	23	model	model	NOUN
easat-10684	93	24	.	.	PUNCT
easat-10684	94	1	hossain	hossain	PROPN
easat-10684	94	2	et	et	PROPN
easat-10684	94	3	al	al	PROPN
easat-10684	94	4	.	.	PUNCT
easat-10684	95	1	[	[	X
easat-10684	95	2	5	5	NUM
easat-10684	95	3	]	]	PUNCT
easat-10684	95	4	developed	develop	VERB
easat-10684	95	5	a	a	DET
easat-10684	95	6	bi	bi	ADJ
easat-10684	95	7	-	-	ADJ
easat-10684	95	8	lstm	lstm	ADJ
easat-10684	95	9	model	model	NOUN
easat-10684	95	10	with	with	ADP
easat-10684	95	11	glove	glove	NOUN
easat-10684	95	12	,	,	PUNCT
easat-10684	95	13	fasttext	fasttext	NOUN
easat-10684	95	14	word	word	NOUN
easat-10684	95	15	embedding	embed	VERB
easat-10684	95	16	,	,	PUNCT
easat-10684	95	17	and	and	CCONJ
easat-10684	95	18	applied	apply	VERB
easat-10684	95	19	cross	cross	ADJ
easat-10684	95	20	-	-	ADJ
easat-10684	95	21	fold	fold	ADJ
easat-10684	95	22	validation	validation	NOUN
easat-10684	95	23	to	to	PART
easat-10684	95	24	detect	detect	VERB
easat-10684	95	25	bangla	bangla	PROPN
easat-10684	95	26	fake	fake	ADJ
easat-10684	95	27	news	news	NOUN
easat-10684	95	28	.	.	PUNCT
easat-10684	96	1	a	a	DET
easat-10684	96	2	gated	gate	VERB
easat-10684	96	3	recurrent	recurrent	ADJ
easat-10684	96	4	unit	unit	NOUN
easat-10684	96	5	(	(	PUNCT
easat-10684	96	6	gru	gru	NOUN
easat-10684	96	7	)	)	PUNCT
easat-10684	96	8	deep	deep	ADJ
easat-10684	96	9	learning	learning	NOUN
easat-10684	96	10	model	model	NOUN
easat-10684	96	11	is	be	AUX
easat-10684	96	12	introduced	introduce	VERB
easat-10684	96	13	to	to	PART
easat-10684	96	14	detect	detect	VERB
easat-10684	96	15	fake	fake	ADJ
easat-10684	96	16	news	news	NOUN
easat-10684	96	17	.	.	PUNCT
easat-10684	97	1	based	base	VERB
easat-10684	97	2	on	on	ADP
easat-10684	97	3	the	the	DET
easat-10684	97	4	experimental	experimental	ADJ
easat-10684	97	5	analysis	analysis	NOUN
easat-10684	97	6	conducted	conduct	VERB
easat-10684	97	7	,	,	PUNCT
easat-10684	97	8	an	an	DET
easat-10684	97	9	1518	1518	NUM
easat-10684	97	10	edelweiss	edelweiss	PROPN
easat-10684	97	11	applied	apply	VERB
easat-10684	97	12	science	science	NOUN
easat-10684	97	13	and	and	CCONJ
easat-10684	97	14	technology	technology	NOUN
easat-10684	97	15	issn	issn	PROPN
easat-10684	97	16	:	:	PUNCT
easat-10684	97	17	2576	2576	NUM
easat-10684	97	18	-	-	SYM
easat-10684	97	19	8484	8484	NUM
easat-10684	97	20	vol	vol	NOUN
easat-10684	97	21	.	.	PROPN
easat-10684	98	1	9	9	NUM
easat-10684	98	2	,	,	PUNCT
easat-10684	98	3	no	no	INTJ
easat-10684	98	4	.	.	NOUN
easat-10684	98	5	10	10	NUM
easat-10684	98	6	:	:	PUNCT
easat-10684	98	7	1515	1515	NUM
easat-10684	98	8	-	-	SYM
easat-10684	98	9	1540	1540	NUM
easat-10684	98	10	,	,	PUNCT
easat-10684	98	11	2025	2025	NUM
easat-10684	98	12	doi	doi	NOUN
easat-10684	98	13	:	:	PUNCT
easat-10684	98	14	10.55214/2576	10.55214/2576	NUM
easat-10684	98	15	-	-	SYM
easat-10684	98	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	99	1	©	©	PROPN
easat-10684	99	2	2025	2025	NUM
easat-10684	99	3	by	by	ADP
easat-10684	99	4	the	the	DET
easat-10684	99	5	authors	author	NOUN
easat-10684	99	6	;	;	PUNCT
easat-10684	99	7	licensee	licensee	PROPN
easat-10684	99	8	learning	learn	VERB
easat-10684	99	9	gate	gate	PROPN
easat-10684	99	10	accuracy	accuracy	NOUN
easat-10684	99	11	of	of	ADP
easat-10684	99	12	96	96	NUM
easat-10684	99	13	%	%	NOUN
easat-10684	99	14	was	be	AUX
easat-10684	99	15	achieved	achieve	VERB
easat-10684	99	16	using	use	VERB
easat-10684	99	17	the	the	DET
easat-10684	99	18	bi	bi	ADJ
easat-10684	99	19	-	-	ADJ
easat-10684	99	20	lstm	lstm	ADJ
easat-10684	99	21	deep	deep	ADJ
easat-10684	99	22	learning	learning	NOUN
easat-10684	99	23	model	model	NOUN
easat-10684	99	24	,	,	PUNCT
easat-10684	99	25	and	and	CCONJ
easat-10684	99	26	77	77	NUM
easat-10684	99	27	%	%	NOUN
easat-10684	99	28	accuracy	accuracy	NOUN
easat-10684	99	29	was	be	AUX
easat-10684	99	30	achieved	achieve	VERB
easat-10684	99	31	using	use	VERB
easat-10684	99	32	the	the	DET
easat-10684	99	33	gru	gru	NOUN
easat-10684	99	34	model	model	NOUN
easat-10684	99	35	for	for	ADP
easat-10684	99	36	detecting	detect	VERB
easat-10684	99	37	and	and	CCONJ
easat-10684	99	38	classifying	classify	VERB
easat-10684	99	39	fake	fake	ADJ
easat-10684	99	40	news	news	NOUN
easat-10684	99	41	from	from	ADP
easat-10684	99	42	real	real	ADJ
easat-10684	99	43	news	news	NOUN
easat-10684	99	44	.	.	PUNCT
easat-10684	100	1	pandey	pandey	PROPN
easat-10684	100	2	et	et	PROPN
easat-10684	100	3	al	al	PROPN
easat-10684	100	4	.	.	PUNCT
easat-10684	101	1	[	[	X
easat-10684	101	2	6	6	NUM
easat-10684	101	3	]	]	PUNCT
easat-10684	101	4	and	and	CCONJ
easat-10684	101	5	jaybhaye	jaybhaye	NOUN
easat-10684	101	6	[	[	X
easat-10684	101	7	13	13	NUM
easat-10684	101	8	]	]	PUNCT
easat-10684	101	9	introduce	introduce	NOUN
easat-10684	101	10	machine	machine	NOUN
easat-10684	101	11	learning	learn	VERB
easat-10684	101	12	techniques	technique	NOUN
easat-10684	101	13	such	such	ADJ
easat-10684	101	14	as	as	ADP
easat-10684	101	15	k	k	NOUN
easat-10684	101	16	-	-	PUNCT
easat-10684	101	17	nearest	near	ADJ
easat-10684	101	18	neighbour	neighbour	NOUN
easat-10684	101	19	,	,	PUNCT
easat-10684	101	20	support	support	NOUN
easat-10684	101	21	vector	vector	NOUN
easat-10684	101	22	machine	machine	NOUN
easat-10684	101	23	,	,	PUNCT
easat-10684	101	24	decision	decision	NOUN
easat-10684	101	25	tree	tree	NOUN
easat-10684	101	26	,	,	PUNCT
easat-10684	101	27	and	and	CCONJ
easat-10684	101	28	logistic	logistic	ADJ
easat-10684	101	29	regression	regression	NOUN
easat-10684	101	30	classifiers	classifier	NOUN
easat-10684	101	31	to	to	PART
easat-10684	101	32	predict	predict	VERB
easat-10684	101	33	and	and	CCONJ
easat-10684	101	34	identify	identify	VERB
easat-10684	101	35	fake	fake	ADJ
easat-10684	101	36	news	news	NOUN
easat-10684	101	37	from	from	ADP
easat-10684	101	38	real	real	ADJ
easat-10684	101	39	news	news	NOUN
easat-10684	101	40	accurately	accurately	ADV
easat-10684	101	41	.	.	PUNCT
easat-10684	102	1	performance	performance	NOUN
easat-10684	102	2	comparison	comparison	NOUN
easat-10684	102	3	of	of	ADP
easat-10684	102	4	pandey	pandey	PROPN
easat-10684	102	5	et	et	PROPN
easat-10684	102	6	al	al	PROPN
easat-10684	102	7	.	.	PUNCT
easat-10684	103	1	[	[	X
easat-10684	103	2	6	6	X
easat-10684	103	3	]	]	PUNCT
easat-10684	103	4	shows	show	VERB
easat-10684	103	5	that	that	SCONJ
easat-10684	103	6	logistic	logistic	ADJ
easat-10684	103	7	regression	regression	NOUN
easat-10684	103	8	performs	perform	VERB
easat-10684	103	9	best	well	ADV
easat-10684	103	10	with	with	ADP
easat-10684	103	11	90.48	90.48	NUM
easat-10684	103	12	%	%	NOUN
easat-10684	103	13	accuracy	accuracy	NOUN
easat-10684	103	14	.	.	PUNCT
easat-10684	104	1	however	however	ADV
easat-10684	104	2	,	,	PUNCT
easat-10684	104	3	jaybhaye	jaybhaye	VERB
easat-10684	104	4	[	[	X
easat-10684	104	5	13	13	NUM
easat-10684	104	6	]	]	PUNCT
easat-10684	104	7	face	face	NOUN
easat-10684	104	8	challenges	challenge	NOUN
easat-10684	104	9	of	of	ADP
easat-10684	104	10	data	datum	NOUN
easat-10684	104	11	overfitting	overfitte	VERB
easat-10684	104	12	during	during	ADP
easat-10684	104	13	training	training	NOUN
easat-10684	104	14	,	,	PUNCT
easat-10684	104	15	which	which	PRON
easat-10684	104	16	resulted	result	VERB
easat-10684	104	17	in	in	ADP
easat-10684	104	18	poor	poor	ADJ
easat-10684	104	19	performance	performance	NOUN
easat-10684	104	20	on	on	ADP
easat-10684	104	21	testing	test	VERB
easat-10684	104	22	data	datum	NOUN
easat-10684	104	23	.	.	PUNCT
easat-10684	105	1	the	the	DET
easat-10684	105	2	researcher	researcher	NOUN
easat-10684	105	3	obtained	obtain	VERB
easat-10684	105	4	86	86	NUM
easat-10684	105	5	%	%	NOUN
easat-10684	105	6	accuracy	accuracy	NOUN
easat-10684	105	7	using	use	VERB
easat-10684	105	8	the	the	DET
easat-10684	105	9	lstm	lstm	PROPN
easat-10684	105	10	model	model	NOUN
easat-10684	105	11	,	,	PUNCT
easat-10684	105	12	and	and	CCONJ
easat-10684	105	13	it	it	PRON
easat-10684	105	14	was	be	AUX
easat-10684	105	15	suggested	suggest	VERB
easat-10684	105	16	that	that	SCONJ
easat-10684	105	17	the	the	DET
easat-10684	105	18	researcher	researcher	NOUN
easat-10684	105	19	build	build	VERB
easat-10684	105	20	a	a	DET
easat-10684	105	21	fake	fake	ADJ
easat-10684	105	22	news	news	NOUN
easat-10684	105	23	model	model	NOUN
easat-10684	105	24	due	due	ADP
easat-10684	105	25	to	to	ADP
easat-10684	105	26	its	its	PRON
easat-10684	105	27	better	well	ADJ
easat-10684	105	28	performance	performance	NOUN
easat-10684	105	29	than	than	ADP
easat-10684	105	30	the	the	DET
easat-10684	105	31	machine	machine	NOUN
easat-10684	105	32	learning	learn	VERB
easat-10684	105	33	model	model	NOUN
easat-10684	105	34	.	.	PUNCT
easat-10684	106	1	however	however	ADV
easat-10684	106	2	,	,	PUNCT
easat-10684	106	3	kaliyar	kaliyar	PROPN
easat-10684	106	4	et	et	PROPN
easat-10684	106	5	al	al	PROPN
easat-10684	106	6	.	.	PUNCT
easat-10684	107	1	[	[	X
easat-10684	107	2	14	14	NUM
easat-10684	107	3	]	]	PUNCT
easat-10684	107	4	revealed	reveal	VERB
easat-10684	107	5	that	that	SCONJ
easat-10684	107	6	the	the	DET
easat-10684	107	7	use	use	NOUN
easat-10684	107	8	of	of	ADP
easat-10684	107	9	bidirectional	bidirectional	ADJ
easat-10684	107	10	encoding	encoding	NOUN
easat-10684	107	11	for	for	ADP
easat-10684	107	12	fake	fake	ADJ
easat-10684	107	13	news	news	NOUN
easat-10684	107	14	detection	detection	NOUN
easat-10684	107	15	performs	perform	VERB
easat-10684	107	16	better	well	ADJ
easat-10684	107	17	than	than	ADP
easat-10684	107	18	the	the	DET
easat-10684	107	19	unidirectional	unidirectional	ADJ
easat-10684	107	20	-	-	PUNCT
easat-10684	107	21	based	base	VERB
easat-10684	107	22	model	model	NOUN
easat-10684	107	23	.	.	PUNCT
easat-10684	108	1	hence	hence	ADV
easat-10684	108	2	,	,	PUNCT
easat-10684	108	3	the	the	DET
easat-10684	108	4	researcher	researcher	NOUN
easat-10684	108	5	proposed	propose	VERB
easat-10684	108	6	a	a	DET
easat-10684	108	7	bidirectional	bidirectional	ADJ
easat-10684	108	8	encoder	encoder	NOUN
easat-10684	108	9	representation	representation	NOUN
easat-10684	108	10	from	from	ADP
easat-10684	108	11	a	a	DET
easat-10684	108	12	transformer	transformer	NOUN
easat-10684	108	13	(	(	PUNCT
easat-10684	108	14	fakebert	fakebert	NOUN
easat-10684	108	15	)	)	PUNCT
easat-10684	108	16	model	model	NOUN
easat-10684	108	17	integrated	integrate	VERB
easat-10684	108	18	with	with	ADP
easat-10684	108	19	a	a	DET
easat-10684	108	20	single	single	ADJ
easat-10684	108	21	-	-	PUNCT
easat-10684	108	22	layer	layer	NOUN
easat-10684	108	23	deep	deep	ADJ
easat-10684	108	24	cnn	cnn	PROPN
easat-10684	108	25	network	network	NOUN
easat-10684	108	26	.	.	PUNCT
easat-10684	109	1	as	as	ADP
easat-10684	109	2	a	a	DET
easat-10684	109	3	result	result	NOUN
easat-10684	109	4	,	,	PUNCT
easat-10684	109	5	the	the	DET
easat-10684	109	6	developed	develop	VERB
easat-10684	109	7	model	model	NOUN
easat-10684	109	8	achieved	achieve	VERB
easat-10684	109	9	an	an	DET
easat-10684	109	10	accuracy	accuracy	NOUN
easat-10684	109	11	of	of	ADP
easat-10684	109	12	98.90	98.90	NUM
easat-10684	109	13	%	%	NOUN
easat-10684	109	14	,	,	PUNCT
easat-10684	109	15	which	which	PRON
easat-10684	109	16	outperformed	outperform	VERB
easat-10684	109	17	the	the	DET
easat-10684	109	18	existing	exist	VERB
easat-10684	109	19	study	study	NOUN
easat-10684	109	20	.	.	PUNCT
easat-10684	110	1	7furthermore	7furthermore	X
easat-10684	110	2	,	,	PUNCT
easat-10684	110	3	nasir	nasir	PROPN
easat-10684	110	4	et	et	PROPN
easat-10684	110	5	al	al	PROPN
easat-10684	110	6	.	.	PUNCT
easat-10684	111	1	[	[	X
easat-10684	111	2	23	23	NUM
easat-10684	111	3	]	]	PUNCT
easat-10684	111	4	adopt	adopt	VERB
easat-10684	111	5	the	the	DET
easat-10684	111	6	hybridization	hybridization	NOUN
easat-10684	111	7	of	of	ADP
easat-10684	111	8	convolutional	convolutional	ADJ
easat-10684	111	9	neural	neural	ADJ
easat-10684	111	10	network	network	NOUN
easat-10684	111	11	(	(	PUNCT
easat-10684	111	12	cnn	cnn	PROPN
easat-10684	111	13	)	)	PUNCT
easat-10684	111	14	and	and	CCONJ
easat-10684	111	15	recurrent	recurrent	ADJ
easat-10684	111	16	neural	neural	ADJ
easat-10684	111	17	network	network	NOUN
easat-10684	111	18	(	(	PUNCT
easat-10684	111	19	rnn	rnn	PROPN
easat-10684	111	20	)	)	PUNCT
easat-10684	111	21	for	for	ADP
easat-10684	111	22	detecting	detect	VERB
easat-10684	111	23	fake	fake	ADJ
easat-10684	111	24	news	news	NOUN
easat-10684	111	25	.	.	PUNCT
easat-10684	112	1	the	the	DET
easat-10684	112	2	proposed	propose	VERB
easat-10684	112	3	cnn	cnn	PROPN
easat-10684	112	4	-	-	PUNCT
easat-10684	112	5	rnn	rnn	PROPN
easat-10684	112	6	model	model	NOUN
easat-10684	112	7	was	be	AUX
easat-10684	112	8	trained	train	VERB
easat-10684	112	9	on	on	ADP
easat-10684	112	10	both	both	CCONJ
easat-10684	112	11	isot	isot	NOUN
easat-10684	112	12	and	and	CCONJ
easat-10684	112	13	fake	fake	ADJ
easat-10684	112	14	datasets	dataset	NOUN
easat-10684	112	15	,	,	PUNCT
easat-10684	112	16	with	with	ADP
easat-10684	112	17	approximately	approximately	ADV
easat-10684	112	18	100	100	NUM
easat-10684	112	19	%	%	NOUN
easat-10684	112	20	accuracy	accuracy	NOUN
easat-10684	112	21	achieved	achieve	VERB
easat-10684	112	22	using	use	VERB
easat-10684	112	23	the	the	DET
easat-10684	112	24	isot	isot	NOUN
easat-10684	112	25	dataset	dataset	NOUN
easat-10684	112	26	and	and	CCONJ
easat-10684	112	27	60	60	NUM
easat-10684	112	28	%	%	NOUN
easat-10684	112	29	accuracy	accuracy	NOUN
easat-10684	112	30	using	use	VERB
easat-10684	112	31	the	the	DET
easat-10684	112	32	fake	fake	ADJ
easat-10684	112	33	dataset	dataset	NOUN
easat-10684	112	34	.	.	PUNCT
easat-10684	113	1	thota	thota	PROPN
easat-10684	113	2	et	et	PROPN
easat-10684	113	3	al	al	PROPN
easat-10684	113	4	.	.	PUNCT
easat-10684	114	1	[	[	X
easat-10684	114	2	24	24	NUM
easat-10684	114	3	]	]	PUNCT
easat-10684	114	4	also	also	ADV
easat-10684	114	5	address	address	VERB
easat-10684	114	6	the	the	DET
easat-10684	114	7	same	same	ADJ
easat-10684	114	8	issues	issue	NOUN
easat-10684	114	9	using	use	VERB
easat-10684	114	10	the	the	DET
easat-10684	114	11	hybridization	hybridization	NOUN
easat-10684	114	12	of	of	ADP
easat-10684	114	13	bidirectional	bidirectional	ADJ
easat-10684	114	14	encoder	encoder	NOUN
easat-10684	114	15	representations	representation	VERB
easat-10684	114	16	from	from	ADP
easat-10684	114	17	transformers	transformer	NOUN
easat-10684	114	18	(	(	PUNCT
easat-10684	114	19	bert	bert	PROPN
easat-10684	114	20	)	)	PUNCT
easat-10684	114	21	,	,	PUNCT
easat-10684	114	22	alite	alite	NOUN
easat-10684	114	23	-	-	PUNCT
easat-10684	114	24	bert	bert	NOUN
easat-10684	114	25	,	,	PUNCT
easat-10684	114	26	and	and	CCONJ
easat-10684	114	27	transformer	transformer	NOUN
easat-10684	114	28	-	-	PUNCT
easat-10684	114	29	xl	xl	PROPN
easat-10684	114	30	style	style	NOUN
easat-10684	114	31	segmented	segment	VERB
easat-10684	114	32	memory	memory	NOUN
easat-10684	114	33	(	(	PUNCT
easat-10684	114	34	xlnet	xlnet	PROPN
easat-10684	114	35	)	)	PUNCT
easat-10684	114	36	.	.	PUNCT
easat-10684	115	1	the	the	DET
easat-10684	115	2	model	model	NOUN
easat-10684	115	3	performs	perform	VERB
easat-10684	115	4	efficiently	efficiently	ADV
easat-10684	115	5	with	with	ADP
easat-10684	115	6	98	98	NUM
easat-10684	115	7	%	%	NOUN
easat-10684	115	8	precision	precision	NOUN
easat-10684	115	9	,	,	PUNCT
easat-10684	115	10	recall	recall	NOUN
easat-10684	115	11	,	,	PUNCT
easat-10684	115	12	accuracy	accuracy	NOUN
easat-10684	115	13	,	,	PUNCT
easat-10684	115	14	and	and	CCONJ
easat-10684	115	15	f1	f1	NOUN
easat-10684	115	16	-	-	PUNCT
easat-10684	115	17	score	score	NOUN
easat-10684	115	18	.	.	PUNCT
easat-10684	116	1	in	in	ADP
easat-10684	116	2	today	today	NOUN
easat-10684	116	3	’s	’s	PART
easat-10684	116	4	world	world	NOUN
easat-10684	116	5	,	,	PUNCT
easat-10684	116	6	social	social	ADJ
easat-10684	116	7	media	medium	NOUN
easat-10684	116	8	platforms	platform	NOUN
easat-10684	116	9	are	be	AUX
easat-10684	116	10	ranked	rank	VERB
easat-10684	116	11	as	as	ADP
easat-10684	116	12	the	the	DET
easat-10684	116	13	primary	primary	ADJ
easat-10684	116	14	internet	internet	NOUN
easat-10684	116	15	tools	tool	NOUN
easat-10684	116	16	used	use	VERB
easat-10684	116	17	to	to	PART
easat-10684	116	18	spread	spread	VERB
easat-10684	116	19	news	news	NOUN
easat-10684	116	20	across	across	ADP
easat-10684	116	21	the	the	DET
easat-10684	116	22	globe	globe	NOUN
easat-10684	116	23	.	.	PUNCT
easat-10684	117	1	choudhary	choudhary	PROPN
easat-10684	117	2	and	and	CCONJ
easat-10684	117	3	arora	arora	PROPN
easat-10684	118	1	[	[	X
easat-10684	118	2	25	25	NUM
easat-10684	118	3	]	]	PUNCT
easat-10684	118	4	claim	claim	NOUN
easat-10684	118	5	that	that	SCONJ
easat-10684	118	6	social	social	ADJ
easat-10684	118	7	media	media	NOUN
easat-10684	118	8	is	be	AUX
easat-10684	118	9	an	an	DET
easat-10684	118	10	essential	essential	ADJ
easat-10684	118	11	tool	tool	NOUN
easat-10684	118	12	for	for	ADP
easat-10684	118	13	disseminating	disseminate	VERB
easat-10684	118	14	information	information	NOUN
easat-10684	118	15	and	and	CCONJ
easat-10684	118	16	creating	create	VERB
easat-10684	118	17	awareness	awareness	NOUN
easat-10684	118	18	among	among	ADP
easat-10684	118	19	the	the	DET
easat-10684	118	20	public	public	NOUN
easat-10684	118	21	,	,	PUNCT
easat-10684	118	22	but	but	CCONJ
easat-10684	118	23	it	it	PRON
easat-10684	118	24	has	have	VERB
easat-10684	118	25	no	no	DET
easat-10684	118	26	authentication	authentication	NOUN
easat-10684	118	27	schemes	scheme	NOUN
easat-10684	118	28	for	for	ADP
easat-10684	118	29	validating	validate	VERB
easat-10684	118	30	the	the	DET
easat-10684	118	31	origin	origin	NOUN
easat-10684	118	32	and	and	CCONJ
easat-10684	118	33	content	content	NOUN
easat-10684	118	34	of	of	ADP
easat-10684	118	35	all	all	DET
easat-10684	118	36	published	publish	VERB
easat-10684	118	37	news	news	NOUN
easat-10684	118	38	.	.	PUNCT
easat-10684	119	1	a	a	DET
easat-10684	119	2	linguistic	linguistic	ADJ
easat-10684	119	3	model	model	NOUN
easat-10684	119	4	for	for	ADP
easat-10684	119	5	extracting	extract	VERB
easat-10684	119	6	essential	essential	ADJ
easat-10684	119	7	syntactic	syntactic	ADJ
easat-10684	119	8	and	and	CCONJ
easat-10684	119	9	semantic	semantic	ADJ
easat-10684	119	10	information	information	NOUN
easat-10684	119	11	describing	describe	VERB
easat-10684	119	12	the	the	DET
easat-10684	119	13	features	feature	NOUN
easat-10684	119	14	of	of	ADP
easat-10684	119	15	fake	fake	ADJ
easat-10684	119	16	news	news	NOUN
easat-10684	119	17	from	from	ADP
easat-10684	119	18	real	real	ADJ
easat-10684	119	19	news	news	NOUN
easat-10684	119	20	is	be	AUX
easat-10684	119	21	proposed	propose	VERB
easat-10684	119	22	by	by	ADP
easat-10684	119	23	the	the	DET
easat-10684	119	24	researchers	researcher	NOUN
easat-10684	119	25	to	to	PART
easat-10684	119	26	address	address	VERB
easat-10684	119	27	the	the	DET
easat-10684	119	28	widespread	widespread	ADJ
easat-10684	119	29	spread	spread	NOUN
easat-10684	119	30	of	of	ADP
easat-10684	119	31	phony	phony	ADJ
easat-10684	119	32	news	news	NOUN
easat-10684	119	33	uploaded	upload	VERB
easat-10684	119	34	on	on	ADP
easat-10684	119	35	the	the	DET
easat-10684	119	36	internet	internet	NOUN
easat-10684	119	37	.	.	PUNCT
easat-10684	120	1	the	the	DET
easat-10684	120	2	developed	develop	VERB
easat-10684	120	3	model	model	NOUN
easat-10684	120	4	was	be	AUX
easat-10684	120	5	able	able	ADJ
easat-10684	120	6	to	to	PART
easat-10684	120	7	achieve	achieve	VERB
easat-10684	120	8	an	an	DET
easat-10684	120	9	average	average	ADJ
easat-10684	120	10	accuracy	accuracy	NOUN
easat-10684	120	11	of	of	ADP
easat-10684	120	12	86	86	NUM
easat-10684	120	13	%	%	NOUN
easat-10684	120	14	for	for	ADP
easat-10684	120	15	detecting	detect	VERB
easat-10684	120	16	and	and	CCONJ
easat-10684	120	17	classifying	classify	VERB
easat-10684	120	18	phony	phony	ADJ
easat-10684	120	19	news	news	NOUN
easat-10684	120	20	.	.	PUNCT
easat-10684	121	1	the	the	DET
easat-10684	121	2	comparative	comparative	ADJ
easat-10684	121	3	analysis	analysis	NOUN
easat-10684	121	4	shows	show	VERB
easat-10684	121	5	that	that	SCONJ
easat-10684	121	6	the	the	DET
easat-10684	121	7	developed	develop	VERB
easat-10684	121	8	model	model	NOUN
easat-10684	121	9	performs	perform	VERB
easat-10684	121	10	more	more	ADV
easat-10684	121	11	efficiently	efficiently	ADV
easat-10684	121	12	than	than	ADP
easat-10684	121	13	the	the	DET
easat-10684	121	14	existing	exist	VERB
easat-10684	121	15	models	model	NOUN
easat-10684	121	16	.	.	PUNCT
easat-10684	122	1	vijjali	vijjali	PROPN
easat-10684	122	2	et	et	PROPN
easat-10684	122	3	al	al	PROPN
easat-10684	122	4	.	.	PUNCT
easat-10684	123	1	[	[	X
easat-10684	123	2	26	26	NUM
easat-10684	123	3	]	]	PUNCT
easat-10684	123	4	explore	explore	VERB
easat-10684	123	5	the	the	DET
easat-10684	123	6	alarming	alarming	ADJ
easat-10684	123	7	rate	rate	NOUN
easat-10684	123	8	of	of	ADP
easat-10684	123	9	fake	fake	ADJ
easat-10684	123	10	news	news	NOUN
easat-10684	123	11	occurrence	occurrence	NOUN
easat-10684	123	12	during	during	ADP
easat-10684	123	13	the	the	DET
easat-10684	123	14	covid-19	covid-19	PROPN
easat-10684	123	15	pandemic	pandemic	NOUN
easat-10684	123	16	,	,	PUNCT
easat-10684	123	17	which	which	PRON
easat-10684	123	18	results	result	VERB
easat-10684	123	19	in	in	ADP
easat-10684	123	20	people	people	NOUN
easat-10684	123	21	finding	find	VERB
easat-10684	123	22	it	it	PRON
easat-10684	123	23	difficult	difficult	ADJ
easat-10684	123	24	to	to	PART
easat-10684	123	25	differentiate	differentiate	VERB
easat-10684	123	26	false	false	ADJ
easat-10684	123	27	information	information	NOUN
easat-10684	123	28	from	from	ADP
easat-10684	123	29	real	real	ADJ
easat-10684	123	30	covid-19	covid-19	PROPN
easat-10684	123	31	news	news	NOUN
easat-10684	123	32	.	.	PUNCT
easat-10684	124	1	hence	hence	ADV
easat-10684	124	2	,	,	PUNCT
easat-10684	124	3	the	the	DET
easat-10684	124	4	researcher	researcher	NOUN
easat-10684	124	5	proposed	propose	VERB
easat-10684	124	6	a	a	DET
easat-10684	124	7	state	state	NOUN
easat-10684	124	8	-	-	PUNCT
easat-10684	124	9	of	of	ADP
easat-10684	124	10	-	-	PUNCT
easat-10684	124	11	the	the	DET
easat-10684	124	12	-	-	PUNCT
easat-10684	124	13	art	art	NOUN
easat-10684	124	14	machine	machine	NOUN
easat-10684	124	15	learning	learning	NOUN
easat-10684	124	16	model	model	NOUN
easat-10684	124	17	based	base	VERB
easat-10684	124	18	on	on	ADP
easat-10684	124	19	a	a	DET
easat-10684	124	20	two	two	NUM
easat-10684	124	21	-	-	PUNCT
easat-10684	124	22	level	level	NOUN
easat-10684	124	23	automated	automate	VERB
easat-10684	124	24	pipeline	pipeline	NOUN
easat-10684	124	25	for	for	ADP
easat-10684	124	26	detecting	detect	VERB
easat-10684	124	27	covid-19	covid-19	PROPN
easat-10684	124	28	fake	fake	ADJ
easat-10684	124	29	news	news	NOUN
easat-10684	124	30	.	.	PUNCT
easat-10684	125	1	the	the	DET
easat-10684	125	2	study	study	NOUN
easat-10684	125	3	reveals	reveal	VERB
easat-10684	125	4	that	that	SCONJ
easat-10684	125	5	the	the	DET
easat-10684	125	6	bidirectional	bidirectional	ADJ
easat-10684	125	7	encoder	encoder	NOUN
easat-10684	125	8	representations	representation	VERB
easat-10684	125	9	from	from	ADP
easat-10684	125	10	transformers	transformer	NOUN
easat-10684	125	11	(	(	PUNCT
easat-10684	125	12	bert	bert	PROPN
easat-10684	125	13	)	)	PUNCT
easat-10684	125	14	(	(	PUNCT
easat-10684	125	15	92	92	NUM
easat-10684	125	16	%	%	NOUN
easat-10684	125	17	)	)	PUNCT
easat-10684	125	18	and	and	CCONJ
easat-10684	125	19	the	the	DET
easat-10684	125	20	albert	albert	PROPN
easat-10684	125	21	(	(	PUNCT
easat-10684	125	22	94	94	NUM
easat-10684	125	23	%	%	NOUN
easat-10684	125	24	)	)	PUNCT
easat-10684	125	25	models	model	NOUN
easat-10684	125	26	perform	perform	VERB
easat-10684	125	27	satisfactorily	satisfactorily	ADV
easat-10684	125	28	.	.	PUNCT
easat-10684	126	1	furthermore	furthermore	ADV
easat-10684	126	2	,	,	PUNCT
easat-10684	126	3	most	most	ADV
easat-10684	126	4	fake	fake	ADJ
easat-10684	126	5	information	information	NOUN
easat-10684	126	6	created	create	VERB
easat-10684	126	7	on	on	ADP
easat-10684	126	8	the	the	DET
easat-10684	126	9	internet	internet	NOUN
easat-10684	126	10	today	today	NOUN
easat-10684	126	11	is	be	AUX
easat-10684	126	12	a	a	DET
easat-10684	126	13	result	result	NOUN
easat-10684	126	14	of	of	ADP
easat-10684	126	15	manipulated	manipulate	VERB
easat-10684	126	16	image	image	NOUN
easat-10684	126	17	,	,	PUNCT
easat-10684	126	18	audio	audio	NOUN
easat-10684	126	19	,	,	PUNCT
easat-10684	126	20	text	text	NOUN
easat-10684	126	21	,	,	PUNCT
easat-10684	126	22	and	and	CCONJ
easat-10684	126	23	video	video	NOUN
easat-10684	126	24	data	datum	NOUN
easat-10684	126	25	being	be	AUX
easat-10684	126	26	uploaded	upload	VERB
easat-10684	126	27	online	online	ADV
easat-10684	126	28	.	.	PUNCT
easat-10684	127	1	singhal	singhal	PROPN
easat-10684	127	2	et	et	PROPN
easat-10684	127	3	al	al	PROPN
easat-10684	127	4	.	.	PUNCT
easat-10684	128	1	[	[	X
easat-10684	128	2	27	27	NUM
easat-10684	128	3	]	]	PUNCT
easat-10684	128	4	revealed	reveal	VERB
easat-10684	128	5	that	that	SCONJ
easat-10684	128	6	manipulated	manipulate	VERB
easat-10684	128	7	data	datum	NOUN
easat-10684	128	8	could	could	AUX
easat-10684	128	9	easily	easily	ADV
easat-10684	128	10	spread	spread	VERB
easat-10684	128	11	via	via	ADP
easat-10684	128	12	social	social	ADJ
easat-10684	128	13	media	medium	NOUN
easat-10684	128	14	due	due	ADP
easat-10684	128	15	to	to	ADP
easat-10684	128	16	the	the	DET
easat-10684	128	17	fast	fast	ADV
easat-10684	128	18	-	-	PUNCT
easat-10684	128	19	growing	grow	VERB
easat-10684	128	20	nature	nature	NOUN
easat-10684	128	21	of	of	ADP
easat-10684	128	22	social	social	ADJ
easat-10684	128	23	media	medium	NOUN
easat-10684	128	24	users	user	NOUN
easat-10684	128	25	.	.	PUNCT
easat-10684	129	1	the	the	DET
easat-10684	129	2	researcher	researcher	NOUN
easat-10684	129	3	proposed	propose	VERB
easat-10684	129	4	a	a	DET
easat-10684	129	5	multimodal	multimodal	NOUN
easat-10684	129	6	-	-	PUNCT
easat-10684	129	7	based	base	VERB
easat-10684	129	8	fake	fake	ADJ
easat-10684	129	9	news	news	NOUN
easat-10684	129	10	detection	detection	NOUN
easat-10684	129	11	system	system	NOUN
easat-10684	129	12	,	,	PUNCT
easat-10684	129	13	which	which	PRON
easat-10684	129	14	internally	internally	ADV
easat-10684	129	15	utilized	utilize	VERB
easat-10684	129	16	a	a	DET
easat-10684	129	17	bidirectional	bidirectional	ADJ
easat-10684	129	18	encoder	encoder	NOUN
easat-10684	129	19	from	from	ADP
easat-10684	129	20	transformer	transformer	NOUN
easat-10684	129	21	(	(	PUNCT
easat-10684	129	22	bert	bert	NOUN
easat-10684	129	23	)	)	PUNCT
easat-10684	129	24	architecture	architecture	NOUN
easat-10684	129	25	to	to	PART
easat-10684	129	26	extract	extract	VERB
easat-10684	129	27	features	feature	NOUN
easat-10684	129	28	from	from	ADP
easat-10684	129	29	text	text	NOUN
easat-10684	129	30	and	and	CCONJ
easat-10684	129	31	utilized	utilize	VERB
easat-10684	129	32	the	the	DET
easat-10684	129	33	vgg19	vgg19	NOUN
easat-10684	129	34	pre	pre	ADJ
easat-10684	129	35	-	-	ADJ
easat-10684	129	36	trained	train	VERB
easat-10684	129	37	model	model	NOUN
easat-10684	129	38	for	for	ADP
easat-10684	129	39	image	image	NOUN
easat-10684	129	40	feature	feature	NOUN
easat-10684	129	41	extraction	extraction	NOUN
easat-10684	129	42	.	.	PUNCT
easat-10684	130	1	as	as	ADP
easat-10684	130	2	a	a	DET
easat-10684	130	3	result	result	NOUN
easat-10684	130	4	,	,	PUNCT
easat-10684	130	5	the	the	DET
easat-10684	130	6	developed	develop	VERB
easat-10684	130	7	model	model	NOUN
easat-10684	130	8	outperformed	outperform	VERB
easat-10684	130	9	the	the	DET
easat-10684	130	10	existing	exist	VERB
easat-10684	130	11	state	state	NOUN
easat-10684	130	12	-	-	PUNCT
easat-10684	130	13	of	of	ADP
easat-10684	130	14	-	-	PUNCT
easat-10684	130	15	the	the	DET
easat-10684	130	16	-	-	PUNCT
easat-10684	130	17	art	art	NOUN
easat-10684	130	18	model	model	NOUN
easat-10684	130	19	trained	train	VERB
easat-10684	130	20	on	on	ADP
easat-10684	130	21	twitter	twitter	NOUN
easat-10684	130	22	and	and	CCONJ
easat-10684	130	23	weibo	weibo	NOUN
easat-10684	130	24	datasets	dataset	NOUN
easat-10684	130	25	with	with	ADP
easat-10684	130	26	an	an	DET
easat-10684	130	27	accuracy	accuracy	NOUN
easat-10684	130	28	increase	increase	NOUN
easat-10684	130	29	of	of	ADP
easat-10684	130	30	3.27	3.27	NUM
easat-10684	130	31	%	%	NOUN
easat-10684	130	32	and	and	CCONJ
easat-10684	130	33	6.83	6.83	NUM
easat-10684	130	34	%	%	NOUN
easat-10684	130	35	.	.	PUNCT
easat-10684	131	1	kumari	kumari	PROPN
easat-10684	131	2	and	and	CCONJ
easat-10684	131	3	singh	singh	PROPN
easat-10684	132	1	[	[	X
easat-10684	132	2	28	28	NUM
easat-10684	132	3	]	]	PUNCT
easat-10684	132	4	presented	present	VERB
easat-10684	132	5	a	a	DET
easat-10684	132	6	dl	dl	PROPN
easat-10684	132	7	multimodal	multimodal	NOUN
easat-10684	132	8	framework	framework	NOUN
easat-10684	132	9	for	for	ADP
easat-10684	132	10	fake	fake	ADJ
easat-10684	132	11	news	news	NOUN
easat-10684	132	12	detection	detection	NOUN
easat-10684	132	13	that	that	PRON
easat-10684	132	14	significantly	significantly	ADV
easat-10684	132	15	outperforms	outperform	VERB
easat-10684	132	16	existing	exist	VERB
easat-10684	132	17	state	state	NOUN
easat-10684	132	18	-	-	PUNCT
easat-10684	132	19	of	of	ADP
easat-10684	132	20	-	-	PUNCT
easat-10684	132	21	the	the	DET
easat-10684	132	22	-	-	PUNCT
easat-10684	132	23	art	art	NOUN
easat-10684	132	24	techniques	technique	NOUN
easat-10684	132	25	.	.	PUNCT
easat-10684	133	1	it	it	PRON
easat-10684	133	2	achieves	achieve	VERB
easat-10684	133	3	an	an	DET
easat-10684	133	4	accuracy	accuracy	NOUN
easat-10684	133	5	of	of	ADP
easat-10684	133	6	99.22	99.22	NUM
easat-10684	133	7	%	%	NOUN
easat-10684	133	8	for	for	ADP
easat-10684	133	9	text	text	NOUN
easat-10684	133	10	analysis	analysis	NOUN
easat-10684	133	11	and	and	CCONJ
easat-10684	133	12	93.12	93.12	NUM
easat-10684	133	13	%	%	NOUN
easat-10684	133	14	when	when	SCONJ
easat-10684	133	15	combining	combine	VERB
easat-10684	133	16	text	text	NOUN
easat-10684	133	17	and	and	CCONJ
easat-10684	133	18	images	image	NOUN
easat-10684	133	19	.	.	PUNCT
easat-10684	134	1	this	this	PRON
easat-10684	134	2	demonstrates	demonstrate	VERB
easat-10684	134	3	that	that	SCONJ
easat-10684	134	4	advanced	advance	VERB
easat-10684	134	5	ml	ml	X
easat-10684	134	6	and	and	CCONJ
easat-10684	134	7	dl	dl	PROPN
easat-10684	134	8	models	model	NOUN
easat-10684	134	9	can	can	AUX
easat-10684	134	10	effectively	effectively	ADV
easat-10684	134	11	address	address	VERB
easat-10684	134	12	the	the	DET
easat-10684	134	13	challenges	challenge	NOUN
easat-10684	134	14	of	of	ADP
easat-10684	134	15	fake	fake	ADJ
easat-10684	134	16	news	news	NOUN
easat-10684	134	17	detection	detection	NOUN
easat-10684	134	18	,	,	PUNCT
easat-10684	134	19	surpassing	surpass	VERB
easat-10684	134	20	traditional	traditional	ADJ
easat-10684	134	21	methods	method	NOUN
easat-10684	134	22	by	by	ADP
easat-10684	134	23	leveraging	leverage	VERB
easat-10684	134	24	nlp	nlp	NOUN
easat-10684	134	25	and	and	CCONJ
easat-10684	134	26	image	image	VERB
easat-10684	134	27	analysis	analysis	NOUN
easat-10684	134	28	through	through	ADP
easat-10684	134	29	lstm	lstm	NOUN
easat-10684	134	30	networks	network	NOUN
easat-10684	134	31	and	and	CCONJ
easat-10684	134	32	the	the	DET
easat-10684	134	33	clip	clip	NOUN
easat-10684	134	34	model	model	NOUN
easat-10684	134	35	.	.	PUNCT
easat-10684	135	1	gul	gul	PROPN
easat-10684	135	2	et	et	PROPN
easat-10684	135	3	al	al	PROPN
easat-10684	135	4	.	.	PUNCT
easat-10684	136	1	[	[	X
easat-10684	136	2	29	29	NUM
easat-10684	136	3	]	]	PUNCT
easat-10684	136	4	introduced	introduce	VERB
easat-10684	136	5	an	an	DET
easat-10684	136	6	innovative	innovative	ADJ
easat-10684	136	7	approach	approach	NOUN
easat-10684	136	8	that	that	PRON
easat-10684	136	9	combines	combine	VERB
easat-10684	136	10	gnns	gnn	NOUN
easat-10684	136	11	with	with	ADP
easat-10684	136	12	advanced	advanced	ADJ
easat-10684	136	13	dl	dl	PROPN
easat-10684	136	14	techniques	technique	NOUN
easat-10684	136	15	to	to	PART
easat-10684	136	16	enhance	enhance	VERB
easat-10684	136	17	the	the	DET
easat-10684	136	18	detection	detection	NOUN
easat-10684	136	19	of	of	ADP
easat-10684	136	20	fake	fake	ADJ
easat-10684	136	21	news	news	NOUN
easat-10684	136	22	,	,	PUNCT
easat-10684	136	23	moving	move	VERB
easat-10684	136	24	beyond	beyond	ADP
easat-10684	136	25	traditional	traditional	ADJ
easat-10684	136	26	methods	method	NOUN
easat-10684	136	27	that	that	PRON
easat-10684	136	28	focus	focus	VERB
easat-10684	136	29	mainly	mainly	ADV
easat-10684	136	30	on	on	ADP
easat-10684	136	31	text	text	NOUN
easat-10684	136	32	analysis	analysis	NOUN
easat-10684	136	33	and	and	CCONJ
easat-10684	136	34	1519	1519	NUM
easat-10684	136	35	edelweiss	edelweiss	PROPN
easat-10684	136	36	applied	apply	VERB
easat-10684	136	37	science	science	NOUN
easat-10684	136	38	and	and	CCONJ
easat-10684	136	39	technology	technology	NOUN
easat-10684	136	40	issn	issn	PROPN
easat-10684	136	41	:	:	PUNCT
easat-10684	136	42	2576	2576	NUM
easat-10684	136	43	-	-	SYM
easat-10684	136	44	8484	8484	NUM
easat-10684	136	45	vol	vol	NOUN
easat-10684	136	46	.	.	PROPN
easat-10684	137	1	9	9	NUM
easat-10684	137	2	,	,	PUNCT
easat-10684	137	3	no	no	INTJ
easat-10684	137	4	.	.	NOUN
easat-10684	137	5	10	10	NUM
easat-10684	137	6	:	:	PUNCT
easat-10684	137	7	1515	1515	NUM
easat-10684	137	8	-	-	SYM
easat-10684	137	9	1540	1540	NUM
easat-10684	137	10	,	,	PUNCT
easat-10684	137	11	2025	2025	NUM
easat-10684	137	12	doi	doi	NOUN
easat-10684	137	13	:	:	PUNCT
easat-10684	137	14	10.55214/2576	10.55214/2576	NUM
easat-10684	137	15	-	-	SYM
easat-10684	137	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	138	1	©	©	PROPN
easat-10684	138	2	2025	2025	NUM
easat-10684	138	3	by	by	ADP
easat-10684	138	4	the	the	DET
easat-10684	138	5	authors	author	NOUN
easat-10684	138	6	;	;	PUNCT
easat-10684	138	7	licensee	licensee	PROPN
easat-10684	138	8	learning	learning	NOUN
easat-10684	138	9	gate	gate	PROPN
easat-10684	138	10	source	source	NOUN
easat-10684	138	11	credibility	credibility	NOUN
easat-10684	138	12	assessment	assessment	NOUN
easat-10684	138	13	.	.	PUNCT
easat-10684	139	1	this	this	DET
easat-10684	139	2	proposed	propose	VERB
easat-10684	139	3	model	model	NOUN
easat-10684	139	4	leverages	leverage	VERB
easat-10684	139	5	the	the	DET
easat-10684	139	6	structural	structural	ADJ
easat-10684	139	7	information	information	NOUN
easat-10684	139	8	of	of	ADP
easat-10684	139	9	news	news	NOUN
easat-10684	139	10	propagation	propagation	NOUN
easat-10684	139	11	networks	network	NOUN
easat-10684	139	12	to	to	PART
easat-10684	139	13	better	well	ADV
easat-10684	139	14	understand	understand	VERB
easat-10684	139	15	the	the	DET
easat-10684	139	16	connections	connection	NOUN
easat-10684	139	17	and	and	CCONJ
easat-10684	139	18	patterns	pattern	NOUN
easat-10684	139	19	indicative	indicative	ADJ
easat-10684	139	20	of	of	ADP
easat-10684	139	21	misinformation	misinformation	NOUN
easat-10684	139	22	.	.	PUNCT
easat-10684	140	1	sari	sari	NOUN
easat-10684	140	2	et	et	PROPN
easat-10684	140	3	al	al	PROPN
easat-10684	140	4	.	.	PUNCT
easat-10684	141	1	[	[	X
easat-10684	141	2	30	30	NUM
easat-10684	141	3	]	]	PUNCT
easat-10684	141	4	developed	develop	VERB
easat-10684	141	5	a	a	DET
easat-10684	141	6	fake	fake	ADJ
easat-10684	141	7	news	news	NOUN
easat-10684	141	8	detection	detection	NOUN
easat-10684	141	9	model	model	NOUN
easat-10684	141	10	utilizing	utilize	VERB
easat-10684	141	11	an	an	DET
easat-10684	141	12	enhanced	enhanced	ADJ
easat-10684	141	13	cnn	cnn	PROPN
easat-10684	141	14	-	-	PUNCT
easat-10684	141	15	bilstm	bilstm	NOUN
easat-10684	141	16	architecture	architecture	NOUN
easat-10684	141	17	combined	combine	VERB
easat-10684	141	18	with	with	ADP
easat-10684	141	19	glove	glove	NOUN
easat-10684	141	20	word	word	NOUN
easat-10684	141	21	embedding	embed	VERB
easat-10684	141	22	techniques	technique	NOUN
easat-10684	141	23	,	,	PUNCT
easat-10684	141	24	achieving	achieve	VERB
easat-10684	141	25	a	a	DET
easat-10684	141	26	high	high	ADJ
easat-10684	141	27	accuracy	accuracy	NOUN
easat-10684	141	28	of	of	ADP
easat-10684	141	29	96	96	NUM
easat-10684	141	30	%	%	NOUN
easat-10684	141	31	in	in	ADP
easat-10684	141	32	distinguishing	distinguish	VERB
easat-10684	141	33	fake	fake	ADJ
easat-10684	141	34	news	news	NOUN
easat-10684	141	35	from	from	ADP
easat-10684	141	36	real	real	ADJ
easat-10684	141	37	news	news	NOUN
easat-10684	141	38	.	.	PUNCT
easat-10684	142	1	this	this	PRON
easat-10684	142	2	demonstrates	demonstrate	VERB
easat-10684	142	3	the	the	DET
easat-10684	142	4	effectiveness	effectiveness	NOUN
easat-10684	142	5	of	of	ADP
easat-10684	142	6	the	the	DET
easat-10684	142	7	proposed	propose	VERB
easat-10684	142	8	model	model	NOUN
easat-10684	142	9	in	in	ADP
easat-10684	142	10	addressing	address	VERB
easat-10684	142	11	the	the	DET
easat-10684	142	12	challenges	challenge	NOUN
easat-10684	142	13	posed	pose	VERB
easat-10684	142	14	by	by	ADP
easat-10684	142	15	the	the	DET
easat-10684	142	16	spread	spread	NOUN
easat-10684	142	17	of	of	ADP
easat-10684	142	18	fake	fake	ADJ
easat-10684	142	19	news	news	NOUN
easat-10684	142	20	in	in	ADP
easat-10684	142	21	the	the	DET
easat-10684	142	22	digital	digital	ADJ
easat-10684	142	23	age	age	NOUN
easat-10684	142	24	.	.	PUNCT
easat-10684	143	1	the	the	DET
easat-10684	143	2	research	research	NOUN
easat-10684	143	3	highlights	highlight	VERB
easat-10684	143	4	the	the	DET
easat-10684	143	5	importance	importance	NOUN
easat-10684	143	6	of	of	ADP
easat-10684	143	7	training	training	NOUN
easat-10684	143	8	data	datum	NOUN
easat-10684	143	9	ratios	ratio	NOUN
easat-10684	143	10	,	,	PUNCT
easat-10684	143	11	showing	show	VERB
easat-10684	143	12	that	that	SCONJ
easat-10684	143	13	a	a	DET
easat-10684	143	14	larger	large	ADJ
easat-10684	143	15	training	training	NOUN
easat-10684	143	16	data	datum	NOUN
easat-10684	143	17	ratio	ratio	NOUN
easat-10684	143	18	leads	lead	VERB
easat-10684	143	19	to	to	ADP
easat-10684	143	20	better	well	ADJ
easat-10684	143	21	model	model	NOUN
easat-10684	143	22	performance	performance	NOUN
easat-10684	143	23	.	.	PUNCT
easat-10684	144	1	brinda	brinda	PROPN
easat-10684	144	2	et	et	PROPN
easat-10684	144	3	al	al	PROPN
easat-10684	144	4	.	.	PUNCT
easat-10684	145	1	[	[	X
easat-10684	145	2	31	31	NUM
easat-10684	145	3	]	]	PUNCT
easat-10684	145	4	proposed	propose	VERB
easat-10684	145	5	a	a	DET
easat-10684	145	6	heterogeneous	heterogeneous	ADJ
easat-10684	145	7	deep	deep	ADJ
easat-10684	145	8	learning	learning	NOUN
easat-10684	145	9	model	model	NOUN
easat-10684	145	10	that	that	PRON
easat-10684	145	11	enhances	enhance	VERB
easat-10684	145	12	fake	fake	ADJ
easat-10684	145	13	news	news	NOUN
easat-10684	145	14	detection	detection	NOUN
easat-10684	145	15	by	by	ADP
easat-10684	145	16	incorporating	incorporate	VERB
easat-10684	145	17	dissemination	dissemination	NOUN
easat-10684	145	18	structure	structure	NOUN
easat-10684	145	19	features	feature	NOUN
easat-10684	145	20	from	from	ADP
easat-10684	145	21	social	social	ADJ
easat-10684	145	22	networks	network	NOUN
easat-10684	145	23	,	,	PUNCT
easat-10684	145	24	such	such	ADJ
easat-10684	145	25	as	as	ADP
easat-10684	145	26	user	user	NOUN
easat-10684	145	27	interactions	interaction	NOUN
easat-10684	145	28	,	,	PUNCT
easat-10684	145	29	sharing	share	VERB
easat-10684	145	30	patterns	pattern	NOUN
easat-10684	145	31	,	,	PUNCT
easat-10684	145	32	and	and	CCONJ
easat-10684	145	33	network	network	NOUN
easat-10684	145	34	topology	topology	NOUN
easat-10684	145	35	,	,	PUNCT
easat-10684	145	36	which	which	PRON
easat-10684	145	37	provide	provide	VERB
easat-10684	145	38	deeper	deep	ADJ
easat-10684	145	39	insights	insight	NOUN
easat-10684	145	40	into	into	ADP
easat-10684	145	41	how	how	SCONJ
easat-10684	145	42	information	information	NOUN
easat-10684	145	43	spreads	spread	VERB
easat-10684	145	44	beyond	beyond	ADP
easat-10684	145	45	just	just	ADV
easat-10684	145	46	text	text	NOUN
easat-10684	145	47	analysis	analysis	NOUN
easat-10684	145	48	.	.	PUNCT
easat-10684	146	1	deepthi	deepthi	NOUN
easat-10684	146	2	and	and	CCONJ
easat-10684	146	3	shastry	shastry	NOUN
easat-10684	147	1	[	[	X
easat-10684	147	2	32	32	NUM
easat-10684	147	3	]	]	PUNCT
easat-10684	147	4	proposed	propose	VERB
easat-10684	147	5	an	an	DET
easat-10684	147	6	unsupervised	unsupervised	ADJ
easat-10684	147	7	misinformation	misinformation	NOUN
easat-10684	147	8	detection	detection	NOUN
easat-10684	147	9	technique	technique	NOUN
easat-10684	147	10	(	(	PUNCT
easat-10684	147	11	umd	umd	NOUN
easat-10684	147	12	)	)	PUNCT
easat-10684	147	13	that	that	PRON
easat-10684	147	14	achieves	achieve	VERB
easat-10684	147	15	an	an	DET
easat-10684	147	16	overall	overall	ADJ
easat-10684	147	17	accuracy	accuracy	NOUN
easat-10684	147	18	of	of	ADP
easat-10684	147	19	92.82	92.82	NUM
easat-10684	147	20	%	%	NOUN
easat-10684	147	21	,	,	PUNCT
easat-10684	147	22	outperforming	outperform	VERB
easat-10684	147	23	existing	exist	VERB
easat-10684	147	24	models	model	NOUN
easat-10684	147	25	such	such	ADJ
easat-10684	147	26	as	as	ADP
easat-10684	147	27	transformer	transformer	NOUN
easat-10684	147	28	and	and	CCONJ
easat-10684	147	29	hybrid	hybrid	ADJ
easat-10684	147	30	cnn	cnn	PROPN
easat-10684	147	31	-	-	PUNCT
easat-10684	147	32	rnn	rnn	NOUN
easat-10684	147	33	models	model	NOUN
easat-10684	147	34	by	by	ADP
easat-10684	147	35	approximately	approximately	ADV
easat-10684	147	36	3	3	NUM
easat-10684	147	37	%	%	NOUN
easat-10684	147	38	,	,	PUNCT
easat-10684	147	39	demonstrating	demonstrate	VERB
easat-10684	147	40	the	the	DET
easat-10684	147	41	effectiveness	effectiveness	NOUN
easat-10684	147	42	of	of	ADP
easat-10684	147	43	deep	deep	ADJ
easat-10684	147	44	learning	learning	NOUN
easat-10684	147	45	in	in	ADP
easat-10684	147	46	fake	fake	ADJ
easat-10684	147	47	news	news	NOUN
easat-10684	147	48	detection	detection	NOUN
easat-10684	147	49	.	.	PUNCT
easat-10684	148	1	zhu	zhu	X
easat-10684	149	1	[	[	X
easat-10684	149	2	33	33	NUM
easat-10684	149	3	]	]	PUNCT
easat-10684	149	4	highlighted	highlight	VERB
easat-10684	149	5	that	that	SCONJ
easat-10684	149	6	dl	dl	PROPN
easat-10684	149	7	models	model	NOUN
easat-10684	149	8	,	,	PUNCT
easat-10684	149	9	such	such	ADJ
easat-10684	149	10	as	as	ADP
easat-10684	149	11	opcnn	opcnn	NOUN
easat-10684	149	12	-	-	PUNCT
easat-10684	149	13	fake	fake	NOUN
easat-10684	149	14	,	,	PUNCT
easat-10684	149	15	dc	dc	PROPN
easat-10684	149	16	-	-	PUNCT
easat-10684	149	17	cnn	cnn	PROPN
easat-10684	149	18	,	,	PUNCT
easat-10684	149	19	and	and	CCONJ
easat-10684	149	20	cnn	cnn	PROPN
easat-10684	149	21	-	-	PUNCT
easat-10684	149	22	lstm	lstm	PROPN
easat-10684	149	23	,	,	PUNCT
easat-10684	149	24	show	show	VERB
easat-10684	149	25	significant	significant	ADJ
easat-10684	149	26	accuracy	accuracy	NOUN
easat-10684	149	27	advantages	advantage	NOUN
easat-10684	149	28	over	over	ADP
easat-10684	149	29	traditional	traditional	ADJ
easat-10684	149	30	machine	machine	NOUN
easat-10684	149	31	learning	learning	NOUN
easat-10684	149	32	methods	method	NOUN
easat-10684	149	33	in	in	ADP
easat-10684	149	34	fake	fake	ADJ
easat-10684	149	35	news	news	NOUN
easat-10684	149	36	detection	detection	NOUN
easat-10684	149	37	,	,	PUNCT
easat-10684	149	38	addressing	address	VERB
easat-10684	149	39	the	the	DET
easat-10684	149	40	limitations	limitation	NOUN
easat-10684	149	41	of	of	ADP
easat-10684	149	42	conventional	conventional	ADJ
easat-10684	149	43	approaches	approach	NOUN
easat-10684	149	44	in	in	ADP
easat-10684	149	45	handling	handle	VERB
easat-10684	149	46	expanding	expand	VERB
easat-10684	149	47	data	datum	NOUN
easat-10684	149	48	sets	set	NOUN
easat-10684	149	49	and	and	CCONJ
easat-10684	149	50	streams	stream	NOUN
easat-10684	149	51	.	.	PUNCT
easat-10684	150	1	ladouceur	ladouceur	PROPN
easat-10684	150	2	et	et	PROPN
easat-10684	150	3	al	al	PROPN
easat-10684	150	4	.	.	PUNCT
easat-10684	151	1	[	[	X
easat-10684	151	2	34	34	NUM
easat-10684	151	3	]	]	PUNCT
easat-10684	151	4	highlight	highlight	VERB
easat-10684	151	5	that	that	SCONJ
easat-10684	151	6	dl	dl	PROPN
easat-10684	151	7	algorithms	algorithms	PROPN
easat-10684	151	8	,	,	PUNCT
easat-10684	151	9	specifically	specifically	ADV
easat-10684	151	10	bert	bert	PROPN
easat-10684	151	11	and	and	CCONJ
easat-10684	151	12	lstm	lstm	PROPN
easat-10684	151	13	,	,	PUNCT
easat-10684	151	14	have	have	AUX
easat-10684	151	15	demonstrated	demonstrate	VERB
easat-10684	151	16	superior	superior	ADJ
easat-10684	151	17	performance	performance	NOUN
easat-10684	151	18	in	in	ADP
easat-10684	151	19	fake	fake	ADJ
easat-10684	151	20	news	news	NOUN
easat-10684	151	21	detection	detection	NOUN
easat-10684	151	22	compared	compare	VERB
easat-10684	151	23	to	to	ADP
easat-10684	151	24	traditional	traditional	ADJ
easat-10684	151	25	methods	method	NOUN
easat-10684	151	26	,	,	PUNCT
easat-10684	151	27	particularly	particularly	ADV
easat-10684	151	28	when	when	SCONJ
easat-10684	151	29	incorporating	incorporate	VERB
easat-10684	151	30	context	context	NOUN
easat-10684	151	31	features	feature	NOUN
easat-10684	151	32	,	,	PUNCT
easat-10684	151	33	achieving	achieve	VERB
easat-10684	151	34	better	well	ADJ
easat-10684	151	35	accuracy	accuracy	NOUN
easat-10684	151	36	in	in	ADP
easat-10684	151	37	classifying	classify	VERB
easat-10684	151	38	news	news	NOUN
easat-10684	151	39	from	from	ADP
easat-10684	151	40	the	the	DET
easat-10684	151	41	covid-19	covid-19	PROPN
easat-10684	151	42	dataset	dataset	NOUN
easat-10684	151	43	.	.	PUNCT
easat-10684	152	1	porya	porya	PROPN
easat-10684	152	2	et	et	PROPN
easat-10684	152	3	al	al	PROPN
easat-10684	152	4	.	.	PUNCT
easat-10684	153	1	[	[	X
easat-10684	153	2	35	35	NUM
easat-10684	153	3	]	]	PUNCT
easat-10684	153	4	investigated	investigate	VERB
easat-10684	153	5	various	various	ADJ
easat-10684	153	6	models	model	NOUN
easat-10684	153	7	,	,	PUNCT
easat-10684	153	8	including	include	VERB
easat-10684	153	9	deep	deep	ADJ
easat-10684	153	10	learning	learning	NOUN
easat-10684	153	11	algorithms	algorithm	NOUN
easat-10684	153	12	such	such	ADJ
easat-10684	153	13	as	as	ADP
easat-10684	153	14	cnn	cnn	PROPN
easat-10684	153	15	and	and	CCONJ
easat-10684	153	16	lstm	lstm	PROPN
easat-10684	153	17	,	,	PUNCT
easat-10684	153	18	alongside	alongside	ADP
easat-10684	153	19	conventional	conventional	ADJ
easat-10684	153	20	methods	method	NOUN
easat-10684	153	21	like	like	ADP
easat-10684	153	22	rf	rf	NOUN
easat-10684	153	23	and	and	CCONJ
easat-10684	153	24	gb	gb	NOUN
easat-10684	153	25	.	.	PUNCT
easat-10684	154	1	it	it	PRON
easat-10684	154	2	highlights	highlight	VERB
easat-10684	154	3	challenges	challenge	NOUN
easat-10684	154	4	in	in	ADP
easat-10684	154	5	achieving	achieve	VERB
easat-10684	154	6	effective	effective	ADJ
easat-10684	154	7	fake	fake	ADJ
easat-10684	154	8	news	news	NOUN
easat-10684	154	9	detection	detection	NOUN
easat-10684	154	10	,	,	PUNCT
easat-10684	154	11	indicating	indicate	VERB
easat-10684	154	12	that	that	SCONJ
easat-10684	154	13	robust	robust	ADJ
easat-10684	154	14	performance	performance	NOUN
easat-10684	154	15	remains	remain	VERB
easat-10684	154	16	elusive	elusive	ADJ
easat-10684	154	17	despite	despite	SCONJ
easat-10684	154	18	advancements	advancement	NOUN
easat-10684	154	19	.	.	PUNCT
easat-10684	155	1	traditional	traditional	ADJ
easat-10684	155	2	ml	ml	ADV
easat-10684	155	3	-	-	PUNCT
easat-10684	155	4	based	base	VERB
easat-10684	155	5	models	model	NOUN
easat-10684	155	6	often	often	ADV
easat-10684	155	7	require	require	VERB
easat-10684	155	8	extensive	extensive	ADJ
easat-10684	155	9	labeled	label	VERB
easat-10684	155	10	datasets	dataset	NOUN
easat-10684	155	11	,	,	PUNCT
easat-10684	155	12	which	which	PRON
easat-10684	155	13	can	can	AUX
easat-10684	155	14	be	be	AUX
easat-10684	155	15	costly	costly	ADJ
easat-10684	155	16	and	and	CCONJ
easat-10684	155	17	time	time	NOUN
easat-10684	155	18	-	-	PUNCT
easat-10684	155	19	consuming	consume	VERB
easat-10684	155	20	to	to	PART
easat-10684	155	21	attain	attain	VERB
easat-10684	155	22	[	[	X
easat-10684	155	23	36	36	NUM
easat-10684	155	24	]	]	PUNCT
easat-10684	155	25	.	.	PUNCT
easat-10684	156	1	to	to	PART
easat-10684	156	2	address	address	VERB
easat-10684	156	3	this	this	PRON
easat-10684	156	4	,	,	PUNCT
easat-10684	156	5	transfer	transfer	NOUN
easat-10684	156	6	learning	learning	NOUN
easat-10684	156	7	has	have	AUX
easat-10684	156	8	emerged	emerge	VERB
easat-10684	156	9	as	as	ADP
easat-10684	156	10	a	a	DET
easat-10684	156	11	promising	promising	ADJ
easat-10684	156	12	approach	approach	NOUN
easat-10684	156	13	,	,	PUNCT
easat-10684	156	14	leveraging	leverage	VERB
easat-10684	156	15	pre	pre	ADJ
easat-10684	156	16	-	-	ADJ
easat-10684	156	17	trained	train	VERB
easat-10684	156	18	models	model	NOUN
easat-10684	156	19	to	to	PART
easat-10684	156	20	adapt	adapt	VERB
easat-10684	156	21	to	to	ADP
easat-10684	156	22	new	new	ADJ
easat-10684	156	23	tasks	task	NOUN
easat-10684	156	24	with	with	ADP
easat-10684	156	25	minimal	minimal	ADJ
easat-10684	156	26	data	datum	NOUN
easat-10684	156	27	requirements	requirement	NOUN
easat-10684	156	28	.	.	PUNCT
easat-10684	157	1	by	by	ADP
easat-10684	157	2	utilizing	utilize	VERB
easat-10684	157	3	pretrained	pretraine	VERB
easat-10684	157	4	language	language	NOUN
easat-10684	157	5	models	model	NOUN
easat-10684	157	6	like	like	ADP
easat-10684	157	7	bert	bert	PROPN
easat-10684	157	8	,	,	PUNCT
easat-10684	157	9	roberta	roberta	PROPN
easat-10684	157	10	,	,	PUNCT
easat-10684	157	11	or	or	CCONJ
easat-10684	157	12	xlnet	xlnet	NOUN
easat-10684	157	13	,	,	PUNCT
easat-10684	157	14	transfer	transfer	NOUN
easat-10684	157	15	learning	learning	NOUN
easat-10684	157	16	enables	enable	VERB
easat-10684	157	17	the	the	DET
easat-10684	157	18	extraction	extraction	NOUN
easat-10684	157	19	of	of	ADP
easat-10684	157	20	deep	deep	ADJ
easat-10684	157	21	semantic	semantic	ADJ
easat-10684	157	22	features	feature	NOUN
easat-10684	157	23	,	,	PUNCT
easat-10684	157	24	enhancing	enhance	VERB
easat-10684	157	25	the	the	DET
easat-10684	157	26	accuracy	accuracy	NOUN
easat-10684	157	27	and	and	CCONJ
easat-10684	157	28	efficiency	efficiency	NOUN
easat-10684	157	29	of	of	ADP
easat-10684	157	30	fake	fake	ADJ
easat-10684	157	31	news	news	NOUN
easat-10684	157	32	detection	detection	NOUN
easat-10684	157	33	systems	system	NOUN
easat-10684	157	34	[	[	X
easat-10684	157	35	19	19	NUM
easat-10684	157	36	,	,	PUNCT
easat-10684	157	37	37	37	NUM
easat-10684	157	38	]	]	PUNCT
easat-10684	157	39	.	.	PUNCT
easat-10684	158	1	recent	recent	ADJ
easat-10684	158	2	studies	study	NOUN
easat-10684	158	3	demonstrate	demonstrate	VERB
easat-10684	158	4	that	that	SCONJ
easat-10684	158	5	transfer	transfer	NOUN
easat-10684	158	6	learning	learning	NOUN
easat-10684	158	7	models	model	NOUN
easat-10684	158	8	can	can	AUX
easat-10684	158	9	outperform	outperform	VERB
easat-10684	158	10	conventional	conventional	ADJ
easat-10684	158	11	methods	method	NOUN
easat-10684	158	12	,	,	PUNCT
easat-10684	158	13	achieving	achieve	VERB
easat-10684	158	14	robust	robust	ADJ
easat-10684	158	15	performance	performance	NOUN
easat-10684	158	16	across	across	ADP
easat-10684	158	17	diverse	diverse	ADJ
easat-10684	158	18	fake	fake	ADJ
easat-10684	158	19	news	news	NOUN
easat-10684	158	20	datasets	dataset	NOUN
easat-10684	158	21	[	[	X
easat-10684	158	22	38	38	NUM
easat-10684	158	23	-	-	SYM
easat-10684	158	24	42	42	NUM
easat-10684	158	25	]	]	PUNCT
easat-10684	158	26	.	.	PUNCT
easat-10684	159	1	this	this	DET
easat-10684	159	2	section	section	NOUN
easat-10684	159	3	explores	explore	VERB
easat-10684	159	4	the	the	DET
easat-10684	159	5	latest	late	ADJ
easat-10684	159	6	advancements	advancement	NOUN
easat-10684	159	7	in	in	ADP
easat-10684	159	8	transfer	transfer	NOUN
easat-10684	159	9	learning	learning	NOUN
easat-10684	159	10	for	for	ADP
easat-10684	159	11	fake	fake	ADJ
easat-10684	159	12	news	news	NOUN
easat-10684	159	13	detection	detection	NOUN
easat-10684	159	14	,	,	PUNCT
easat-10684	159	15	with	with	ADP
easat-10684	159	16	a	a	DET
easat-10684	159	17	focus	focus	NOUN
easat-10684	159	18	on	on	ADP
easat-10684	159	19	model	model	NOUN
easat-10684	159	20	architectures	architecture	NOUN
easat-10684	159	21	,	,	PUNCT
easat-10684	159	22	dataset	dataset	ADJ
easat-10684	159	23	utilization	utilization	NOUN
easat-10684	159	24	,	,	PUNCT
easat-10684	159	25	and	and	CCONJ
easat-10684	159	26	cross	cross	ADJ
easat-10684	159	27	-	-	ADJ
easat-10684	159	28	domain	domain	ADJ
easat-10684	159	29	adaptability	adaptability	NOUN
easat-10684	159	30	,	,	PUNCT
easat-10684	159	31	drawing	draw	VERB
easat-10684	159	32	insights	insight	NOUN
easat-10684	159	33	from	from	ADP
easat-10684	159	34	recent	recent	ADJ
easat-10684	159	35	research	research	NOUN
easat-10684	159	36	in	in	ADP
easat-10684	159	37	this	this	DET
easat-10684	159	38	dynamic	dynamic	ADJ
easat-10684	159	39	field	field	NOUN
easat-10684	159	40	.	.	PUNCT
easat-10684	160	1	1520	1520	NUM
easat-10684	160	2	edelweiss	edelweiss	PROPN
easat-10684	160	3	applied	apply	VERB
easat-10684	160	4	science	science	NOUN
easat-10684	160	5	and	and	CCONJ
easat-10684	160	6	technology	technology	NOUN
easat-10684	160	7	issn	issn	PROPN
easat-10684	160	8	:	:	PUNCT
easat-10684	160	9	2576	2576	NUM
easat-10684	160	10	-	-	SYM
easat-10684	160	11	8484	8484	NUM
easat-10684	160	12	vol	vol	NOUN
easat-10684	160	13	.	.	PROPN
easat-10684	161	1	9	9	NUM
easat-10684	161	2	,	,	PUNCT
easat-10684	161	3	no	no	INTJ
easat-10684	161	4	.	.	NOUN
easat-10684	161	5	10	10	NUM
easat-10684	161	6	:	:	PUNCT
easat-10684	161	7	1515	1515	NUM
easat-10684	161	8	-	-	SYM
easat-10684	161	9	1540	1540	NUM
easat-10684	161	10	,	,	PUNCT
easat-10684	161	11	2025	2025	NUM
easat-10684	161	12	doi	doi	NOUN
easat-10684	161	13	:	:	PUNCT
easat-10684	161	14	10.55214/2576	10.55214/2576	NUM
easat-10684	161	15	-	-	SYM
easat-10684	161	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	162	1	©	©	PROPN
easat-10684	162	2	2025	2025	NUM
easat-10684	162	3	by	by	ADP
easat-10684	162	4	the	the	DET
easat-10684	162	5	authors	author	NOUN
easat-10684	162	6	;	;	PUNCT
easat-10684	162	7	licensee	licensee	PROPN
easat-10684	162	8	learning	learn	VERB
easat-10684	162	9	gate	gate	NOUN
easat-10684	162	10	table	table	NOUN
easat-10684	162	11	1	1	NUM
easat-10684	162	12	.	.	PUNCT
easat-10684	162	13	summarised	summarise	VERB
easat-10684	162	14	the	the	DET
easat-10684	162	15	related	related	ADJ
easat-10684	162	16	work	work	NOUN
easat-10684	162	17	in	in	ADP
easat-10684	162	18	this	this	DET
easat-10684	162	19	study	study	NOUN
easat-10684	162	20	authors	author	NOUN
easat-10684	162	21	methods	method	VERB
easat-10684	162	22	dataset	dataset	VERB
easat-10684	162	23	contributions	contribution	NOUN
easat-10684	162	24	research	research	NOUN
easat-10684	162	25	gap	gap	NOUN
easat-10684	162	26	roumeliotis	roumeliotis	INTJ
easat-10684	162	27	et	et	PROPN
easat-10684	162	28	al	al	PROPN
easat-10684	162	29	.	.	PUNCT
easat-10684	163	1	[	[	X
easat-10684	163	2	17	17	NUM
easat-10684	163	3	]	]	X
easat-10684	163	4	cnn	cnn	PROPN
easat-10684	163	5	,	,	PUNCT
easat-10684	163	6	bert	bert	PROPN
easat-10684	163	7	,	,	PUNCT
easat-10684	163	8	and	and	CCONJ
easat-10684	163	9	gpt	gpt	NOUN
easat-10684	163	10	self	self	NOUN
easat-10684	163	11	-	-	PUNCT
easat-10684	163	12	dataset	dataset	VERB
easat-10684	163	13	the	the	DET
easat-10684	163	14	research	research	NOUN
easat-10684	163	15	addresses	address	VERB
easat-10684	163	16	the	the	DET
easat-10684	163	17	growing	grow	VERB
easat-10684	163	18	need	need	NOUN
easat-10684	163	19	for	for	ADP
easat-10684	163	20	scalable	scalable	ADJ
easat-10684	163	21	and	and	CCONJ
easat-10684	163	22	unbiased	unbiased	ADJ
easat-10684	163	23	tools	tool	NOUN
easat-10684	163	24	to	to	PART
easat-10684	163	25	identify	identify	VERB
easat-10684	163	26	fake	fake	ADJ
easat-10684	163	27	news	news	NOUN
easat-10684	163	28	across	across	ADP
easat-10684	163	29	diverse	diverse	ADJ
easat-10684	163	30	platforms	platform	NOUN
easat-10684	163	31	and	and	CCONJ
easat-10684	163	32	languages	language	NOUN
easat-10684	163	33	,	,	PUNCT
easat-10684	163	34	which	which	PRON
easat-10684	163	35	could	could	AUX
easat-10684	163	36	assist	assist	VERB
easat-10684	163	37	journalists	journalist	NOUN
easat-10684	163	38	,	,	PUNCT
easat-10684	163	39	social	social	ADJ
easat-10684	163	40	media	medium	NOUN
easat-10684	163	41	platforms	platform	NOUN
easat-10684	163	42	,	,	PUNCT
easat-10684	163	43	and	and	CCONJ
easat-10684	163	44	policymakers	policymaker	NOUN
easat-10684	163	45	in	in	ADP
easat-10684	163	46	mitigating	mitigate	VERB
easat-10684	163	47	the	the	DET
easat-10684	163	48	spread	spread	NOUN
easat-10684	163	49	of	of	ADP
easat-10684	163	50	misinformation	misinformation	NOUN
easat-10684	163	51	.	.	PUNCT
easat-10684	164	1	the	the	DET
easat-10684	164	2	study	study	NOUN
easat-10684	164	3	does	do	AUX
easat-10684	164	4	not	not	PART
easat-10684	164	5	explicitly	explicitly	ADV
easat-10684	164	6	address	address	VERB
easat-10684	164	7	the	the	DET
easat-10684	164	8	limitations	limitation	NOUN
easat-10684	164	9	or	or	CCONJ
easat-10684	164	10	challenges	challenge	NOUN
easat-10684	164	11	faced	face	VERB
easat-10684	164	12	by	by	ADP
easat-10684	164	13	each	each	DET
easat-10684	164	14	ml	ml	PROPN
easat-10684	164	15	model	model	NOUN
easat-10684	164	16	(	(	PUNCT
easat-10684	164	17	cnn	cnn	PROPN
easat-10684	164	18	,	,	PUNCT
easat-10684	164	19	bert	bert	PROPN
easat-10684	164	20	,	,	PUNCT
easat-10684	164	21	and	and	CCONJ
easat-10684	164	22	gpt	gpt	NOUN
easat-10684	164	23	)	)	PUNCT
easat-10684	164	24	in	in	ADP
easat-10684	164	25	the	the	DET
easat-10684	164	26	context	context	NOUN
easat-10684	164	27	of	of	ADP
easat-10684	164	28	fake	fake	ADJ
easat-10684	164	29	news	news	NOUN
easat-10684	164	30	.	.	PUNCT
easat-10684	165	1	almandouh	almandouh	VERB
easat-10684	165	2	et	et	PROPN
easat-10684	165	3	al	al	PROPN
easat-10684	165	4	.	.	PUNCT
easat-10684	166	1	[	[	X
easat-10684	166	2	19	19	NUM
easat-10684	166	3	]	]	X
easat-10684	166	4	cnn	cnn	PROPN
easat-10684	166	5	-	-	PUNCT
easat-10684	166	6	lstm	lstm	PROPN
easat-10684	166	7	,	,	PUNCT
easat-10684	166	8	rnn	rnn	PROPN
easat-10684	166	9	-	-	PUNCT
easat-10684	166	10	cnn	cnn	PROPN
easat-10684	166	11	,	,	PUNCT
easat-10684	166	12	rnn	rnn	NOUN
easat-10684	166	13	-	-	PUNCT
easat-10684	166	14	lstm	lstm	NOUN
easat-10684	166	15	,	,	PUNCT
easat-10684	166	16	and	and	CCONJ
easat-10684	166	17	bi	bi	ADJ
easat-10684	166	18	-	-	ADJ
easat-10684	166	19	grubi	grubi	ADJ
easat-10684	166	20	-	-	PUNCT
easat-10684	166	21	lstm	lstm	ADJ
easat-10684	166	22	afnd	afnd	NOUN
easat-10684	166	23	and	and	CCONJ
easat-10684	166	24	arabicfak	arabicfak	NOUN
easat-10684	166	25	etweets	etweet	VERB
easat-10684	166	26	the	the	DET
easat-10684	166	27	research	research	NOUN
easat-10684	166	28	integrates	integrate	VERB
easat-10684	166	29	fasttext	fasttext	NOUN
easat-10684	166	30	word	word	NOUN
easat-10684	166	31	embeddings	embedding	NOUN
easat-10684	166	32	with	with	ADP
easat-10684	166	33	various	various	ADJ
easat-10684	166	34	ml	ml	NOUN
easat-10684	166	35	and	and	CCONJ
easat-10684	166	36	dl	dl	PROPN
easat-10684	166	37	methods	method	NOUN
easat-10684	166	38	,	,	PUNCT
easat-10684	166	39	and	and	CCONJ
easat-10684	166	40	employs	employ	VERB
easat-10684	166	41	advanced	advanced	ADJ
easat-10684	166	42	transformer	transformer	NOUN
easat-10684	166	43	-	-	PUNCT
easat-10684	166	44	based	base	VERB
easat-10684	166	45	models	model	NOUN
easat-10684	166	46	,	,	PUNCT
easat-10684	166	47	optimizing	optimize	VERB
easat-10684	166	48	their	their	PRON
easat-10684	166	49	performance	performance	NOUN
easat-10684	166	50	through	through	ADP
easat-10684	166	51	hyperparameter	hyperparameter	NOUN
easat-10684	166	52	tuning	tune	VERB
easat-10684	166	53	for	for	ADP
easat-10684	166	54	effective	effective	ADJ
easat-10684	166	55	arabic	arabic	ADJ
easat-10684	166	56	fake	fake	ADJ
easat-10684	166	57	news	news	NOUN
easat-10684	166	58	detection	detection	NOUN
easat-10684	166	59	.	.	PUNCT
easat-10684	167	1	the	the	DET
easat-10684	167	2	research	research	NOUN
easat-10684	167	3	primarily	primarily	ADV
easat-10684	167	4	focuses	focus	VERB
easat-10684	167	5	on	on	ADP
easat-10684	167	6	arabic	arabic	ADJ
easat-10684	167	7	fake	fake	ADJ
easat-10684	167	8	news	news	NOUN
easat-10684	167	9	detection	detection	NOUN
easat-10684	167	10	,	,	PUNCT
easat-10684	167	11	indicating	indicate	VERB
easat-10684	167	12	a	a	DET
easat-10684	167	13	gap	gap	NOUN
easat-10684	167	14	in	in	ADP
easat-10684	167	15	the	the	DET
easat-10684	167	16	application	application	NOUN
easat-10684	167	17	of	of	ADP
easat-10684	167	18	the	the	DET
easat-10684	167	19	developed	develop	VERB
easat-10684	167	20	models	model	NOUN
easat-10684	167	21	to	to	ADP
easat-10684	167	22	other	other	ADJ
easat-10684	167	23	languages	language	NOUN
easat-10684	167	24	.	.	PUNCT
easat-10684	168	1	future	future	ADJ
easat-10684	168	2	research	research	NOUN
easat-10684	168	3	could	could	AUX
easat-10684	168	4	expand	expand	VERB
easat-10684	168	5	the	the	DET
easat-10684	168	6	methodologies	methodology	NOUN
easat-10684	168	7	and	and	CCONJ
easat-10684	168	8	models	model	NOUN
easat-10684	168	9	to	to	PART
easat-10684	168	10	accommodate	accommodate	VERB
easat-10684	168	11	multilingual	multilingual	ADJ
easat-10684	168	12	fake	fake	ADJ
easat-10684	168	13	news	news	NOUN
easat-10684	168	14	detection	detection	NOUN
easat-10684	168	15	,	,	PUNCT
easat-10684	168	16	addressing	address	VERB
easat-10684	168	17	the	the	DET
easat-10684	168	18	global	global	ADJ
easat-10684	168	19	nature	nature	NOUN
easat-10684	168	20	of	of	ADP
easat-10684	168	21	misinformation	misinformation	NOUN
easat-10684	168	22	.	.	PUNCT
easat-10684	169	1	al	al	PROPN
easat-10684	169	2	-	-	PUNCT
easat-10684	169	3	alshaqi	alshaqi	PROPN
easat-10684	169	4	et	et	PROPN
easat-10684	169	5	al	al	PROPN
easat-10684	169	6	.	.	PUNCT
easat-10684	170	1	[	[	X
easat-10684	170	2	20	20	NUM
easat-10684	170	3	]	]	SYM
easat-10684	170	4	ml	ml	NOUN
easat-10684	170	5	-	-	PUNCT
easat-10684	170	6	based	base	VERB
easat-10684	170	7	model	model	NOUN
easat-10684	170	8	,	,	PUNCT
easat-10684	170	9	bert+cnn	bert+cnn	PROPN
easat-10684	170	10	isot	isot	NOUN
easat-10684	170	11	&	&	CCONJ
easat-10684	170	12	mediaeval	mediaeval	ADJ
easat-10684	170	13	2016	2016	NUM
easat-10684	170	14	the	the	DET
easat-10684	170	15	study	study	NOUN
easat-10684	170	16	proposes	propose	VERB
easat-10684	170	17	a	a	DET
easat-10684	170	18	comprehensive	comprehensive	ADJ
easat-10684	170	19	framework	framework	NOUN
easat-10684	170	20	for	for	ADP
easat-10684	170	21	fake	fake	ADJ
easat-10684	170	22	news	news	NOUN
easat-10684	170	23	detection	detection	NOUN
easat-10684	170	24	that	that	PRON
easat-10684	170	25	integrates	integrate	VERB
easat-10684	170	26	multiple	multiple	ADJ
easat-10684	170	27	modalities	modality	NOUN
easat-10684	170	28	,	,	PUNCT
easat-10684	170	29	including	include	VERB
easat-10684	170	30	text	text	NOUN
easat-10684	170	31	,	,	PUNCT
easat-10684	170	32	images	image	NOUN
easat-10684	170	33	,	,	PUNCT
easat-10684	170	34	and	and	CCONJ
easat-10684	170	35	videos	video	NOUN
easat-10684	170	36	,	,	PUNCT
easat-10684	170	37	utilizing	utilize	VERB
easat-10684	170	38	both	both	PRON
easat-10684	170	39	ml	ml	NOUN
easat-10684	170	40	and	and	CCONJ
easat-10684	170	41	dl	dl	PROPN
easat-10684	170	42	techniques	technique	NOUN
easat-10684	170	43	.	.	PUNCT
easat-10684	171	1	this	this	DET
easat-10684	171	2	approach	approach	NOUN
easat-10684	171	3	addresses	address	VERB
easat-10684	171	4	the	the	DET
easat-10684	171	5	complexities	complexity	NOUN
easat-10684	171	6	of	of	ADP
easat-10684	171	7	modern	modern	ADJ
easat-10684	171	8	information	information	NOUN
easat-10684	171	9	dissemination	dissemination	NOUN
easat-10684	171	10	and	and	CCONJ
easat-10684	171	11	manipulation	manipulation	NOUN
easat-10684	171	12	.	.	PUNCT
easat-10684	172	1	while	while	SCONJ
easat-10684	172	2	the	the	DET
easat-10684	172	3	random	random	ADJ
easat-10684	172	4	forest	forest	NOUN
easat-10684	172	5	classifier	classifier	NOUN
easat-10684	172	6	achieved	achieve	VERB
easat-10684	172	7	high	high	ADJ
easat-10684	172	8	accuracy	accuracy	NOUN
easat-10684	172	9	for	for	ADP
easat-10684	172	10	textual	textual	ADJ
easat-10684	172	11	data	datum	NOUN
easat-10684	172	12	,	,	PUNCT
easat-10684	172	13	the	the	DET
easat-10684	172	14	research	research	NOUN
easat-10684	172	15	does	do	AUX
easat-10684	172	16	not	not	PART
easat-10684	172	17	explore	explore	VERB
easat-10684	172	18	the	the	DET
easat-10684	172	19	limitations	limitation	NOUN
easat-10684	172	20	or	or	CCONJ
easat-10684	172	21	potential	potential	ADJ
easat-10684	172	22	biases	bias	NOUN
easat-10684	172	23	of	of	ADP
easat-10684	172	24	the	the	DET
easat-10684	172	25	classifiers	classifier	NOUN
easat-10684	172	26	used	use	VERB
easat-10684	172	27	,	,	PUNCT
easat-10684	172	28	nor	nor	CCONJ
easat-10684	172	29	does	do	AUX
easat-10684	172	30	it	it	PRON
easat-10684	172	31	provide	provide	VERB
easat-10684	172	32	insights	insight	NOUN
easat-10684	172	33	into	into	ADP
easat-10684	172	34	the	the	DET
easat-10684	172	35	interpretability	interpretability	NOUN
easat-10684	172	36	of	of	ADP
easat-10684	172	37	the	the	DET
easat-10684	172	38	models	model	NOUN
easat-10684	172	39	,	,	PUNCT
easat-10684	172	40	which	which	PRON
easat-10684	172	41	are	be	AUX
easat-10684	172	42	critical	critical	ADJ
easat-10684	172	43	for	for	ADP
easat-10684	172	44	understanding	understand	VERB
easat-10684	172	45	the	the	DET
easat-10684	172	46	decisionmaking	decisionmake	VERB
easat-10684	172	47	process	process	NOUN
easat-10684	172	48	in	in	ADP
easat-10684	172	49	fake	fake	ADJ
easat-10684	172	50	news	news	NOUN
easat-10684	172	51	detection	detection	NOUN
easat-10684	172	52	.	.	PUNCT
easat-10684	173	1	al	al	PROPN
easat-10684	173	2	-	-	PUNCT
easat-10684	173	3	tarawneh	tarawneh	PROPN
easat-10684	173	4	et	et	PROPN
easat-10684	173	5	al	al	PROPN
easat-10684	173	6	.	.	PUNCT
easat-10684	174	1	[	[	X
easat-10684	174	2	21	21	NUM
easat-10684	174	3	]	]	X
easat-10684	174	4	svms	svms	NOUN
easat-10684	174	5	,	,	PUNCT
easat-10684	174	6	mlps	mlp	NOUN
easat-10684	174	7	,	,	PUNCT
easat-10684	174	8	&	&	CCONJ
easat-10684	174	9	cnns	cnns	PROPN
easat-10684	174	10	truthseeker	truthseeker	PROPN
easat-10684	175	1	the	the	DET
easat-10684	175	2	study	study	NOUN
easat-10684	175	3	emphasizes	emphasize	VERB
easat-10684	175	4	the	the	DET
easat-10684	175	5	importance	importance	NOUN
easat-10684	175	6	of	of	ADP
easat-10684	175	7	selecting	select	VERB
easat-10684	175	8	appropriate	appropriate	ADJ
easat-10684	175	9	embedding	embed	VERB
easat-10684	175	10	techniques	technique	NOUN
easat-10684	175	11	based	base	VERB
easat-10684	175	12	on	on	ADP
easat-10684	175	13	model	model	NOUN
easat-10684	175	14	architecture	architecture	NOUN
easat-10684	175	15	to	to	PART
easat-10684	175	16	maximize	maximize	VERB
easat-10684	175	17	fake	fake	ADJ
easat-10684	175	18	news	news	NOUN
easat-10684	175	19	detection	detection	NOUN
easat-10684	175	20	performance	performance	NOUN
easat-10684	175	21	.	.	PUNCT
easat-10684	176	1	the	the	DET
easat-10684	176	2	study	study	NOUN
easat-10684	176	3	suggests	suggest	VERB
easat-10684	176	4	future	future	ADJ
easat-10684	176	5	research	research	NOUN
easat-10684	176	6	directions	direction	NOUN
easat-10684	176	7	,	,	PUNCT
easat-10684	176	8	such	such	ADJ
easat-10684	176	9	as	as	ADP
easat-10684	176	10	integrating	integrate	VERB
easat-10684	176	11	contextual	contextual	ADJ
easat-10684	176	12	embeddings	embedding	NOUN
easat-10684	176	13	and	and	CCONJ
easat-10684	176	14	exploring	explore	VERB
easat-10684	176	15	hybrid	hybrid	ADJ
easat-10684	176	16	model	model	NOUN
easat-10684	176	17	architectures	architecture	NOUN
easat-10684	176	18	to	to	PART
easat-10684	176	19	further	far	ADV
easat-10684	176	20	improve	improve	VERB
easat-10684	176	21	detection	detection	NOUN
easat-10684	176	22	capabilities	capability	NOUN
easat-10684	176	23	against	against	ADP
easat-10684	176	24	misinformation	misinformation	NOUN
easat-10684	176	25	.	.	PUNCT
easat-10684	177	1	kumari	kumari	PROPN
easat-10684	177	2	and	and	CCONJ
easat-10684	177	3	singh	singh	PROPN
easat-10684	177	4	[	[	X
easat-10684	177	5	28	28	NUM
easat-10684	177	6	]	]	X
easat-10684	177	7	nlp	nlp	NOUN
easat-10684	177	8	,	,	PUNCT
easat-10684	177	9	lstm	lstm	NOUN
easat-10684	177	10	,	,	PUNCT
easat-10684	177	11	&	&	CCONJ
easat-10684	177	12	clip	clip	PROPN
easat-10684	177	13	fakeddit	fakeddit	VERB
easat-10684	177	14	the	the	DET
easat-10684	177	15	study	study	NOUN
easat-10684	177	16	introduces	introduce	VERB
easat-10684	177	17	a	a	DET
easat-10684	177	18	comprehensive	comprehensive	ADJ
easat-10684	177	19	multimodal	multimodal	NOUN
easat-10684	177	20	framework	framework	NOUN
easat-10684	177	21	for	for	ADP
easat-10684	177	22	fake	fake	ADJ
easat-10684	177	23	news	news	NOUN
easat-10684	177	24	detection	detection	NOUN
easat-10684	177	25	that	that	PRON
easat-10684	177	26	integrates	integrate	VERB
easat-10684	177	27	advanced	advanced	ADJ
easat-10684	177	28	techniques	technique	NOUN
easat-10684	177	29	to	to	PART
easat-10684	177	30	analyze	analyze	VERB
easat-10684	177	31	both	both	PRON
easat-10684	177	32	multilingual	multilingual	ADJ
easat-10684	177	33	text	text	NOUN
easat-10684	177	34	and	and	CCONJ
easat-10684	177	35	visual	visual	ADJ
easat-10684	177	36	data	datum	NOUN
easat-10684	177	37	,	,	PUNCT
easat-10684	177	38	addressing	address	VERB
easat-10684	177	39	the	the	DET
easat-10684	177	40	limitations	limitation	NOUN
easat-10684	177	41	of	of	ADP
easat-10684	177	42	existing	exist	VERB
easat-10684	177	43	methods	method	NOUN
easat-10684	177	44	that	that	PRON
easat-10684	177	45	primarily	primarily	ADV
easat-10684	177	46	focus	focus	VERB
easat-10684	177	47	on	on	ADP
easat-10684	177	48	linguistic	linguistic	ADJ
easat-10684	177	49	characteristics	characteristic	NOUN
easat-10684	177	50	.	.	PUNCT
easat-10684	178	1	the	the	DET
easat-10684	178	2	paper	paper	NOUN
easat-10684	178	3	highlights	highlight	VERB
easat-10684	178	4	the	the	DET
easat-10684	178	5	need	need	NOUN
easat-10684	178	6	for	for	ADP
easat-10684	178	7	a	a	DET
easat-10684	178	8	more	more	ADV
easat-10684	178	9	efficient	efficient	ADJ
easat-10684	178	10	model	model	NOUN
easat-10684	178	11	that	that	PRON
easat-10684	178	12	integrates	integrate	VERB
easat-10684	178	13	both	both	PRON
easat-10684	178	14	text	text	NOUN
easat-10684	178	15	and	and	CCONJ
easat-10684	178	16	visual	visual	ADJ
easat-10684	178	17	data	datum	NOUN
easat-10684	178	18	analysis	analysis	NOUN
easat-10684	178	19	,	,	PUNCT
easat-10684	178	20	suggesting	suggest	VERB
easat-10684	178	21	that	that	SCONJ
easat-10684	178	22	current	current	ADJ
easat-10684	178	23	approaches	approach	NOUN
easat-10684	178	24	may	may	AUX
easat-10684	178	25	lack	lack	VERB
easat-10684	178	26	a	a	DET
easat-10684	178	27	comprehensive	comprehensive	ADJ
easat-10684	178	28	multimodal	multimodal	NOUN
easat-10684	178	29	framework	framework	NOUN
easat-10684	178	30	that	that	PRON
easat-10684	178	31	leverages	leverage	VERB
easat-10684	178	32	advanced	advanced	ADJ
easat-10684	178	33	techniques	technique	NOUN
easat-10684	178	34	for	for	ADP
easat-10684	178	35	improved	improved	ADJ
easat-10684	178	36	detection	detection	NOUN
easat-10684	178	37	accuracy	accuracy	NOUN
easat-10684	178	38	.	.	PUNCT
easat-10684	179	1	gul	gul	PROPN
easat-10684	179	2	et	et	PROPN
easat-10684	179	3	al	al	PROPN
easat-10684	179	4	.	.	PUNCT
easat-10684	180	1	[	[	X
easat-10684	180	2	29	29	NUM
easat-10684	180	3	]	]	X
easat-10684	180	4	gnns	gnns	NOUN
easat-10684	180	5	the	the	DET
easat-10684	180	6	model	model	NOUN
easat-10684	180	7	highlights	highlight	VERB
easat-10684	180	8	the	the	DET
easat-10684	180	9	effectiveness	effectiveness	NOUN
easat-10684	180	10	of	of	ADP
easat-10684	180	11	gnns	gnn	NOUN
easat-10684	180	12	in	in	ADP
easat-10684	180	13	addressing	address	VERB
easat-10684	180	14	the	the	DET
easat-10684	180	15	challenges	challenge	NOUN
easat-10684	180	16	posed	pose	VERB
easat-10684	180	17	by	by	ADP
easat-10684	180	18	the	the	DET
easat-10684	180	19	spread	spread	NOUN
easat-10684	180	20	of	of	ADP
easat-10684	180	21	false	false	ADJ
easat-10684	180	22	information	information	NOUN
easat-10684	180	23	.	.	PUNCT
easat-10684	181	1	the	the	DET
easat-10684	181	2	paper	paper	NOUN
easat-10684	181	3	does	do	AUX
easat-10684	181	4	not	not	PART
easat-10684	181	5	address	address	VERB
easat-10684	181	6	the	the	DET
easat-10684	181	7	limitations	limitation	NOUN
easat-10684	181	8	or	or	CCONJ
easat-10684	181	9	potential	potential	ADJ
easat-10684	181	10	biases	bias	NOUN
easat-10684	181	11	in	in	ADP
easat-10684	181	12	the	the	DET
easat-10684	181	13	data	datum	NOUN
easat-10684	181	14	used	use	VERB
easat-10684	181	15	for	for	ADP
easat-10684	181	16	training	train	VERB
easat-10684	181	17	the	the	DET
easat-10684	181	18	gnn	gnn	PROPN
easat-10684	181	19	model	model	NOUN
easat-10684	181	20	,	,	PUNCT
easat-10684	181	21	which	which	PRON
easat-10684	181	22	could	could	AUX
easat-10684	181	23	affect	affect	VERB
easat-10684	181	24	the	the	DET
easat-10684	181	25	generalizability	generalizability	NOUN
easat-10684	181	26	and	and	CCONJ
easat-10684	181	27	reliability	reliability	NOUN
easat-10684	181	28	of	of	ADP
easat-10684	181	29	the	the	DET
easat-10684	181	30	fake	fake	ADJ
easat-10684	181	31	news	news	NOUN
easat-10684	181	32	1521	1521	NUM
easat-10684	181	33	edelweiss	edelweiss	PROPN
easat-10684	181	34	applied	apply	VERB
easat-10684	181	35	science	science	NOUN
easat-10684	181	36	and	and	CCONJ
easat-10684	181	37	technology	technology	NOUN
easat-10684	181	38	issn	issn	PROPN
easat-10684	181	39	:	:	PUNCT
easat-10684	181	40	2576	2576	NUM
easat-10684	181	41	-	-	SYM
easat-10684	181	42	8484	8484	NUM
easat-10684	181	43	vol	vol	NOUN
easat-10684	181	44	.	.	PROPN
easat-10684	182	1	9	9	NUM
easat-10684	182	2	,	,	PUNCT
easat-10684	182	3	no	no	INTJ
easat-10684	182	4	.	.	NOUN
easat-10684	182	5	10	10	NUM
easat-10684	182	6	:	:	PUNCT
easat-10684	182	7	1515	1515	NUM
easat-10684	182	8	-	-	SYM
easat-10684	182	9	1540	1540	NUM
easat-10684	182	10	,	,	PUNCT
easat-10684	182	11	2025	2025	NUM
easat-10684	182	12	doi	doi	NOUN
easat-10684	182	13	:	:	PUNCT
easat-10684	182	14	10.55214/2576	10.55214/2576	NUM
easat-10684	182	15	-	-	SYM
easat-10684	182	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	183	1	©	©	PROPN
easat-10684	183	2	2025	2025	NUM
easat-10684	183	3	by	by	ADP
easat-10684	183	4	the	the	DET
easat-10684	183	5	authors	author	NOUN
easat-10684	183	6	;	;	PUNCT
easat-10684	183	7	licensee	licensee	PROPN
easat-10684	183	8	learning	learn	VERB
easat-10684	183	9	gate	gate	PROPN
easat-10684	183	10	detection	detection	NOUN
easat-10684	183	11	results	result	NOUN
easat-10684	183	12	.	.	PUNCT
easat-10684	184	1	sari	sari	NOUN
easat-10684	184	2	et	et	PROPN
easat-10684	184	3	al	al	PROPN
easat-10684	184	4	.	.	PUNCT
easat-10684	185	1	[	[	X
easat-10684	185	2	30	30	NUM
easat-10684	185	3	]	]	X
easat-10684	185	4	cnn	cnn	PROPN
easat-10684	185	5	-	-	PUNCT
easat-10684	185	6	bilstm	bilstm	NOUN
easat-10684	185	7	welfake	welfake	NOUN
easat-10684	185	8	the	the	DET
easat-10684	185	9	study	study	NOUN
easat-10684	185	10	developed	develop	VERB
easat-10684	185	11	a	a	DET
easat-10684	185	12	fake	fake	ADJ
easat-10684	185	13	news	news	NOUN
easat-10684	185	14	detection	detection	NOUN
easat-10684	185	15	model	model	NOUN
easat-10684	185	16	using	use	VERB
easat-10684	185	17	an	an	DET
easat-10684	185	18	optimized	optimize	VERB
easat-10684	185	19	cnn	cnn	PROPN
easat-10684	185	20	-	-	PUNCT
easat-10684	185	21	bilstm	bilstm	NOUN
easat-10684	185	22	architecture	architecture	NOUN
easat-10684	185	23	,	,	PUNCT
easat-10684	185	24	achieving	achieve	VERB
easat-10684	185	25	an	an	DET
easat-10684	185	26	accuracy	accuracy	NOUN
easat-10684	185	27	of	of	ADP
easat-10684	185	28	96	96	NUM
easat-10684	185	29	%	%	NOUN
easat-10684	185	30	on	on	ADP
easat-10684	185	31	the	the	DET
easat-10684	185	32	welfake	welfake	NOUN
easat-10684	185	33	dataset	dataset	NOUN
easat-10684	185	34	.	.	PUNCT
easat-10684	186	1	the	the	DET
easat-10684	186	2	study	study	NOUN
easat-10684	186	3	recommends	recommend	VERB
easat-10684	186	4	further	further	ADJ
easat-10684	186	5	research	research	NOUN
easat-10684	186	6	to	to	PART
easat-10684	186	7	explore	explore	VERB
easat-10684	186	8	different	different	ADJ
easat-10684	186	9	preprocessing	preprocessing	NOUN
easat-10684	186	10	techniques	technique	NOUN
easat-10684	186	11	,	,	PUNCT
easat-10684	186	12	indicating	indicate	VERB
easat-10684	186	13	that	that	SCONJ
easat-10684	186	14	the	the	DET
easat-10684	186	15	current	current	ADJ
easat-10684	186	16	methods	method	NOUN
easat-10684	186	17	may	may	AUX
easat-10684	186	18	not	not	PART
easat-10684	186	19	be	be	AUX
easat-10684	186	20	exhaustive	exhaustive	ADJ
easat-10684	186	21	and	and	CCONJ
easat-10684	186	22	that	that	SCONJ
easat-10684	186	23	alternative	alternative	ADJ
easat-10684	186	24	approaches	approach	NOUN
easat-10684	186	25	could	could	AUX
easat-10684	186	26	enhance	enhance	VERB
easat-10684	186	27	the	the	DET
easat-10684	186	28	model	model	NOUN
easat-10684	186	29	's	's	PART
easat-10684	186	30	performance	performance	NOUN
easat-10684	186	31	in	in	ADP
easat-10684	186	32	fake	fake	ADJ
easat-10684	186	33	news	news	NOUN
easat-10684	186	34	detection	detection	NOUN
easat-10684	186	35	.	.	PUNCT
easat-10684	187	1	brinda	brinda	PROPN
easat-10684	187	2	et	et	PROPN
easat-10684	187	3	al	al	PROPN
easat-10684	187	4	.	.	PUNCT
easat-10684	188	1	[	[	X
easat-10684	188	2	31	31	NUM
easat-10684	188	3	]	]	PUNCT
easat-10684	188	4	bert+	bert+	PUNCT
easat-10684	188	5	ulmfit	ulmfit	PROPN
easat-10684	188	6	quint	quint	PROPN
easat-10684	188	7	dataset	dataset	PROPN
easat-10684	188	8	,	,	PUNCT
easat-10684	188	9	boom	boom	NOUN
easat-10684	188	10	live	live	VERB
easat-10684	188	11	the	the	DET
easat-10684	188	12	model	model	NOUN
easat-10684	188	13	achieves	achieve	VERB
easat-10684	188	14	high	high	ADJ
easat-10684	188	15	accuracy	accuracy	NOUN
easat-10684	188	16	rates	rate	NOUN
easat-10684	188	17	,	,	PUNCT
easat-10684	188	18	demonstrating	demonstrate	VERB
easat-10684	188	19	its	its	PRON
easat-10684	188	20	effectiveness	effectiveness	NOUN
easat-10684	188	21	in	in	ADP
easat-10684	188	22	combating	combat	VERB
easat-10684	188	23	evolving	evolve	VERB
easat-10684	188	24	fake	fake	ADJ
easat-10684	188	25	news	news	NOUN
easat-10684	188	26	tactics	tactic	NOUN
easat-10684	188	27	,	,	PUNCT
easat-10684	188	28	with	with	ADP
easat-10684	188	29	a	a	DET
easat-10684	188	30	reported	reported	ADJ
easat-10684	188	31	accuracy	accuracy	NOUN
easat-10684	188	32	of	of	ADP
easat-10684	188	33	95.27	95.27	NUM
easat-10684	188	34	%	%	NOUN
easat-10684	188	35	on	on	ADP
easat-10684	188	36	the	the	DET
easat-10684	188	37	boom	boom	NOUN
easat-10684	188	38	live	live	VERB
easat-10684	188	39	dataset	dataset	NOUN
easat-10684	188	40	and	and	CCONJ
easat-10684	188	41	92.1	92.1	NUM
easat-10684	188	42	%	%	NOUN
easat-10684	188	43	on	on	ADP
easat-10684	188	44	the	the	DET
easat-10684	188	45	quint	quint	NOUN
easat-10684	188	46	dataset	dataset	NOUN
easat-10684	188	47	,	,	PUNCT
easat-10684	188	48	showcasing	showcase	VERB
easat-10684	188	49	its	its	PRON
easat-10684	188	50	generalizability	generalizability	NOUN
easat-10684	188	51	and	and	CCONJ
easat-10684	188	52	robustness	robustness	NOUN
easat-10684	188	53	in	in	ADP
easat-10684	188	54	detecting	detect	VERB
easat-10684	188	55	fake	fake	ADJ
easat-10684	188	56	news	news	NOUN
easat-10684	188	57	.	.	PUNCT
easat-10684	189	1	the	the	DET
easat-10684	189	2	paper	paper	NOUN
easat-10684	189	3	highlights	highlight	VERB
easat-10684	189	4	the	the	DET
easat-10684	189	5	limitations	limitation	NOUN
easat-10684	189	6	of	of	ADP
easat-10684	189	7	traditional	traditional	ADJ
easat-10684	189	8	machine	machine	NOUN
easat-10684	189	9	learning	learning	NOUN
easat-10684	189	10	methods	method	NOUN
easat-10684	189	11	that	that	PRON
easat-10684	189	12	rely	rely	VERB
easat-10684	189	13	solely	solely	ADV
easat-10684	189	14	on	on	ADP
easat-10684	189	15	semantic	semantic	ADJ
easat-10684	189	16	text	text	NOUN
easat-10684	189	17	analysis	analysis	NOUN
easat-10684	189	18	,	,	PUNCT
easat-10684	189	19	indicating	indicate	VERB
easat-10684	189	20	a	a	DET
easat-10684	189	21	research	research	NOUN
easat-10684	189	22	gap	gap	NOUN
easat-10684	189	23	in	in	ADP
easat-10684	189	24	developing	develop	VERB
easat-10684	189	25	more	more	ADV
easat-10684	189	26	robust	robust	ADJ
easat-10684	189	27	detection	detection	NOUN
easat-10684	189	28	methods	method	NOUN
easat-10684	189	29	that	that	PRON
easat-10684	189	30	incorporate	incorporate	VERB
easat-10684	189	31	additional	additional	ADJ
easat-10684	189	32	features	feature	NOUN
easat-10684	189	33	beyond	beyond	ADP
easat-10684	189	34	text	text	NOUN
easat-10684	189	35	,	,	PUNCT
easat-10684	189	36	such	such	ADJ
easat-10684	189	37	as	as	ADP
easat-10684	189	38	dissemination	dissemination	NOUN
easat-10684	189	39	structure	structure	NOUN
easat-10684	189	40	features	feature	VERB
easat-10684	189	41	from	from	ADP
easat-10684	189	42	social	social	ADJ
easat-10684	189	43	networks	network	NOUN
easat-10684	189	44	.	.	PUNCT
easat-10684	190	1	deepthi	deepthi	PROPN
easat-10684	190	2	and	and	CCONJ
easat-10684	190	3	shastry	shastry	NOUN
easat-10684	191	1	[	[	X
easat-10684	191	2	32	32	NUM
easat-10684	191	3	]	]	SYM
easat-10684	191	4	bi	bi	NOUN
easat-10684	191	5	-	-	NOUN
easat-10684	191	6	gru	gru	PROPN
easat-10684	191	7	not	not	PART
easat-10684	191	8	provided	provide	VERB
easat-10684	191	9	the	the	DET
easat-10684	191	10	study	study	NOUN
easat-10684	191	11	highlights	highlight	VERB
easat-10684	191	12	the	the	DET
easat-10684	191	13	effectiveness	effectiveness	NOUN
easat-10684	191	14	of	of	ADP
easat-10684	191	15	deep	deep	ADJ
easat-10684	191	16	learning	learning	NOUN
easat-10684	191	17	in	in	ADP
easat-10684	191	18	detecting	detect	VERB
easat-10684	191	19	fake	fake	ADJ
easat-10684	191	20	news	news	NOUN
easat-10684	191	21	.	.	PUNCT
easat-10684	192	1	subjectivity	subjectivity	NOUN
easat-10684	192	2	and	and	CCONJ
easat-10684	192	3	bias	bias	NOUN
easat-10684	192	4	in	in	ADP
easat-10684	192	5	news	news	NOUN
easat-10684	192	6	authenticity	authenticity	NOUN
easat-10684	192	7	comparison	comparison	NOUN
easat-10684	192	8	,	,	PUNCT
easat-10684	192	9	and	and	CCONJ
easat-10684	192	10	challenges	challenge	NOUN
easat-10684	192	11	in	in	ADP
easat-10684	192	12	contextualising	contextualise	VERB
easat-10684	192	13	news	news	NOUN
easat-10684	192	14	content	content	NOUN
easat-10684	192	15	effectively	effectively	ADV
easat-10684	192	16	.	.	PUNCT
easat-10684	193	1	zhu	zhu	X
easat-10684	194	1	[	[	X
easat-10684	194	2	33	33	NUM
easat-10684	194	3	]	]	PUNCT
easat-10684	194	4	opcnnfake	opcnnfake	NOUN
easat-10684	194	5	,	,	PUNCT
easat-10684	194	6	dccnn	dccnn	PROPN
easat-10684	194	7	,	,	PUNCT
easat-10684	194	8	and	and	CCONJ
easat-10684	194	9	cnn	cnn	PROPN
easat-10684	194	10	-	-	PUNCT
easat-10684	194	11	lstm	lstm	PROPN
easat-10684	194	12	not	not	PART
easat-10684	194	13	provided	provide	VERB
easat-10684	194	14	these	these	DET
easat-10684	194	15	models	model	NOUN
easat-10684	194	16	effectively	effectively	ADV
easat-10684	194	17	address	address	VERB
easat-10684	194	18	the	the	DET
easat-10684	194	19	limitations	limitation	NOUN
easat-10684	194	20	of	of	ADP
easat-10684	194	21	conventional	conventional	ADJ
easat-10684	194	22	approaches	approach	NOUN
easat-10684	194	23	in	in	ADP
easat-10684	194	24	managing	manage	VERB
easat-10684	194	25	growing	grow	VERB
easat-10684	194	26	data	datum	NOUN
easat-10684	194	27	sets	set	NOUN
easat-10684	194	28	and	and	CCONJ
easat-10684	194	29	continuous	continuous	ADJ
easat-10684	194	30	data	datum	NOUN
easat-10684	194	31	streams	stream	NOUN
easat-10684	194	32	.	.	PUNCT
easat-10684	195	1	poor	poor	ADJ
easat-10684	195	2	scalability	scalability	NOUN
easat-10684	195	3	of	of	ADP
easat-10684	195	4	dl	dl	PROPN
easat-10684	195	5	models	model	NOUN
easat-10684	195	6	for	for	ADP
easat-10684	195	7	fake	fake	ADJ
easat-10684	195	8	news	news	NOUN
easat-10684	195	9	and	and	CCONJ
easat-10684	195	10	slow	slow	ADJ
easat-10684	195	11	training	training	NOUN
easat-10684	195	12	speed	speed	NOUN
easat-10684	195	13	of	of	ADP
easat-10684	195	14	existing	exist	VERB
easat-10684	195	15	dl	dl	PROPN
easat-10684	195	16	models	model	NOUN
easat-10684	195	17	.	.	PUNCT
easat-10684	196	1	ladouceur	ladouceur	PROPN
easat-10684	196	2	,	,	PUNCT
easat-10684	196	3	et	et	PROPN
easat-10684	196	4	al	al	PROPN
easat-10684	196	5	.	.	PUNCT
easat-10684	197	1	[	[	X
easat-10684	197	2	34	34	NUM
easat-10684	197	3	]	]	AUX
easat-10684	197	4	bert+lstm	bert+lstm	VERB
easat-10684	197	5	5,000	5,000	NUM
easat-10684	197	6	news	news	NOUN
easat-10684	197	7	articles	article	NOUN
easat-10684	197	8	extracted	extract	VERB
easat-10684	197	9	from	from	ADP
easat-10684	197	10	the	the	DET
easat-10684	197	11	covid-19	covid-19	PROPN
easat-10684	197	12	dataset	dataset	VERB
easat-10684	197	13	the	the	DET
easat-10684	197	14	model	model	NOUN
easat-10684	197	15	demonstrated	demonstrate	VERB
easat-10684	197	16	superior	superior	ADJ
easat-10684	197	17	performance	performance	NOUN
easat-10684	197	18	in	in	ADP
easat-10684	197	19	fake	fake	ADJ
easat-10684	197	20	news	news	NOUN
easat-10684	197	21	detection	detection	NOUN
easat-10684	197	22	compared	compare	VERB
easat-10684	197	23	to	to	ADP
easat-10684	197	24	traditional	traditional	ADJ
easat-10684	197	25	methods	method	NOUN
easat-10684	197	26	.	.	PUNCT
easat-10684	198	1	problem	problem	NOUN
easat-10684	198	2	of	of	ADP
easat-10684	198	3	feature	feature	NOUN
easat-10684	198	4	selection	selection	NOUN
easat-10684	198	5	porya	porya	NOUN
easat-10684	198	6	,	,	PUNCT
easat-10684	198	7	et	et	PROPN
easat-10684	198	8	al	al	PROPN
easat-10684	198	9	.	.	PUNCT
easat-10684	199	1	[	[	X
easat-10684	199	2	35	35	NUM
easat-10684	199	3	]	]	SYM
easat-10684	199	4	rf	rf	NOUN
easat-10684	199	5	,	,	PUNCT
easat-10684	199	6	gb	gb	PROPN
easat-10684	199	7	,	,	PUNCT
easat-10684	199	8	cnn	cnn	PROPN
easat-10684	199	9	,	,	PUNCT
easat-10684	199	10	&	&	CCONJ
easat-10684	199	11	lstm	lstm	PROPN
easat-10684	199	12	two	two	NUM
easat-10684	199	13	widely	widely	ADV
easat-10684	199	14	used	use	VERB
easat-10684	199	15	datasets	dataset	NOUN
easat-10684	199	16	the	the	DET
easat-10684	199	17	paper	paper	NOUN
easat-10684	199	18	explores	explore	NOUN
easat-10684	199	19	preprocessing	preprocesse	VERB
easat-10684	199	20	techniques	technique	NOUN
easat-10684	199	21	and	and	CCONJ
easat-10684	199	22	model	model	NOUN
easat-10684	199	23	architectures	architecture	NOUN
easat-10684	199	24	,	,	PUNCT
easat-10684	199	25	tested	test	VERB
easat-10684	199	26	on	on	ADP
easat-10684	199	27	two	two	NUM
easat-10684	199	28	datasets	dataset	NOUN
easat-10684	199	29	,	,	PUNCT
easat-10684	199	30	contributing	contribute	VERB
easat-10684	199	31	to	to	ADP
easat-10684	199	32	fake	fake	ADJ
easat-10684	199	33	news	news	NOUN
easat-10684	199	34	detection	detection	NOUN
easat-10684	199	35	research	research	NOUN
easat-10684	199	36	and	and	CCONJ
easat-10684	199	37	guiding	guide	VERB
easat-10684	199	38	future	future	ADJ
easat-10684	199	39	advancements	advancement	NOUN
easat-10684	199	40	for	for	ADP
easat-10684	199	41	newcomers	newcomer	NOUN
easat-10684	199	42	.	.	PUNCT
easat-10684	200	1	model	model	NOUN
easat-10684	200	2	implementation	implementation	NOUN
easat-10684	200	3	issues	issue	NOUN
easat-10684	200	4	hinder	hinder	VERB
easat-10684	200	5	effective	effective	ADJ
easat-10684	200	6	fake	fake	ADJ
easat-10684	200	7	news	news	NOUN
easat-10684	200	8	detection	detection	NOUN
easat-10684	200	9	,	,	PUNCT
easat-10684	200	10	and	and	CCONJ
easat-10684	200	11	a	a	DET
easat-10684	200	12	lack	lack	NOUN
easat-10684	200	13	of	of	ADP
easat-10684	200	14	clean	clean	ADJ
easat-10684	200	15	,	,	PUNCT
easat-10684	200	16	unbiased	unbiased	ADJ
easat-10684	200	17	data	datum	NOUN
easat-10684	200	18	poses	pose	VERB
easat-10684	200	19	a	a	DET
easat-10684	200	20	challenge	challenge	NOUN
easat-10684	200	21	in	in	ADP
easat-10684	200	22	research	research	NOUN
easat-10684	200	23	.	.	PUNCT
easat-10684	201	1	3	3	X
easat-10684	201	2	.	.	X
easat-10684	201	3	research	research	NOUN
easat-10684	201	4	methodology	methodology	NOUN
easat-10684	201	5	the	the	DET
easat-10684	201	6	study	study	NOUN
easat-10684	201	7	proposed	propose	VERB
easat-10684	201	8	a	a	DET
easat-10684	201	9	transfer	transfer	NOUN
easat-10684	201	10	learning	learning	NOUN
easat-10684	201	11	-	-	PUNCT
easat-10684	201	12	based	base	VERB
easat-10684	201	13	bidirectional	bidirectional	ADJ
easat-10684	201	14	long	long	ADJ
easat-10684	201	15	-	-	PUNCT
easat-10684	201	16	short	short	ADJ
easat-10684	201	17	-	-	PUNCT
easat-10684	201	18	term	term	NOUN
easat-10684	201	19	memory	memory	NOUN
easat-10684	201	20	(	(	PUNCT
easat-10684	201	21	bi	bi	NOUN
easat-10684	201	22	-	-	ADJ
easat-10684	201	23	lstm	lstm	NOUN
easat-10684	201	24	)	)	PUNCT
easat-10684	201	25	for	for	ADP
easat-10684	201	26	a	a	DET
easat-10684	201	27	machine	machine	NOUN
easat-10684	201	28	learning	learning	NOUN
easat-10684	201	29	(	(	PUNCT
easat-10684	201	30	ml	ml	NOUN
easat-10684	201	31	)	)	PUNCT
easat-10684	201	32	fake	fake	ADJ
easat-10684	201	33	news	news	NOUN
easat-10684	201	34	detection	detection	NOUN
easat-10684	201	35	model	model	NOUN
easat-10684	201	36	(	(	PUNCT
easat-10684	201	37	bi	bi	NOUN
easat-10684	201	38	-	-	ADJ
easat-10684	201	39	lstm-2	lstm-2	NOUN
easat-10684	201	40	-	-	PUNCT
easat-10684	201	41	ml	ml	NOUN
easat-10684	201	42	model	model	NOUN
easat-10684	201	43	)	)	PUNCT
easat-10684	201	44	.	.	PUNCT
easat-10684	202	1	technically	technically	ADV
easat-10684	202	2	,	,	PUNCT
easat-10684	202	3	the	the	DET
easat-10684	202	4	bilstm	bilstm	NOUN
easat-10684	202	5	architecture	architecture	NOUN
easat-10684	202	6	is	be	AUX
easat-10684	202	7	used	use	VERB
easat-10684	202	8	to	to	PART
easat-10684	202	9	pre	pre	VERB
easat-10684	202	10	-	-	VERB
easat-10684	202	11	train	train	VERB
easat-10684	202	12	the	the	DET
easat-10684	202	13	fake	fake	ADJ
easat-10684	202	14	news	news	NOUN
easat-10684	202	15	detection	detection	NOUN
easat-10684	202	16	model	model	NOUN
easat-10684	202	17	to	to	PART
easat-10684	202	18	generate	generate	VERB
easat-10684	202	19	highly	highly	ADV
easat-10684	202	20	contextualized	contextualized	ADJ
easat-10684	202	21	bidirectional	bidirectional	ADJ
easat-10684	202	22	feature	feature	NOUN
easat-10684	202	23	encodings	encoding	NOUN
easat-10684	202	24	or	or	CCONJ
easat-10684	202	25	vectors	vector	NOUN
easat-10684	202	26	,	,	PUNCT
easat-10684	202	27	which	which	PRON
easat-10684	202	28	are	be	AUX
easat-10684	202	29	later	later	ADV
easat-10684	202	30	transferred	transfer	VERB
easat-10684	202	31	to	to	ADP
easat-10684	202	32	a	a	DET
easat-10684	202	33	machine	machine	NOUN
easat-10684	202	34	learning	learn	VERB
easat-10684	202	35	algorithm	algorithm	NOUN
easat-10684	202	36	for	for	ADP
easat-10684	202	37	further	further	ADJ
easat-10684	202	38	training	training	NOUN
easat-10684	202	39	.	.	PUNCT
easat-10684	203	1	3.1	3.1	NUM
easat-10684	203	2	.	.	NOUN
easat-10684	203	3	methodological	methodological	ADJ
easat-10684	203	4	process	process	NOUN
easat-10684	203	5	figure	figure	NOUN
easat-10684	203	6	1	1	NUM
easat-10684	203	7	illustrates	illustrate	VERB
easat-10684	203	8	the	the	DET
easat-10684	203	9	conceptual	conceptual	ADJ
easat-10684	203	10	framework	framework	NOUN
easat-10684	203	11	of	of	ADP
easat-10684	203	12	how	how	SCONJ
easat-10684	203	13	the	the	DET
easat-10684	203	14	bidirectional	bidirectional	ADJ
easat-10684	203	15	long	long	ADJ
easat-10684	203	16	short	short	ADJ
easat-10684	203	17	-	-	PUNCT
easat-10684	203	18	term	term	NOUN
easat-10684	203	19	memory	memory	NOUN
easat-10684	203	20	(	(	PUNCT
easat-10684	203	21	bi	bi	NOUN
easat-10684	203	22	-	-	ADJ
easat-10684	203	23	lstm	lstm	NOUN
easat-10684	203	24	)	)	PUNCT
easat-10684	203	25	to	to	ADP
easat-10684	203	26	machine	machine	NOUN
easat-10684	203	27	learning	learning	NOUN
easat-10684	203	28	(	(	PUNCT
easat-10684	203	29	ml	ml	NOUN
easat-10684	203	30	)	)	PUNCT
easat-10684	203	31	transfer	transfer	NOUN
easat-10684	203	32	learning	learning	NOUN
easat-10684	203	33	-	-	PUNCT
easat-10684	203	34	based	base	VERB
easat-10684	203	35	model	model	NOUN
easat-10684	203	36	(	(	PUNCT
easat-10684	203	37	bi	bi	NOUN
easat-10684	203	38	-	-	ADJ
easat-10684	203	39	lsmt-2	lsmt-2	PROPN
easat-10684	203	40	-	-	PUNCT
easat-10684	203	41	ml	ml	NOUN
easat-10684	203	42	)	)	PUNCT
easat-10684	203	43	for	for	ADP
easat-10684	203	44	detecting	detect	VERB
easat-10684	203	45	fake	fake	ADJ
easat-10684	203	46	news	news	NOUN
easat-10684	203	47	will	will	AUX
easat-10684	203	48	be	be	AUX
easat-10684	203	49	designed	design	VERB
easat-10684	203	50	and	and	CCONJ
easat-10684	203	51	developed	develop	VERB
easat-10684	203	52	.	.	PUNCT
easat-10684	204	1	the	the	DET
easat-10684	204	2	various	various	ADJ
easat-10684	204	3	technologies	technology	NOUN
easat-10684	204	4	,	,	PUNCT
easat-10684	204	5	entities	entity	NOUN
easat-10684	204	6	,	,	PUNCT
easat-10684	204	7	and	and	CCONJ
easat-10684	204	8	tools	tool	NOUN
easat-10684	204	9	that	that	PRON
easat-10684	204	10	will	will	AUX
easat-10684	204	11	be	be	AUX
easat-10684	204	12	utilized	utilize	VERB
easat-10684	204	13	in	in	ADP
easat-10684	204	14	designing	design	VERB
easat-10684	204	15	and	and	CCONJ
easat-10684	204	16	developing	develop	VERB
easat-10684	204	17	the	the	DET
easat-10684	204	18	bi	bi	NOUN
easat-10684	204	19	-	-	PUNCT
easat-10684	204	20	lstm-2	lstm-2	NOUN
easat-10684	204	21	-	-	PUNCT
easat-10684	204	22	ml	ml	ADV
easat-10684	204	23	fake	fake	ADJ
easat-10684	204	24	news	news	NOUN
easat-10684	204	25	detection	detection	NOUN
easat-10684	204	26	model	model	NOUN
easat-10684	204	27	are	be	AUX
easat-10684	204	28	also	also	ADV
easat-10684	204	29	described	describe	VERB
easat-10684	204	30	1522	1522	NUM
easat-10684	204	31	edelweiss	edelweiss	PROPN
easat-10684	204	32	applied	apply	VERB
easat-10684	204	33	science	science	NOUN
easat-10684	204	34	and	and	CCONJ
easat-10684	204	35	technology	technology	NOUN
easat-10684	204	36	issn	issn	PROPN
easat-10684	204	37	:	:	PUNCT
easat-10684	204	38	2576	2576	NUM
easat-10684	204	39	-	-	SYM
easat-10684	204	40	8484	8484	NUM
easat-10684	204	41	vol	vol	NOUN
easat-10684	204	42	.	.	PROPN
easat-10684	205	1	9	9	NUM
easat-10684	205	2	,	,	PUNCT
easat-10684	205	3	no	no	INTJ
easat-10684	205	4	.	.	NOUN
easat-10684	205	5	10	10	NUM
easat-10684	205	6	:	:	PUNCT
easat-10684	205	7	1515	1515	NUM
easat-10684	205	8	-	-	SYM
easat-10684	205	9	1540	1540	NUM
easat-10684	205	10	,	,	PUNCT
easat-10684	205	11	2025	2025	NUM
easat-10684	205	12	doi	doi	NOUN
easat-10684	205	13	:	:	PUNCT
easat-10684	205	14	10.55214/2576	10.55214/2576	NUM
easat-10684	205	15	-	-	SYM
easat-10684	205	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	206	1	©	©	PROPN
easat-10684	206	2	2025	2025	NUM
easat-10684	206	3	by	by	ADP
easat-10684	206	4	the	the	DET
easat-10684	206	5	authors	author	NOUN
easat-10684	206	6	;	;	PUNCT
easat-10684	206	7	licensee	licensee	PROPN
easat-10684	206	8	learning	learning	NOUN
easat-10684	206	9	gate	gate	NOUN
easat-10684	206	10	in	in	ADP
easat-10684	206	11	figure	figure	NOUN
easat-10684	206	12	1	1	NUM
easat-10684	206	13	.	.	PUNCT
easat-10684	207	1	the	the	DET
easat-10684	207	2	fake	fake	ADJ
easat-10684	207	3	news	news	NOUN
easat-10684	207	4	dataset	dataset	NOUN
easat-10684	207	5	is	be	AUX
easat-10684	207	6	downloaded	download	VERB
easat-10684	207	7	from	from	ADP
easat-10684	207	8	the	the	DET
easat-10684	207	9	kaggle	kaggle	ADJ
easat-10684	207	10	repository	repository	NOUN
easat-10684	207	11	in	in	ADP
easat-10684	207	12	a	a	DET
easat-10684	207	13	comma	comma	NOUN
easat-10684	207	14	separated	separate	VERB
easat-10684	207	15	values	value	NOUN
easat-10684	207	16	(	(	PUNCT
easat-10684	207	17	csv	csv	NOUN
easat-10684	207	18	)	)	PUNCT
easat-10684	207	19	format	format	NOUN
easat-10684	207	20	,	,	PUNCT
easat-10684	207	21	and	and	CCONJ
easat-10684	207	22	the	the	DET
easat-10684	207	23	data	data	NOUN
easat-10684	207	24	scientist	scientist	NOUN
easat-10684	207	25	uses	use	VERB
easat-10684	207	26	the	the	DET
easat-10684	207	27	jupyter	jupyter	ADJ
easat-10684	207	28	notebook	notebook	NOUN
easat-10684	207	29	environment	environment	NOUN
easat-10684	207	30	to	to	PART
easat-10684	207	31	load	load	VERB
easat-10684	207	32	the	the	DET
easat-10684	207	33	downloaded	download	VERB
easat-10684	207	34	fake	fake	ADJ
easat-10684	207	35	news	news	NOUN
easat-10684	207	36	dataset	dataset	VERB
easat-10684	207	37	into	into	ADP
easat-10684	207	38	the	the	DET
easat-10684	207	39	ide	ide	NOUN
easat-10684	207	40	using	use	VERB
easat-10684	207	41	the	the	DET
easat-10684	207	42	python	python	NOUN
easat-10684	207	43	‘	'	PUNCT
easat-10684	207	44	load_csv	load_csv	NUM
easat-10684	207	45	’	'	PUNCT
easat-10684	207	46	function	function	NOUN
easat-10684	207	47	.	.	PUNCT
easat-10684	208	1	furthermore	furthermore	ADV
easat-10684	208	2	,	,	PUNCT
easat-10684	208	3	the	the	DET
easat-10684	208	4	loaded	loaded	ADJ
easat-10684	208	5	fake	fake	ADJ
easat-10684	208	6	news	news	NOUN
easat-10684	208	7	data	datum	NOUN
easat-10684	208	8	will	will	AUX
easat-10684	208	9	be	be	AUX
easat-10684	208	10	forwarded	forward	VERB
easat-10684	208	11	to	to	PART
easat-10684	208	12	stage	stage	NOUN
easat-10684	208	13	(	(	PUNCT
easat-10684	208	14	1	1	NUM
easat-10684	208	15	)	)	PUNCT
easat-10684	208	16	,	,	PUNCT
easat-10684	208	17	or	or	CCONJ
easat-10684	208	18	the	the	DET
easat-10684	208	19	preprocessing	preprocessing	NOUN
easat-10684	208	20	stage	stage	NOUN
easat-10684	208	21	,	,	PUNCT
easat-10684	208	22	which	which	PRON
easat-10684	208	23	includes	include	VERB
easat-10684	208	24	data	datum	NOUN
easat-10684	208	25	exploration	exploration	NOUN
easat-10684	208	26	,	,	PUNCT
easat-10684	208	27	data	datum	NOUN
easat-10684	208	28	cleaning	cleaning	NOUN
easat-10684	208	29	,	,	PUNCT
easat-10684	208	30	and	and	CCONJ
easat-10684	208	31	data	datum	NOUN
easat-10684	208	32	visualization	visualization	NOUN
easat-10684	208	33	.	.	PUNCT
easat-10684	209	1	the	the	DET
easat-10684	209	2	exploration	exploration	NOUN
easat-10684	209	3	helps	help	VERB
easat-10684	209	4	to	to	PART
easat-10684	209	5	gain	gain	VERB
easat-10684	209	6	more	more	ADJ
easat-10684	209	7	insights	insight	NOUN
easat-10684	209	8	into	into	ADP
easat-10684	209	9	the	the	DET
easat-10684	209	10	dataset	dataset	NOUN
easat-10684	209	11	and	and	CCONJ
easat-10684	209	12	identify	identify	VERB
easat-10684	209	13	possible	possible	ADJ
easat-10684	209	14	correlations	correlation	NOUN
easat-10684	209	15	,	,	PUNCT
easat-10684	209	16	relationships	relationship	NOUN
easat-10684	209	17	,	,	PUNCT
easat-10684	209	18	or	or	CCONJ
easat-10684	209	19	features	feature	NOUN
easat-10684	209	20	that	that	PRON
easat-10684	209	21	can	can	AUX
easat-10684	209	22	assist	assist	VERB
easat-10684	209	23	during	during	ADP
easat-10684	209	24	machine	machine	NOUN
easat-10684	209	25	learning	learning	NOUN
easat-10684	209	26	prediction	prediction	NOUN
easat-10684	209	27	.	.	PUNCT
easat-10684	210	1	the	the	DET
easat-10684	210	2	x	x	NOUN
easat-10684	210	3	and	and	CCONJ
easat-10684	210	4	y	y	PROPN
easat-10684	210	5	variables	variable	NOUN
easat-10684	210	6	(	(	PUNCT
easat-10684	210	7	independent	independent	ADJ
easat-10684	210	8	and	and	CCONJ
easat-10684	210	9	dependent	dependent	ADJ
easat-10684	210	10	variables	variable	NOUN
easat-10684	210	11	)	)	PUNCT
easat-10684	210	12	of	of	ADP
easat-10684	210	13	the	the	DET
easat-10684	210	14	fake	fake	ADJ
easat-10684	210	15	news	news	NOUN
easat-10684	210	16	dataset	dataset	NOUN
easat-10684	210	17	are	be	AUX
easat-10684	210	18	extracted	extract	VERB
easat-10684	210	19	in	in	ADP
easat-10684	210	20	stage	stage	NOUN
easat-10684	210	21	2	2	NUM
easat-10684	210	22	.	.	PUNCT
easat-10684	211	1	furthermore	furthermore	ADV
easat-10684	211	2	,	,	PUNCT
easat-10684	211	3	a	a	DET
easat-10684	211	4	transformation	transformation	NOUN
easat-10684	211	5	(	(	PUNCT
easat-10684	211	6	word2vec	word2vec	X
easat-10684	211	7	)	)	PUNCT
easat-10684	211	8	will	will	AUX
easat-10684	211	9	be	be	AUX
easat-10684	211	10	conducted	conduct	VERB
easat-10684	211	11	on	on	ADP
easat-10684	211	12	the	the	DET
easat-10684	211	13	raw	raw	ADJ
easat-10684	211	14	text	text	NOUN
easat-10684	211	15	to	to	PART
easat-10684	211	16	convert	convert	VERB
easat-10684	211	17	the	the	DET
easat-10684	211	18	text	text	NOUN
easat-10684	211	19	data	datum	NOUN
easat-10684	211	20	into	into	ADP
easat-10684	211	21	numerical	numerical	ADJ
easat-10684	211	22	features	feature	NOUN
easat-10684	211	23	or	or	CCONJ
easat-10684	211	24	word	word	NOUN
easat-10684	211	25	embeddings	embedding	NOUN
easat-10684	211	26	.	.	PUNCT
easat-10684	212	1	after	after	ADP
easat-10684	212	2	feature	feature	NOUN
easat-10684	212	3	extraction	extraction	NOUN
easat-10684	212	4	,	,	PUNCT
easat-10684	212	5	the	the	DET
easat-10684	212	6	next	next	ADJ
easat-10684	212	7	step	step	NOUN
easat-10684	212	8	is	be	AUX
easat-10684	212	9	to	to	PART
easat-10684	212	10	split	split	VERB
easat-10684	212	11	the	the	DET
easat-10684	212	12	dataset	dataset	NOUN
easat-10684	212	13	into	into	ADP
easat-10684	212	14	training	training	NOUN
easat-10684	212	15	and	and	CCONJ
easat-10684	212	16	testing	testing	NOUN
easat-10684	212	17	samples	sample	NOUN
easat-10684	212	18	.	.	PUNCT
easat-10684	213	1	the	the	DET
easat-10684	213	2	training	training	NOUN
easat-10684	213	3	sample	sample	NOUN
easat-10684	213	4	will	will	AUX
easat-10684	213	5	be	be	AUX
easat-10684	213	6	utilized	utilize	VERB
easat-10684	213	7	to	to	PART
easat-10684	213	8	train	train	VERB
easat-10684	213	9	the	the	DET
easat-10684	213	10	bi	bi	NOUN
easat-10684	213	11	-	-	ADJ
easat-10684	213	12	lstm	lstm	ADJ
easat-10684	213	13	deep	deep	ADJ
easat-10684	213	14	learning	learning	NOUN
easat-10684	213	15	model	model	NOUN
easat-10684	213	16	to	to	PART
easat-10684	213	17	extract	extract	VERB
easat-10684	213	18	highly	highly	ADV
easat-10684	213	19	contextualized	contextualized	ADJ
easat-10684	213	20	information	information	NOUN
easat-10684	213	21	about	about	ADP
easat-10684	213	22	the	the	DET
easat-10684	213	23	fake	fake	ADJ
easat-10684	213	24	news	news	NOUN
easat-10684	213	25	data	datum	NOUN
easat-10684	213	26	,	,	PUNCT
easat-10684	213	27	then	then	ADV
easat-10684	213	28	use	use	VERB
easat-10684	213	29	the	the	DET
easat-10684	213	30	pre	pre	ADJ
easat-10684	213	31	-	-	ADJ
easat-10684	213	32	trained	train	VERB
easat-10684	213	33	bi	bi	NOUN
easat-10684	213	34	-	-	ADJ
easat-10684	213	35	lstm	lstm	ADJ
easat-10684	213	36	weights	weight	NOUN
easat-10684	213	37	to	to	PART
easat-10684	213	38	train	train	VERB
easat-10684	213	39	the	the	DET
easat-10684	213	40	machine	machine	NOUN
easat-10684	213	41	learning	learn	VERB
easat-10684	213	42	algorithm	algorithm	NOUN
easat-10684	213	43	in	in	ADP
easat-10684	213	44	a	a	DET
easat-10684	213	45	transfer	transfer	NOUN
easat-10684	213	46	learning	learning	NOUN
easat-10684	213	47	approach	approach	NOUN
easat-10684	213	48	.	.	PUNCT
easat-10684	214	1	the	the	DET
easat-10684	214	2	developed	develop	VERB
easat-10684	214	3	model	model	NOUN
easat-10684	214	4	will	will	AUX
easat-10684	214	5	then	then	ADV
easat-10684	214	6	be	be	AUX
easat-10684	214	7	evaluated	evaluate	VERB
easat-10684	214	8	using	use	VERB
easat-10684	214	9	various	various	ADJ
easat-10684	214	10	performance	performance	NOUN
easat-10684	214	11	metrics	metric	NOUN
easat-10684	214	12	such	such	ADJ
easat-10684	214	13	as	as	ADP
easat-10684	214	14	accuracy	accuracy	NOUN
easat-10684	214	15	,	,	PUNCT
easat-10684	214	16	precision	precision	NOUN
easat-10684	214	17	,	,	PUNCT
easat-10684	214	18	recall	recall	NOUN
easat-10684	214	19	,	,	PUNCT
easat-10684	214	20	and	and	CCONJ
easat-10684	214	21	f1score	f1score	NOUN
easat-10684	214	22	.	.	PUNCT
easat-10684	215	1	figure	figure	NOUN
easat-10684	215	2	1	1	NUM
easat-10684	215	3	.	.	PUNCT
easat-10684	216	1	transfer	transfer	NOUN
easat-10684	216	2	learning	learn	VERB
easat-10684	216	3	bi	bi	NOUN
easat-10684	216	4	-	-	PUNCT
easat-10684	216	5	lstm-2	lstm-2	NOUN
easat-10684	216	6	-	-	PUNCT
easat-10684	216	7	ml	ml	ADV
easat-10684	216	8	proposed	propose	VERB
easat-10684	216	9	framework	framework	NOUN
easat-10684	216	10	.	.	PUNCT
easat-10684	217	1	figure	figure	NOUN
easat-10684	217	2	1	1	NUM
easat-10684	217	3	illustrates	illustrate	VERB
easat-10684	217	4	the	the	DET
easat-10684	217	5	methodological	methodological	ADJ
easat-10684	217	6	process	process	NOUN
easat-10684	217	7	involved	involve	VERB
easat-10684	217	8	in	in	ADP
easat-10684	217	9	developing	develop	VERB
easat-10684	217	10	the	the	DET
easat-10684	217	11	proposed	propose	VERB
easat-10684	217	12	transfer	transfer	NOUN
easat-10684	217	13	learning	learn	VERB
easat-10684	217	14	bi	bi	NOUN
easat-10684	217	15	-	-	PUNCT
easat-10684	217	16	lstm-2	lstm-2	NOUN
easat-10684	217	17	-	-	PUNCT
easat-10684	217	18	ml	ml	ADV
easat-10684	217	19	fake	fake	ADJ
easat-10684	217	20	news	news	NOUN
easat-10684	217	21	detection	detection	NOUN
easat-10684	217	22	model	model	NOUN
easat-10684	217	23	using	use	VERB
easat-10684	217	24	transfer	transfer	NOUN
easat-10684	217	25	learning	learning	NOUN
easat-10684	217	26	.	.	PUNCT
easat-10684	218	1	the	the	DET
easat-10684	218	2	proposed	propose	VERB
easat-10684	218	3	stages	stage	NOUN
easat-10684	218	4	include	include	VERB
easat-10684	218	5	the	the	DET
easat-10684	218	6	following	follow	VERB
easat-10684	218	7	:	:	PUNCT
easat-10684	219	1	3.1.1	3.1.1	X
easat-10684	219	2	.	.	PUNCT
easat-10684	219	3	fake	fake	ADJ
easat-10684	219	4	news	news	NOUN
easat-10684	219	5	data	datum	NOUN
easat-10684	219	6	collection	collection	NOUN
easat-10684	219	7	this	this	DET
easat-10684	219	8	first	first	ADJ
easat-10684	219	9	stage	stage	NOUN
easat-10684	219	10	includes	include	VERB
easat-10684	219	11	data	datum	NOUN
easat-10684	219	12	collection	collection	NOUN
easat-10684	219	13	and	and	CCONJ
easat-10684	219	14	gathering	gathering	NOUN
easat-10684	219	15	from	from	ADP
easat-10684	219	16	the	the	DET
easat-10684	219	17	kaggle	kaggle	ADJ
easat-10684	219	18	repository	repository	NOUN
easat-10684	219	19	.	.	PUNCT
easat-10684	220	1	the	the	DET
easat-10684	220	2	dataset	dataset	NOUN
easat-10684	220	3	is	be	AUX
easat-10684	220	4	downloaded	download	VERB
easat-10684	220	5	in	in	ADP
easat-10684	220	6	a	a	DET
easat-10684	220	7	comma	comma	NOUN
easat-10684	220	8	-	-	PUNCT
easat-10684	220	9	separated	separate	VERB
easat-10684	220	10	value	value	NOUN
easat-10684	220	11	(	(	PUNCT
easat-10684	220	12	csv	csv	NOUN
easat-10684	220	13	)	)	PUNCT
easat-10684	220	14	format	format	NOUN
easat-10684	220	15	into	into	ADP
easat-10684	220	16	the	the	DET
easat-10684	220	17	local	local	ADJ
easat-10684	220	18	system	system	NOUN
easat-10684	220	19	drive	drive	VERB
easat-10684	220	20	for	for	ADP
easat-10684	220	21	further	further	ADJ
easat-10684	220	22	processing	processing	NOUN
easat-10684	220	23	.	.	PUNCT
easat-10684	221	1	however	however	ADV
easat-10684	221	2	,	,	PUNCT
easat-10684	221	3	the	the	DET
easat-10684	221	4	kaggle	kaggle	ADJ
easat-10684	221	5	repository	repository	NOUN
easat-10684	221	6	provides	provide	VERB
easat-10684	221	7	data	datum	NOUN
easat-10684	221	8	scientists	scientist	NOUN
easat-10684	221	9	with	with	ADP
easat-10684	221	10	a	a	DET
easat-10684	221	11	variety	variety	NOUN
easat-10684	221	12	of	of	ADP
easat-10684	221	13	datasets	dataset	NOUN
easat-10684	221	14	for	for	ADP
easat-10684	221	15	addressing	address	VERB
easat-10684	221	16	different	different	ADJ
easat-10684	221	17	tasks	task	NOUN
easat-10684	221	18	,	,	PUNCT
easat-10684	221	19	a	a	DET
easat-10684	221	20	development	development	NOUN
easat-10684	221	21	environment	environment	NOUN
easat-10684	221	22	(	(	PUNCT
easat-10684	221	23	jupyter	jupyter	ADJ
easat-10684	221	24	notebook	notebook	NOUN
easat-10684	221	25	)	)	PUNCT
easat-10684	221	26	,	,	PUNCT
easat-10684	221	27	and	and	CCONJ
easat-10684	221	28	virtual	virtual	ADJ
easat-10684	221	29	hardware	hardware	NOUN
easat-10684	221	30	processing	processing	NOUN
easat-10684	221	31	units	unit	NOUN
easat-10684	221	32	(	(	PUNCT
easat-10684	221	33	gpu	gpu	PROPN
easat-10684	221	34	,	,	PUNCT
easat-10684	221	35	tpu	tpu	NOUN
easat-10684	221	36	,	,	PUNCT
easat-10684	221	37	and	and	CCONJ
easat-10684	221	38	cpu	cpu	PROPN
easat-10684	221	39	)	)	PUNCT
easat-10684	221	40	for	for	ADP
easat-10684	221	41	solving	solve	VERB
easat-10684	221	42	complex	complex	ADJ
easat-10684	221	43	tasks	task	NOUN
easat-10684	221	44	.	.	PUNCT
easat-10684	222	1	the	the	DET
easat-10684	222	2	fake	fake	ADJ
easat-10684	222	3	news	news	NOUN
easat-10684	222	4	dataset	dataset	NOUN
easat-10684	222	5	link	link	NOUN
easat-10684	222	6	on	on	ADP
easat-10684	222	7	the	the	DET
easat-10684	222	8	kaggle	kaggle	ADJ
easat-10684	222	9	repository	repository	NOUN
easat-10684	222	10	is	be	AUX
easat-10684	222	11	provided	provide	VERB
easat-10684	222	12	on	on	ADP
easat-10684	222	13	the	the	DET
easat-10684	222	14	website	website	NOUN
easat-10684	222	15	.	.	PUNCT
easat-10684	223	1	https://github.com/georgemcintire/fake_real_news_dataset	https://github.com/georgemcintire/fake_real_news_dataset	NOUN
easat-10684	223	2	https://github.com/georgemcintire/fake_real_news_dataset	https://github.com/georgemcintire/fake_real_news_dataset	VERB
easat-10684	223	3	1523	1523	NUM
easat-10684	223	4	edelweiss	edelweiss	PROPN
easat-10684	223	5	applied	apply	VERB
easat-10684	223	6	science	science	NOUN
easat-10684	223	7	and	and	CCONJ
easat-10684	223	8	technology	technology	NOUN
easat-10684	223	9	issn	issn	PROPN
easat-10684	223	10	:	:	PUNCT
easat-10684	223	11	2576	2576	NUM
easat-10684	223	12	-	-	SYM
easat-10684	223	13	8484	8484	NUM
easat-10684	223	14	vol	vol	NOUN
easat-10684	223	15	.	.	PROPN
easat-10684	224	1	9	9	NUM
easat-10684	224	2	,	,	PUNCT
easat-10684	224	3	no	no	INTJ
easat-10684	224	4	.	.	NOUN
easat-10684	224	5	10	10	NUM
easat-10684	224	6	:	:	PUNCT
easat-10684	224	7	1515	1515	NUM
easat-10684	224	8	-	-	SYM
easat-10684	224	9	1540	1540	NUM
easat-10684	224	10	,	,	PUNCT
easat-10684	224	11	2025	2025	NUM
easat-10684	224	12	doi	doi	NOUN
easat-10684	224	13	:	:	PUNCT
easat-10684	224	14	10.55214/2576	10.55214/2576	NUM
easat-10684	224	15	-	-	SYM
easat-10684	224	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	225	1	©	©	PROPN
easat-10684	225	2	2025	2025	NUM
easat-10684	225	3	by	by	ADP
easat-10684	225	4	the	the	DET
easat-10684	225	5	authors	author	NOUN
easat-10684	225	6	;	;	PUNCT
easat-10684	225	7	licensee	licensee	PROPN
easat-10684	225	8	learning	learn	VERB
easat-10684	225	9	gate	gate	NOUN
easat-10684	225	10	3.1.2	3.1.2	PROPN
easat-10684	225	11	.	.	PUNCT
easat-10684	225	12	data	datum	NOUN
easat-10684	225	13	exploration	exploration	NOUN
easat-10684	225	14	and	and	CCONJ
easat-10684	225	15	visualisation	visualisation	NOUN
easat-10684	225	16	this	this	DET
easat-10684	225	17	stage	stage	NOUN
easat-10684	225	18	is	be	AUX
easat-10684	225	19	also	also	ADV
easat-10684	225	20	known	know	VERB
easat-10684	225	21	as	as	ADP
easat-10684	225	22	data	datum	NOUN
easat-10684	225	23	exploratory	exploratory	ADJ
easat-10684	225	24	analysis	analysis	NOUN
easat-10684	225	25	(	(	PUNCT
easat-10684	225	26	eda	eda	PROPN
easat-10684	225	27	)	)	PUNCT
easat-10684	225	28	;	;	PUNCT
easat-10684	225	29	it	it	PRON
easat-10684	225	30	involves	involve	VERB
easat-10684	225	31	conducting	conduct	VERB
easat-10684	225	32	an	an	DET
easat-10684	225	33	in	in	ADP
easat-10684	225	34	-	-	PUNCT
easat-10684	225	35	depth	depth	NOUN
easat-10684	225	36	study	study	NOUN
easat-10684	225	37	of	of	ADP
easat-10684	225	38	the	the	DET
easat-10684	225	39	fake	fake	ADJ
easat-10684	225	40	news	news	NOUN
easat-10684	225	41	dataset	dataset	NOUN
easat-10684	225	42	to	to	PART
easat-10684	225	43	extract	extract	VERB
easat-10684	225	44	important	important	ADJ
easat-10684	225	45	insights	insight	NOUN
easat-10684	225	46	and	and	CCONJ
easat-10684	225	47	explore	explore	VERB
easat-10684	225	48	the	the	DET
easat-10684	225	49	dataset	dataset	NOUN
easat-10684	225	50	structure	structure	NOUN
easat-10684	225	51	.	.	PUNCT
easat-10684	226	1	additionally	additionally	ADV
easat-10684	226	2	,	,	PUNCT
easat-10684	226	3	this	this	DET
easat-10684	226	4	step	step	NOUN
easat-10684	226	5	helps	help	VERB
easat-10684	226	6	uncover	uncover	VERB
easat-10684	226	7	hidden	hide	VERB
easat-10684	226	8	patterns	pattern	NOUN
easat-10684	226	9	,	,	PUNCT
easat-10684	226	10	empty	empty	ADJ
easat-10684	226	11	strings	string	NOUN
easat-10684	226	12	,	,	PUNCT
easat-10684	226	13	duplicate	duplicate	ADJ
easat-10684	226	14	samples	sample	NOUN
easat-10684	226	15	,	,	PUNCT
easat-10684	226	16	stop	stop	VERB
easat-10684	226	17	words	word	NOUN
easat-10684	226	18	,	,	PUNCT
easat-10684	226	19	and	and	CCONJ
easat-10684	226	20	special	special	ADJ
easat-10684	226	21	symbols	symbol	NOUN
easat-10684	226	22	present	present	ADJ
easat-10684	226	23	in	in	ADP
easat-10684	226	24	the	the	DET
easat-10684	226	25	dataset	dataset	NOUN
easat-10684	226	26	.	.	PUNCT
easat-10684	227	1	data	datum	NOUN
easat-10684	227	2	visualization	visualization	NOUN
easat-10684	227	3	refers	refer	VERB
easat-10684	227	4	to	to	ADP
easat-10684	227	5	a	a	DET
easat-10684	227	6	graphical	graphical	ADJ
easat-10684	227	7	representation	representation	NOUN
easat-10684	227	8	of	of	ADP
easat-10684	227	9	information	information	NOUN
easat-10684	227	10	and	and	CCONJ
easat-10684	227	11	exploration	exploration	NOUN
easat-10684	227	12	conducted	conduct	VERB
easat-10684	227	13	on	on	ADP
easat-10684	227	14	the	the	DET
easat-10684	227	15	dataset	dataset	NOUN
easat-10684	227	16	.	.	PUNCT
easat-10684	228	1	this	this	DET
easat-10684	228	2	study	study	NOUN
easat-10684	228	3	will	will	AUX
easat-10684	228	4	use	use	VERB
easat-10684	228	5	various	various	ADJ
easat-10684	228	6	charts	chart	NOUN
easat-10684	228	7	from	from	ADP
easat-10684	228	8	the	the	DET
easat-10684	228	9	python	python	NOUN
easat-10684	228	10	library	library	NOUN
easat-10684	228	11	(	(	PUNCT
easat-10684	228	12	matplotlib.pyplot	matplotlib.pyplot	NOUN
easat-10684	228	13	)	)	PUNCT
easat-10684	228	14	to	to	PART
easat-10684	228	15	perform	perform	VERB
easat-10684	228	16	visual	visual	ADJ
easat-10684	228	17	exploration	exploration	NOUN
easat-10684	228	18	.	.	PUNCT
easat-10684	229	1	firstly	firstly	ADV
easat-10684	229	2	,	,	PUNCT
easat-10684	229	3	the	the	DET
easat-10684	229	4	pandas	panda	NOUN
easat-10684	229	5	‘	'	PUNCT
easat-10684	229	6	read_csv	read_csv	NUM
easat-10684	229	7	’	'	PUNCT
easat-10684	229	8	function	function	NOUN
easat-10684	229	9	is	be	AUX
easat-10684	229	10	used	use	VERB
easat-10684	229	11	to	to	PART
easat-10684	229	12	load	load	VERB
easat-10684	229	13	the	the	DET
easat-10684	229	14	dataset	dataset	NOUN
easat-10684	229	15	into	into	ADP
easat-10684	229	16	a	a	DET
easat-10684	229	17	table	table	NOUN
easat-10684	229	18	of	of	ADP
easat-10684	229	19	rows	row	NOUN
easat-10684	229	20	and	and	CCONJ
easat-10684	229	21	columns	column	NOUN
easat-10684	229	22	,	,	PUNCT
easat-10684	229	23	which	which	PRON
easat-10684	229	24	is	be	AUX
easat-10684	229	25	generally	generally	ADV
easat-10684	229	26	referred	refer	VERB
easat-10684	229	27	to	to	ADP
easat-10684	229	28	as	as	SCONJ
easat-10684	229	29	the	the	DET
easat-10684	229	30	pandas	panda	NOUN
easat-10684	229	31	dataframe	dataframe	ADJ
easat-10684	229	32	.	.	PUNCT
easat-10684	230	1	the	the	DET
easat-10684	230	2	pandas	panda	NOUN
easat-10684	230	3	dataframe	dataframe	PROPN
easat-10684	230	4	's	's	PART
easat-10684	230	5	rows	row	NOUN
easat-10684	230	6	and	and	CCONJ
easat-10684	230	7	columns	column	NOUN
easat-10684	230	8	can	can	AUX
easat-10684	230	9	be	be	AUX
easat-10684	230	10	accessed	access	VERB
easat-10684	230	11	via	via	ADP
easat-10684	230	12	indexing	indexing	NOUN
easat-10684	230	13	,	,	PUNCT
easat-10684	230	14	and	and	CCONJ
easat-10684	230	15	various	various	ADJ
easat-10684	230	16	columns	column	NOUN
easat-10684	230	17	can	can	AUX
easat-10684	230	18	be	be	AUX
easat-10684	230	19	indexed	index	VERB
easat-10684	230	20	using	use	VERB
easat-10684	230	21	column	column	NOUN
easat-10684	230	22	names	name	NOUN
easat-10684	230	23	.	.	PUNCT
easat-10684	231	1	the	the	DET
easat-10684	231	2	sample	sample	NOUN
easat-10684	231	3	fake	fake	ADJ
easat-10684	231	4	news	news	NOUN
easat-10684	231	5	dataset	dataset	NOUN
easat-10684	231	6	is	be	AUX
easat-10684	231	7	indexed	index	VERB
easat-10684	231	8	using	use	VERB
easat-10684	231	9	the	the	DET
easat-10684	231	10	column	column	NOUN
easat-10684	231	11	name	name	NOUN
easat-10684	231	12	‘	'	PUNCT
easat-10684	231	13	text	text	NOUN
easat-10684	231	14	’	'	PUNCT
easat-10684	231	15	,	,	PUNCT
easat-10684	231	16	and	and	CCONJ
easat-10684	231	17	with	with	ADP
easat-10684	231	18	a	a	DET
easat-10684	231	19	numeric	numeric	ADJ
easat-10684	231	20	index	index	NOUN
easat-10684	231	21	of	of	ADP
easat-10684	231	22	‘	'	PUNCT
easat-10684	231	23	0	0	NUM
easat-10684	231	24	’	'	PUNCT
easat-10684	231	25	,	,	PUNCT
easat-10684	231	26	the	the	DET
easat-10684	231	27	first	first	ADJ
easat-10684	231	28	row	row	NOUN
easat-10684	231	29	of	of	ADP
easat-10684	231	30	data	datum	NOUN
easat-10684	231	31	in	in	ADP
easat-10684	231	32	that	that	DET
easat-10684	231	33	column	column	NOUN
easat-10684	231	34	can	can	AUX
easat-10684	231	35	be	be	AUX
easat-10684	231	36	accessed	access	VERB
easat-10684	231	37	.	.	PUNCT
easat-10684	232	1	the	the	DET
easat-10684	232	2	first	first	ADJ
easat-10684	232	3	square	square	ADJ
easat-10684	232	4	bracket	bracket	NOUN
easat-10684	232	5	denotes	denote	NOUN
easat-10684	232	6	the	the	DET
easat-10684	232	7	column	column	NOUN
easat-10684	232	8	name	name	NOUN
easat-10684	232	9	,	,	PUNCT
easat-10684	232	10	while	while	SCONJ
easat-10684	232	11	the	the	DET
easat-10684	232	12	second	second	ADJ
easat-10684	232	13	bracket	bracket	NOUN
easat-10684	232	14	denotes	denote	VERB
easat-10684	232	15	the	the	DET
easat-10684	232	16	row	row	NOUN
easat-10684	232	17	number	number	NOUN
easat-10684	232	18	.	.	PUNCT
easat-10684	233	1	based	base	VERB
easat-10684	233	2	on	on	ADP
easat-10684	233	3	the	the	DET
easat-10684	233	4	sample	sample	NOUN
easat-10684	233	5	text	text	NOUN
easat-10684	233	6	explored	explore	VERB
easat-10684	233	7	,	,	PUNCT
easat-10684	233	8	it	it	PRON
easat-10684	233	9	was	be	AUX
easat-10684	233	10	revealed	reveal	VERB
easat-10684	233	11	that	that	SCONJ
easat-10684	233	12	the	the	DET
easat-10684	233	13	fake	fake	ADJ
easat-10684	233	14	news	news	NOUN
easat-10684	233	15	text	text	NOUN
easat-10684	233	16	samples	sample	NOUN
easat-10684	233	17	contain	contain	VERB
easat-10684	233	18	symbols	symbol	NOUN
easat-10684	233	19	,	,	PUNCT
easat-10684	233	20	punctuations	punctuation	NOUN
easat-10684	233	21	,	,	PUNCT
easat-10684	233	22	escape	escape	NOUN
easat-10684	233	23	characters	character	NOUN
easat-10684	233	24	,	,	PUNCT
easat-10684	233	25	and	and	CCONJ
easat-10684	233	26	stop	stop	VERB
easat-10684	233	27	words	word	NOUN
easat-10684	233	28	that	that	PRON
easat-10684	233	29	have	have	VERB
easat-10684	233	30	no	no	DET
easat-10684	233	31	meaning	meaning	NOUN
easat-10684	233	32	or	or	CCONJ
easat-10684	233	33	contextual	contextual	ADJ
easat-10684	233	34	contribution	contribution	NOUN
easat-10684	233	35	to	to	ADP
easat-10684	233	36	the	the	DET
easat-10684	233	37	fake	fake	ADJ
easat-10684	233	38	news	news	NOUN
easat-10684	233	39	text	text	NOUN
easat-10684	233	40	samples	sample	NOUN
easat-10684	233	41	.	.	PUNCT
easat-10684	234	1	moreover	moreover	ADV
easat-10684	234	2	,	,	PUNCT
easat-10684	234	3	it	it	PRON
easat-10684	234	4	is	be	AUX
easat-10684	234	5	essential	essential	ADJ
easat-10684	234	6	to	to	PART
easat-10684	234	7	explore	explore	VERB
easat-10684	234	8	the	the	DET
easat-10684	234	9	shape	shape	NOUN
easat-10684	234	10	of	of	ADP
easat-10684	234	11	the	the	DET
easat-10684	234	12	dataset	dataset	NOUN
easat-10684	234	13	,	,	PUNCT
easat-10684	234	14	which	which	PRON
easat-10684	234	15	indicates	indicate	VERB
easat-10684	234	16	the	the	DET
easat-10684	234	17	number	number	NOUN
easat-10684	234	18	of	of	ADP
easat-10684	234	19	rows	row	NOUN
easat-10684	234	20	(	(	PUNCT
easat-10684	234	21	samples	sample	NOUN
easat-10684	234	22	)	)	PUNCT
easat-10684	234	23	and	and	CCONJ
easat-10684	234	24	columns	column	NOUN
easat-10684	234	25	(	(	PUNCT
easat-10684	234	26	features	feature	NOUN
easat-10684	234	27	)	)	PUNCT
easat-10684	234	28	in	in	ADP
easat-10684	234	29	the	the	DET
easat-10684	234	30	dataset	dataset	NOUN
easat-10684	234	31	.	.	PUNCT
easat-10684	235	1	figure	figure	NOUN
easat-10684	235	2	2	2	NUM
easat-10684	235	3	exploration	exploration	NOUN
easat-10684	235	4	revealed	reveal	VERB
easat-10684	235	5	that	that	SCONJ
easat-10684	235	6	the	the	DET
easat-10684	235	7	dataset	dataset	NOUN
easat-10684	235	8	contains	contain	VERB
easat-10684	235	9	4,594	4,594	NUM
easat-10684	235	10	news	news	NOUN
easat-10684	235	11	-	-	PUNCT
easat-10684	235	12	labeled	label	VERB
easat-10684	235	13	samples	sample	NOUN
easat-10684	235	14	and	and	CCONJ
easat-10684	235	15	2	2	NUM
easat-10684	235	16	features	feature	NOUN
easat-10684	235	17	(	(	PUNCT
easat-10684	235	18	text	text	NOUN
easat-10684	235	19	,	,	PUNCT
easat-10684	235	20	label	label	NOUN
easat-10684	235	21	)	)	PUNCT
easat-10684	235	22	.	.	PUNCT
easat-10684	236	1	the	the	DET
easat-10684	236	2	labels	label	NOUN
easat-10684	236	3	denote	denote	VERB
easat-10684	236	4	the	the	DET
easat-10684	236	5	response	response	NOUN
easat-10684	236	6	variable	variable	NOUN
easat-10684	236	7	of	of	ADP
easat-10684	236	8	either	either	CCONJ
easat-10684	236	9	fake	fake	ADJ
easat-10684	236	10	or	or	CCONJ
easat-10684	236	11	real	real	ADJ
easat-10684	236	12	news	news	NOUN
easat-10684	236	13	.	.	PUNCT
easat-10684	237	1	figure	figure	NOUN
easat-10684	237	2	2	2	NUM
easat-10684	237	3	.	.	PUNCT
easat-10684	237	4	data	datum	NOUN
easat-10684	237	5	distribution	distribution	NOUN
easat-10684	237	6	based	base	VERB
easat-10684	237	7	on	on	ADP
easat-10684	237	8	label	label	NOUN
easat-10684	237	9	class	class	NOUN
easat-10684	237	10	.	.	PUNCT
easat-10684	238	1	1524	1524	NUM
easat-10684	238	2	edelweiss	edelweiss	PROPN
easat-10684	238	3	applied	apply	VERB
easat-10684	238	4	science	science	NOUN
easat-10684	238	5	and	and	CCONJ
easat-10684	238	6	technology	technology	NOUN
easat-10684	238	7	issn	issn	PROPN
easat-10684	238	8	:	:	PUNCT
easat-10684	238	9	2576	2576	NUM
easat-10684	238	10	-	-	SYM
easat-10684	238	11	8484	8484	NUM
easat-10684	238	12	vol	vol	NOUN
easat-10684	238	13	.	.	PROPN
easat-10684	239	1	9	9	NUM
easat-10684	239	2	,	,	PUNCT
easat-10684	239	3	no	no	INTJ
easat-10684	239	4	.	.	NOUN
easat-10684	239	5	10	10	NUM
easat-10684	239	6	:	:	PUNCT
easat-10684	239	7	1515	1515	NUM
easat-10684	239	8	-	-	SYM
easat-10684	239	9	1540	1540	NUM
easat-10684	239	10	,	,	PUNCT
easat-10684	239	11	2025	2025	NUM
easat-10684	239	12	doi	doi	NOUN
easat-10684	239	13	:	:	PUNCT
easat-10684	239	14	10.55214/2576	10.55214/2576	NUM
easat-10684	239	15	-	-	SYM
easat-10684	239	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	240	1	©	©	PROPN
easat-10684	240	2	2025	2025	NUM
easat-10684	240	3	by	by	ADP
easat-10684	240	4	the	the	DET
easat-10684	240	5	authors	author	NOUN
easat-10684	240	6	;	;	PUNCT
easat-10684	240	7	licensee	licensee	PROPN
easat-10684	240	8	learning	learn	VERB
easat-10684	240	9	gate	gate	PROPN
easat-10684	240	10	figure	figure	NOUN
easat-10684	240	11	2	2	NUM
easat-10684	240	12	shows	show	VERB
easat-10684	240	13	a	a	DET
easat-10684	240	14	bar	bar	NOUN
easat-10684	240	15	chart	chart	NOUN
easat-10684	240	16	revealing	reveal	VERB
easat-10684	240	17	the	the	DET
easat-10684	240	18	label	label	NOUN
easat-10684	240	19	class	class	NOUN
easat-10684	240	20	distribution	distribution	NOUN
easat-10684	240	21	(	(	PUNCT
easat-10684	240	22	fake	fake	ADJ
easat-10684	240	23	sample	sample	NOUN
easat-10684	240	24	count	count	NOUN
easat-10684	240	25	and	and	CCONJ
easat-10684	240	26	real	real	ADJ
easat-10684	240	27	sample	sample	NOUN
easat-10684	240	28	count	count	NOUN
easat-10684	240	29	)	)	PUNCT
easat-10684	240	30	.	.	PUNCT
easat-10684	241	1	this	this	PRON
easat-10684	241	2	reveals	reveal	VERB
easat-10684	241	3	that	that	SCONJ
easat-10684	241	4	the	the	DET
easat-10684	241	5	real	real	ADJ
easat-10684	241	6	and	and	CCONJ
easat-10684	241	7	fake	fake	ADJ
easat-10684	241	8	classes	class	NOUN
easat-10684	241	9	contain	contain	VERB
easat-10684	241	10	equal	equal	ADJ
easat-10684	241	11	numbers	number	NOUN
easat-10684	241	12	of	of	ADP
easat-10684	241	13	data	datum	NOUN
easat-10684	241	14	class	class	NOUN
easat-10684	241	15	groups	group	NOUN
easat-10684	241	16	(	(	PUNCT
easat-10684	241	17	fake	fake	ADJ
easat-10684	241	18	and	and	CCONJ
easat-10684	241	19	real	real	ADJ
easat-10684	241	20	)	)	PUNCT
easat-10684	241	21	.	.	PUNCT
easat-10684	242	1	each	each	DET
easat-10684	242	2	class	class	NOUN
easat-10684	242	3	contains	contain	VERB
easat-10684	242	4	2,297	2,297	NUM
easat-10684	242	5	samples	sample	NOUN
easat-10684	242	6	.	.	PUNCT
easat-10684	243	1	the	the	DET
easat-10684	243	2	exploration	exploration	NOUN
easat-10684	243	3	conducted	conduct	VERB
easat-10684	243	4	shows	show	VERB
easat-10684	243	5	that	that	SCONJ
easat-10684	243	6	the	the	DET
easat-10684	243	7	dataset	dataset	NOUN
easat-10684	243	8	is	be	AUX
easat-10684	243	9	balanced	balanced	ADJ
easat-10684	243	10	.	.	PUNCT
easat-10684	244	1	figure	figure	VERB
easat-10684	244	2	3	3	NUM
easat-10684	244	3	.	.	NOUN
easat-10684	244	4	word	word	NOUN
easat-10684	244	5	frequency	frequency	NOUN
easat-10684	244	6	usage	usage	NOUN
easat-10684	244	7	(	(	PUNCT
easat-10684	244	8	real	real	ADJ
easat-10684	244	9	news	news	NOUN
easat-10684	244	10	class	class	NOUN
easat-10684	244	11	)	)	PUNCT
easat-10684	244	12	.	.	PUNCT
easat-10684	245	1	figure	figure	VERB
easat-10684	245	2	4	4	NUM
easat-10684	245	3	.	.	NOUN
easat-10684	245	4	word	word	NOUN
easat-10684	245	5	frequency	frequency	NOUN
easat-10684	245	6	usage	usage	NOUN
easat-10684	245	7	(	(	PUNCT
easat-10684	245	8	fake	fake	ADJ
easat-10684	245	9	news	news	NOUN
easat-10684	245	10	class	class	NOUN
easat-10684	245	11	)	)	PUNCT
easat-10684	245	12	.	.	PUNCT
easat-10684	246	1	the	the	DET
easat-10684	246	2	frequency	frequency	NOUN
easat-10684	246	3	count	count	NOUN
easat-10684	246	4	analysis	analysis	NOUN
easat-10684	246	5	provides	provide	VERB
easat-10684	246	6	a	a	DET
easat-10684	246	7	statistical	statistical	ADJ
easat-10684	246	8	distribution	distribution	NOUN
easat-10684	246	9	of	of	ADP
easat-10684	246	10	vocabulary	vocabulary	ADJ
easat-10684	246	11	usage	usage	NOUN
easat-10684	246	12	.	.	PUNCT
easat-10684	247	1	with	with	ADP
easat-10684	247	2	the	the	DET
easat-10684	247	3	analysis	analysis	NOUN
easat-10684	247	4	,	,	PUNCT
easat-10684	247	5	the	the	DET
easat-10684	247	6	most	most	ADV
easat-10684	247	7	frequently	frequently	ADV
easat-10684	247	8	used	use	VERB
easat-10684	247	9	terms	term	NOUN
easat-10684	247	10	in	in	ADP
easat-10684	247	11	fake	fake	ADJ
easat-10684	247	12	news	news	NOUN
easat-10684	247	13	samples	sample	NOUN
easat-10684	247	14	and	and	CCONJ
easat-10684	247	15	real	real	ADJ
easat-10684	247	16	news	news	NOUN
easat-10684	247	17	samples	sample	NOUN
easat-10684	247	18	can	can	AUX
easat-10684	247	19	be	be	AUX
easat-10684	247	20	seen	see	VERB
easat-10684	247	21	using	use	VERB
easat-10684	247	22	the	the	DET
easat-10684	247	23	frequency	frequency	NOUN
easat-10684	247	24	line	line	NOUN
easat-10684	247	25	plot	plot	NOUN
easat-10684	247	26	in	in	ADP
easat-10684	247	27	figures	figure	NOUN
easat-10684	247	28	3	3	NUM
easat-10684	247	29	and	and	CCONJ
easat-10684	247	30	4	4	NUM
easat-10684	247	31	.	.	PUNCT
easat-10684	248	1	the	the	DET
easat-10684	248	2	line	line	NOUN
easat-10684	248	3	chart	chart	NOUN
easat-10684	248	4	represents	represent	VERB
easat-10684	248	5	the	the	DET
easat-10684	248	6	word	word	NOUN
easat-10684	248	7	count	count	NOUN
easat-10684	248	8	or	or	CCONJ
easat-10684	248	9	frequency	frequency	NOUN
easat-10684	248	10	.	.	PUNCT
easat-10684	249	1	figure	figure	NOUN
easat-10684	249	2	3	3	NUM
easat-10684	249	3	revealed	reveal	VERB
easat-10684	249	4	that	that	SCONJ
easat-10684	249	5	the	the	DET
easat-10684	249	6	word	word	NOUN
easat-10684	249	7	‘	'	PUNCT
easat-10684	249	8	trump	trump	NOUN
easat-10684	249	9	’	'	PUNCT
easat-10684	249	10	is	be	AUX
easat-10684	249	11	the	the	DET
easat-10684	249	12	most	most	ADV
easat-10684	249	13	frequently	frequently	ADV
easat-10684	249	14	used	use	VERB
easat-10684	249	15	term	term	NOUN
easat-10684	249	16	in	in	ADP
easat-10684	249	17	real	real	ADJ
easat-10684	249	18	news	news	NOUN
easat-10684	249	19	samples	sample	NOUN
easat-10684	249	20	,	,	PUNCT
easat-10684	249	21	followed	follow	VERB
easat-10684	249	22	by	by	ADP
easat-10684	249	23	‘	'	PUNCT
easat-10684	249	24	clinton	clinton	PROPN
easat-10684	249	25	’	'	PUNCT
easat-10684	249	26	,	,	PUNCT
easat-10684	249	27	‘	'	PUNCT
easat-10684	249	28	said	say	VERB
easat-10684	249	29	’	'	PUNCT
easat-10684	249	30	,	,	PUNCT
easat-10684	249	31	‘	'	PUNCT
easat-10684	249	32	state	state	NOUN
easat-10684	249	33	’	'	PUNCT
easat-10684	249	34	,	,	PUNCT
easat-10684	249	35	‘	'	PUNCT
easat-10684	249	36	republican	republican	ADJ
easat-10684	249	37	’	'	PUNCT
easat-10684	249	38	,	,	PUNCT
easat-10684	249	39	and	and	CCONJ
easat-10684	249	40	‘	'	PUNCT
easat-10684	249	41	i	i	PRON
easat-10684	249	42	’	'	PUNCT
easat-10684	249	43	,	,	PUNCT
easat-10684	249	44	with	with	ADP
easat-10684	249	45	a	a	DET
easat-10684	249	46	frequency	frequency	NOUN
easat-10684	249	47	of	of	ADP
easat-10684	249	48	above	above	ADP
easat-10684	249	49	7,500	7,500	NUM
easat-10684	249	50	.	.	PUNCT
easat-10684	250	1	while	while	SCONJ
easat-10684	250	2	‘	'	PUNCT
easat-10684	250	3	the	the	PRON
easat-10684	250	4	’	'	PUNCT
easat-10684	250	5	is	be	AUX
easat-10684	250	6	the	the	DET
easat-10684	250	7	most	most	ADV
easat-10684	250	8	frequently	frequently	ADV
easat-10684	250	9	used	use	VERB
easat-10684	250	10	term	term	NOUN
easat-10684	250	11	in	in	ADP
easat-10684	250	12	fake	fake	ADJ
easat-10684	250	13	news	news	NOUN
easat-10684	250	14	samples	sample	NOUN
easat-10684	250	15	,	,	PUNCT
easat-10684	250	16	as	as	SCONJ
easat-10684	250	17	shown	show	VERB
easat-10684	250	18	in	in	ADP
easat-10684	250	19	figure	figure	NOUN
easat-10684	250	20	4	4	NUM
easat-10684	250	21	,	,	PUNCT
easat-10684	250	22	followed	follow	VERB
easat-10684	250	23	by	by	ADP
easat-10684	250	24	‘	'	PUNCT
easat-10684	250	25	i	i	PRON
easat-10684	250	26	’	'	PUNCT
easat-10684	250	27	,	,	PUNCT
easat-10684	250	28	‘	'	PUNCT
easat-10684	250	29	clinton	clinton	PROPN
easat-10684	250	30	’	'	PUNCT
easat-10684	250	31	,	,	PUNCT
easat-10684	250	32	‘	'	PUNCT
easat-10684	250	33	trump	trump	NOUN
easat-10684	250	34	’	'	PUNCT
easat-10684	250	35	,	,	PUNCT
easat-10684	250	36	‘	'	PUNCT
easat-10684	250	37	state	state	NOUN
easat-10684	250	38	’	'	PUNCT
easat-10684	250	39	,	,	PUNCT
easat-10684	250	40	and	and	CCONJ
easat-10684	250	41	‘	'	PUNCT
easat-10684	250	42	people	people	NOUN
easat-10684	250	43	,	,	PUNCT
easat-10684	250	44	’	'	PUNCT
easat-10684	250	45	with	with	ADP
easat-10684	250	46	a	a	DET
easat-10684	250	47	frequency	frequency	NOUN
easat-10684	250	48	of	of	ADP
easat-10684	250	49	above	above	ADP
easat-10684	250	50	4,000	4,000	NUM
easat-10684	250	51	.	.	PUNCT
easat-10684	251	1	1525	1525	NUM
easat-10684	251	2	edelweiss	edelweiss	PROPN
easat-10684	251	3	applied	apply	VERB
easat-10684	251	4	science	science	NOUN
easat-10684	251	5	and	and	CCONJ
easat-10684	251	6	technology	technology	NOUN
easat-10684	251	7	issn	issn	PROPN
easat-10684	251	8	:	:	PUNCT
easat-10684	251	9	2576	2576	NUM
easat-10684	251	10	-	-	SYM
easat-10684	251	11	8484	8484	NUM
easat-10684	251	12	vol	vol	NOUN
easat-10684	251	13	.	.	PROPN
easat-10684	252	1	9	9	NUM
easat-10684	252	2	,	,	PUNCT
easat-10684	252	3	no	no	INTJ
easat-10684	252	4	.	.	NOUN
easat-10684	252	5	10	10	NUM
easat-10684	252	6	:	:	PUNCT
easat-10684	252	7	1515	1515	NUM
easat-10684	252	8	-	-	SYM
easat-10684	252	9	1540	1540	NUM
easat-10684	252	10	,	,	PUNCT
easat-10684	252	11	2025	2025	NUM
easat-10684	252	12	doi	doi	NOUN
easat-10684	252	13	:	:	PUNCT
easat-10684	252	14	10.55214/2576	10.55214/2576	NUM
easat-10684	252	15	-	-	SYM
easat-10684	252	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	253	1	©	©	PROPN
easat-10684	253	2	2025	2025	NUM
easat-10684	253	3	by	by	ADP
easat-10684	253	4	the	the	DET
easat-10684	253	5	authors	author	NOUN
easat-10684	253	6	;	;	PUNCT
easat-10684	254	1	licensee	licensee	PROPN
easat-10684	254	2	learning	learn	VERB
easat-10684	254	3	gate	gate	PROPN
easat-10684	254	4	figure	figure	NOUN
easat-10684	254	5	5	5	NUM
easat-10684	254	6	.	.	PUNCT
easat-10684	255	1	wordcloud	wordcloud	PROPN
easat-10684	255	2	representation	representation	NOUN
easat-10684	255	3	of	of	ADP
easat-10684	255	4	fake	fake	ADJ
easat-10684	255	5	news	news	NOUN
easat-10684	255	6	and	and	CCONJ
easat-10684	255	7	real	real	ADJ
easat-10684	255	8	news	news	NOUN
easat-10684	255	9	.	.	PUNCT
easat-10684	256	1	figure	figure	NOUN
easat-10684	256	2	5	5	NUM
easat-10684	256	3	shows	show	VERB
easat-10684	256	4	the	the	DET
easat-10684	256	5	wordcloud	wordcloud	PROPN
easat-10684	256	6	representation	representation	NOUN
easat-10684	256	7	of	of	ADP
easat-10684	256	8	word	word	NOUN
easat-10684	256	9	frequency	frequency	NOUN
easat-10684	256	10	usage	usage	NOUN
easat-10684	256	11	in	in	ADP
easat-10684	256	12	both	both	CCONJ
easat-10684	256	13	real	real	ADJ
easat-10684	256	14	and	and	CCONJ
easat-10684	256	15	fake	fake	ADJ
easat-10684	256	16	news	news	NOUN
easat-10684	256	17	vocabulary	vocabulary	NOUN
easat-10684	256	18	.	.	PUNCT
easat-10684	257	1	the	the	DET
easat-10684	257	2	word	word	NOUN
easat-10684	257	3	cloud	cloud	NOUN
easat-10684	257	4	uses	use	VERB
easat-10684	257	5	text	text	NOUN
easat-10684	257	6	size	size	NOUN
easat-10684	257	7	to	to	PART
easat-10684	257	8	describe	describe	VERB
easat-10684	257	9	or	or	CCONJ
easat-10684	257	10	explore	explore	VERB
easat-10684	257	11	how	how	SCONJ
easat-10684	257	12	words	word	NOUN
easat-10684	257	13	are	be	AUX
easat-10684	257	14	frequently	frequently	ADV
easat-10684	257	15	used	use	VERB
easat-10684	257	16	.	.	PUNCT
easat-10684	258	1	the	the	PRON
easat-10684	258	2	larger	large	ADJ
easat-10684	258	3	the	the	DET
easat-10684	258	4	text	text	NOUN
easat-10684	258	5	,	,	PUNCT
easat-10684	258	6	the	the	PRON
easat-10684	258	7	more	more	ADV
easat-10684	258	8	frequently	frequently	ADV
easat-10684	258	9	the	the	DET
easat-10684	258	10	vocabulary	vocabulary	NOUN
easat-10684	258	11	appears	appear	VERB
easat-10684	258	12	.	.	PUNCT
easat-10684	259	1	based	base	VERB
easat-10684	259	2	on	on	ADP
easat-10684	259	3	figure	figure	NOUN
easat-10684	259	4	5	5	NUM
easat-10684	259	5	,	,	PUNCT
easat-10684	259	6	it	it	PRON
easat-10684	259	7	shows	show	VERB
easat-10684	259	8	that	that	SCONJ
easat-10684	259	9	the	the	DET
easat-10684	259	10	words	word	NOUN
easat-10684	259	11	‘	'	PUNCT
easat-10684	259	12	trump	trump	NOUN
easat-10684	259	13	,	,	PUNCT
easat-10684	259	14	’	'	PUNCT
easat-10684	259	15	‘	'	PUNCT
easat-10684	259	16	state	state	NOUN
easat-10684	259	17	,	,	PUNCT
easat-10684	259	18	’	'	PUNCT
easat-10684	259	19	‘	'	PUNCT
easat-10684	259	20	clinton	clinton	PROPN
easat-10684	259	21	,	,	PUNCT
easat-10684	259	22	’	'	PUNCT
easat-10684	259	23	and	and	CCONJ
easat-10684	259	24	‘	'	PUNCT
easat-10684	259	25	said	say	VERB
easat-10684	259	26	’	'	PUNCT
easat-10684	259	27	are	be	AUX
easat-10684	259	28	bigger	big	ADJ
easat-10684	259	29	in	in	ADP
easat-10684	259	30	the	the	DET
easat-10684	259	31	text	text	NOUN
easat-10684	259	32	,	,	PUNCT
easat-10684	259	33	revealing	reveal	VERB
easat-10684	259	34	that	that	SCONJ
easat-10684	259	35	they	they	PRON
easat-10684	259	36	are	be	AUX
easat-10684	259	37	mostly	mostly	ADV
easat-10684	259	38	used	use	VERB
easat-10684	259	39	in	in	ADP
easat-10684	259	40	real	real	ADJ
easat-10684	259	41	news	news	NOUN
easat-10684	259	42	vocabulary	vocabulary	NOUN
easat-10684	259	43	.	.	PUNCT
easat-10684	260	1	however	however	ADV
easat-10684	260	2	,	,	PUNCT
easat-10684	260	3	the	the	DET
easat-10684	260	4	same	same	ADJ
easat-10684	260	5	figure	figure	NOUN
easat-10684	260	6	reveals	reveal	VERB
easat-10684	260	7	that	that	SCONJ
easat-10684	260	8	the	the	DET
easat-10684	260	9	terms	term	NOUN
easat-10684	260	10	‘	'	PUNCT
easat-10684	260	11	the	the	PRON
easat-10684	260	12	,	,	PUNCT
easat-10684	260	13	’	'	PUNCT
easat-10684	260	14	‘	'	PUNCT
easat-10684	260	15	i	i	PRON
easat-10684	260	16	,	,	PUNCT
easat-10684	260	17	’	'	PUNCT
easat-10684	260	18	‘	'	PUNCT
easat-10684	260	19	clinton	clinton	PROPN
easat-10684	260	20	,	,	PUNCT
easat-10684	260	21	’	'	PUNCT
easat-10684	260	22	‘	'	PUNCT
easat-10684	260	23	state	state	NOUN
easat-10684	260	24	,	,	PUNCT
easat-10684	260	25	’	'	PUNCT
easat-10684	260	26	‘	'	PUNCT
easat-10684	260	27	people	people	NOUN
easat-10684	260	28	,	,	PUNCT
easat-10684	260	29	’	'	PUNCT
easat-10684	260	30	‘	'	PUNCT
easat-10684	260	31	us	we	PRON
easat-10684	260	32	,	,	PUNCT
easat-10684	260	33	’	'	PUNCT
easat-10684	260	34	and	and	CCONJ
easat-10684	260	35	the	the	DET
easat-10684	260	36	like	like	NOUN
easat-10684	260	37	are	be	AUX
easat-10684	260	38	mostly	mostly	ADV
easat-10684	260	39	used	use	VERB
easat-10684	260	40	in	in	ADP
easat-10684	260	41	fake	fake	ADJ
easat-10684	260	42	news	news	NOUN
easat-10684	260	43	vocabulary	vocabulary	NOUN
easat-10684	260	44	.	.	PUNCT
easat-10684	261	1	3.1.3	3.1.3	NUM
easat-10684	261	2	.	.	PUNCT
easat-10684	261	3	data	datum	NOUN
easat-10684	261	4	preprocessing	preprocessing	NOUN
easat-10684	261	5	and	and	CCONJ
easat-10684	261	6	cleaning	clean	VERB
easat-10684	261	7	the	the	DET
easat-10684	261	8	natural	natural	ADJ
easat-10684	261	9	language	language	NOUN
easat-10684	261	10	tool	tool	NOUN
easat-10684	261	11	kit	kit	NOUN
easat-10684	261	12	(	(	PUNCT
easat-10684	261	13	nltk	nltk	PROPN
easat-10684	261	14	)	)	PUNCT
easat-10684	261	15	is	be	AUX
easat-10684	261	16	an	an	DET
easat-10684	261	17	essential	essential	ADJ
easat-10684	261	18	tool	tool	NOUN
easat-10684	261	19	for	for	ADP
easat-10684	261	20	solving	solve	VERB
easat-10684	261	21	natural	natural	ADJ
easat-10684	261	22	language	language	NOUN
easat-10684	261	23	processing	processing	NOUN
easat-10684	261	24	tasks	task	NOUN
easat-10684	261	25	[	[	X
easat-10684	261	26	8	8	NUM
easat-10684	261	27	]	]	PUNCT
easat-10684	261	28	.	.	PUNCT
easat-10684	262	1	data	datum	NOUN
easat-10684	262	2	preprocessing	preprocessing	NOUN
easat-10684	262	3	and	and	CCONJ
easat-10684	262	4	cleaning	cleaning	NOUN
easat-10684	262	5	are	be	AUX
easat-10684	262	6	conducted	conduct	VERB
easat-10684	262	7	using	use	VERB
easat-10684	262	8	modules	module	NOUN
easat-10684	262	9	(	(	PUNCT
easat-10684	262	10	stopwords	stopword	NOUN
easat-10684	262	11	,	,	PUNCT
easat-10684	262	12	porterstemmer	porterstemmer	NOUN
easat-10684	262	13	,	,	PUNCT
easat-10684	262	14	and	and	CCONJ
easat-10684	262	15	tokenizer	tokenizer	NOUN
easat-10684	262	16	)	)	PUNCT
easat-10684	262	17	from	from	ADP
easat-10684	262	18	the	the	DET
easat-10684	262	19	nltk	nltk	PROPN
easat-10684	262	20	library	library	NOUN
easat-10684	262	21	.	.	PUNCT
easat-10684	263	1	the	the	DET
easat-10684	263	2	cleaning	cleaning	NOUN
easat-10684	263	3	process	process	NOUN
easat-10684	263	4	is	be	AUX
easat-10684	263	5	based	base	VERB
easat-10684	263	6	on	on	ADP
easat-10684	263	7	the	the	DET
easat-10684	263	8	exploration	exploration	NOUN
easat-10684	263	9	previously	previously	ADV
easat-10684	263	10	conducted	conduct	VERB
easat-10684	263	11	on	on	ADP
easat-10684	263	12	the	the	DET
easat-10684	263	13	dataset	dataset	NOUN
easat-10684	263	14	.	.	PUNCT
easat-10684	264	1	the	the	DET
easat-10684	264	2	cleaning	clean	VERB
easat-10684	264	3	steps	step	NOUN
easat-10684	264	4	include	include	VERB
easat-10684	264	5	removing	remove	VERB
easat-10684	264	6	stopwords	stopword	NOUN
easat-10684	264	7	,	,	PUNCT
easat-10684	264	8	punctuation	punctuation	NOUN
easat-10684	264	9	marks	mark	NOUN
easat-10684	264	10	,	,	PUNCT
easat-10684	264	11	special	special	ADJ
easat-10684	264	12	symbols	symbol	NOUN
easat-10684	264	13	,	,	PUNCT
easat-10684	264	14	and	and	CCONJ
easat-10684	264	15	special	special	ADJ
easat-10684	264	16	characters	character	NOUN
easat-10684	264	17	,	,	PUNCT
easat-10684	264	18	while	while	SCONJ
easat-10684	264	19	the	the	DET
easat-10684	264	20	preprocessing	preprocessing	NOUN
easat-10684	264	21	activities	activity	NOUN
easat-10684	264	22	include	include	VERB
easat-10684	264	23	stemming	stem	VERB
easat-10684	264	24	,	,	PUNCT
easat-10684	264	25	tokenizing	tokenizing	NOUN
easat-10684	264	26	,	,	PUNCT
easat-10684	264	27	and	and	CCONJ
easat-10684	264	28	text	text	NOUN
easat-10684	264	29	case	case	NOUN
easat-10684	264	30	conversion	conversion	NOUN
easat-10684	264	31	.	.	PUNCT
easat-10684	265	1	furthermore	furthermore	ADV
easat-10684	265	2	,	,	PUNCT
easat-10684	265	3	to	to	PART
easat-10684	265	4	perform	perform	VERB
easat-10684	265	5	the	the	DET
easat-10684	265	6	preprocessing	preprocessing	NOUN
easat-10684	265	7	operation	operation	NOUN
easat-10684	265	8	,	,	PUNCT
easat-10684	265	9	three	three	NUM
easat-10684	265	10	helper	helper	NOUN
easat-10684	265	11	functions	function	NOUN
easat-10684	265	12	(	(	PUNCT
easat-10684	265	13	remove_punctuation	remove_punctuation	NOUN
easat-10684	265	14	,	,	PUNCT
easat-10684	265	15	remove_stopwords	remove_stopword	NOUN
easat-10684	265	16	,	,	PUNCT
easat-10684	265	17	and	and	CCONJ
easat-10684	265	18	stem_and_lower_text	stem_and_lower_text	PROPN
easat-10684	265	19	)	)	PUNCT
easat-10684	265	20	are	be	AUX
easat-10684	265	21	created	create	VERB
easat-10684	265	22	to	to	PART
easat-10684	265	23	remove	remove	VERB
easat-10684	265	24	punctuations	punctuation	NOUN
easat-10684	265	25	,	,	PUNCT
easat-10684	265	26	special	special	ADJ
easat-10684	265	27	characters	character	NOUN
easat-10684	265	28	,	,	PUNCT
easat-10684	265	29	stop	stop	VERB
easat-10684	265	30	words	word	NOUN
easat-10684	265	31	,	,	PUNCT
easat-10684	265	32	and	and	CCONJ
easat-10684	265	33	reduce	reduce	VERB
easat-10684	265	34	similar	similar	ADJ
easat-10684	265	35	vocabulary	vocabulary	NOUN
easat-10684	265	36	to	to	ADP
easat-10684	265	37	their	their	PRON
easat-10684	265	38	stem	stem	NOUN
easat-10684	265	39	words	word	NOUN
easat-10684	265	40	using	use	VERB
easat-10684	265	41	porterstemmer	porterstemmer	NOUN
easat-10684	265	42	from	from	ADP
easat-10684	265	43	the	the	DET
easat-10684	265	44	nltk	nltk	PROPN
easat-10684	265	45	library	library	NOUN
easat-10684	265	46	.	.	PUNCT
easat-10684	266	1	moreover	moreover	ADV
easat-10684	266	2	,	,	PUNCT
easat-10684	266	3	the	the	DET
easat-10684	266	4	label	label	NOUN
easat-10684	266	5	encoding	encoding	NOUN
easat-10684	266	6	involves	involve	VERB
easat-10684	266	7	the	the	DET
easat-10684	266	8	transformation	transformation	NOUN
easat-10684	266	9	of	of	ADP
easat-10684	266	10	the	the	DET
easat-10684	266	11	fake	fake	ADJ
easat-10684	266	12	news	news	NOUN
easat-10684	266	13	label	label	NOUN
easat-10684	266	14	(	(	PUNCT
easat-10684	266	15	fake	fake	ADJ
easat-10684	266	16	,	,	PUNCT
easat-10684	266	17	true	true	ADJ
easat-10684	266	18	)	)	PUNCT
easat-10684	266	19	into	into	ADP
easat-10684	266	20	a	a	DET
easat-10684	266	21	numeric	numeric	ADJ
easat-10684	266	22	representation	representation	NOUN
easat-10684	266	23	for	for	ADP
easat-10684	266	24	machine	machine	NOUN
easat-10684	266	25	learning	learn	VERB
easat-10684	266	26	processing	processing	NOUN
easat-10684	266	27	.	.	PUNCT
easat-10684	267	1	in	in	ADP
easat-10684	267	2	this	this	DET
easat-10684	267	3	study	study	NOUN
easat-10684	267	4	,	,	PUNCT
easat-10684	267	5	the	the	DET
easat-10684	267	6	fake	fake	ADJ
easat-10684	267	7	label	label	NOUN
easat-10684	267	8	entry	entry	NOUN
easat-10684	267	9	is	be	AUX
easat-10684	267	10	transformed	transform	VERB
easat-10684	267	11	into	into	ADP
easat-10684	267	12	zeros	zero	NOUN
easat-10684	267	13	(	(	PUNCT
easat-10684	267	14	0	0	NUM
easat-10684	267	15	)	)	PUNCT
easat-10684	267	16	,	,	PUNCT
easat-10684	267	17	while	while	SCONJ
easat-10684	267	18	the	the	DET
easat-10684	267	19	true	true	ADJ
easat-10684	267	20	entries	entry	NOUN
easat-10684	267	21	are	be	AUX
easat-10684	267	22	transformed	transform	VERB
easat-10684	267	23	into	into	ADP
easat-10684	267	24	ones	one	NOUN
easat-10684	267	25	(	(	PUNCT
easat-10684	267	26	1	1	NUM
easat-10684	267	27	)	)	PUNCT
easat-10684	267	28	.	.	PUNCT
easat-10684	268	1	the	the	DET
easat-10684	268	2	resulting	result	VERB
easat-10684	268	3	data	datum	NOUN
easat-10684	268	4	frame	frame	NOUN
easat-10684	268	5	after	after	ADP
easat-10684	268	6	transformation	transformation	NOUN
easat-10684	268	7	is	be	AUX
easat-10684	268	8	further	far	ADV
easat-10684	268	9	processed	process	VERB
easat-10684	268	10	during	during	ADP
easat-10684	268	11	feature	feature	NOUN
easat-10684	268	12	extraction	extraction	NOUN
easat-10684	268	13	.	.	PUNCT
easat-10684	269	1	3.1.4	3.1.4	NUM
easat-10684	269	2	.	.	PUNCT
easat-10684	269	3	feature	feature	NOUN
easat-10684	269	4	extraction	extraction	NOUN
easat-10684	269	5	(	(	PUNCT
easat-10684	269	6	wordembeddings	wordembedding	NOUN
easat-10684	269	7	)	)	PUNCT
easat-10684	269	8	the	the	DET
easat-10684	269	9	process	process	NOUN
easat-10684	269	10	of	of	ADP
easat-10684	269	11	transforming	transform	VERB
easat-10684	269	12	text	text	NOUN
easat-10684	269	13	data	datum	NOUN
easat-10684	269	14	into	into	ADP
easat-10684	269	15	numeric	numeric	ADJ
easat-10684	269	16	vectors	vector	NOUN
easat-10684	269	17	is	be	AUX
easat-10684	269	18	known	know	VERB
easat-10684	269	19	as	as	ADP
easat-10684	269	20	vectorization	vectorization	NOUN
easat-10684	269	21	.	.	PUNCT
easat-10684	270	1	however	however	ADV
easat-10684	270	2	,	,	PUNCT
easat-10684	270	3	word	word	NOUN
easat-10684	270	4	embedding	embed	VERB
easat-10684	270	5	refers	refer	NOUN
easat-10684	270	6	to	to	ADP
easat-10684	270	7	feature	feature	NOUN
easat-10684	270	8	vectors	vector	NOUN
easat-10684	270	9	generated	generate	VERB
easat-10684	270	10	using	use	VERB
easat-10684	270	11	deep	deep	ADJ
easat-10684	270	12	learning	learning	NOUN
easat-10684	270	13	-	-	PUNCT
easat-10684	270	14	based	base	VERB
easat-10684	270	15	architectures	architecture	NOUN
easat-10684	270	16	.	.	PUNCT
easat-10684	271	1	standard	standard	ADJ
easat-10684	271	2	techniques	technique	NOUN
easat-10684	271	3	used	use	VERB
easat-10684	271	4	include	include	VERB
easat-10684	271	5	word2vec	word2vec	NOUN
easat-10684	271	6	,	,	PUNCT
easat-10684	271	7	fasttext	fasttext	NOUN
easat-10684	271	8	,	,	PUNCT
easat-10684	271	9	and	and	CCONJ
easat-10684	271	10	glove	glove	NOUN
easat-10684	271	11	[	[	X
easat-10684	271	12	43	43	NUM
easat-10684	271	13	]	]	PUNCT
easat-10684	271	14	.	.	PUNCT
easat-10684	272	1	this	this	DET
easat-10684	272	2	study	study	NOUN
easat-10684	272	3	utilises	utilise	VERB
easat-10684	272	4	the	the	DET
easat-10684	272	5	word2vec	word2vec	X
easat-10684	272	6	(	(	PUNCT
easat-10684	272	7	cbow	cbow	VERB
easat-10684	272	8	)	)	PUNCT
easat-10684	272	9	technique	technique	NOUN
easat-10684	272	10	for	for	ADP
easat-10684	272	11	generating	generate	VERB
easat-10684	272	12	2	2	NUM
easat-10684	272	13	-	-	PUNCT
easat-10684	272	14	dimensional	dimensional	ADJ
easat-10684	272	15	features	feature	NOUN
easat-10684	272	16	from	from	ADP
easat-10684	272	17	the	the	DET
easat-10684	272	18	fake	fake	ADJ
easat-10684	272	19	news	news	NOUN
easat-10684	272	20	dataset	dataset	NOUN
easat-10684	272	21	.	.	PUNCT
easat-10684	273	1	3.1.4.1	3.1.4.1	X
easat-10684	273	2	.	.	PUNCT
easat-10684	274	1	word2vec	word2vec	X
easat-10684	274	2	(	(	PUNCT
easat-10684	274	3	cbow	cbow	VERB
easat-10684	274	4	)	)	PUNCT
easat-10684	274	5	it	it	PRON
easat-10684	274	6	is	be	AUX
easat-10684	274	7	essential	essential	ADJ
easat-10684	274	8	to	to	PART
easat-10684	274	9	extract	extract	VERB
easat-10684	274	10	numerical	numerical	ADJ
easat-10684	274	11	features	feature	NOUN
easat-10684	274	12	from	from	ADP
easat-10684	274	13	textual	textual	ADJ
easat-10684	274	14	data	datum	NOUN
easat-10684	274	15	for	for	ADP
easat-10684	274	16	proper	proper	ADJ
easat-10684	274	17	machine	machine	NOUN
easat-10684	274	18	learning	learn	VERB
easat-10684	274	19	computation	computation	NOUN
easat-10684	274	20	.	.	PUNCT
easat-10684	275	1	this	this	DET
easat-10684	275	2	study	study	NOUN
easat-10684	275	3	utilized	utilize	VERB
easat-10684	275	4	the	the	DET
easat-10684	275	5	word2vec	word2vec	X
easat-10684	275	6	(	(	PUNCT
easat-10684	275	7	cbow	cbow	PROPN
easat-10684	275	8	)	)	PUNCT
easat-10684	275	9	model	model	NOUN
easat-10684	275	10	from	from	ADP
easat-10684	275	11	the	the	DET
easat-10684	275	12	gensim	gensim	NOUN
easat-10684	275	13	library	library	NOUN
easat-10684	275	14	to	to	PART
easat-10684	275	15	extract	extract	VERB
easat-10684	275	16	feature	feature	NOUN
easat-10684	275	17	vectors	vector	NOUN
easat-10684	275	18	for	for	ADP
easat-10684	275	19	each	each	DET
easat-10684	275	20	vocabulary	vocabulary	NOUN
easat-10684	275	21	with	with	ADP
easat-10684	275	22	300	300	NUM
easat-10684	275	23	vectors	vector	NOUN
easat-10684	275	24	(	(	PUNCT
easat-10684	275	25	dimensions	dimension	NOUN
easat-10684	275	26	)	)	PUNCT
easat-10684	275	27	.	.	PUNCT
easat-10684	276	1	however	however	ADV
easat-10684	276	2	,	,	PUNCT
easat-10684	276	3	the	the	DET
easat-10684	276	4	feature	feature	NOUN
easat-10684	276	5	vector	vector	NOUN
easat-10684	276	6	of	of	ADP
easat-10684	276	7	‘	'	PUNCT
easat-10684	276	8	trump	trump	NOUN
easat-10684	276	9	’	'	PUNCT
easat-10684	276	10	is	be	AUX
easat-10684	276	11	explored	explore	VERB
easat-10684	276	12	,	,	PUNCT
easat-10684	276	13	and	and	CCONJ
easat-10684	276	14	it	it	PRON
easat-10684	276	15	is	be	AUX
easat-10684	276	16	revealed	reveal	VERB
easat-10684	276	17	that	that	SCONJ
easat-10684	276	18	each	each	DET
easat-10684	276	19	feature	feature	NOUN
easat-10684	276	20	vector	vector	NOUN
easat-10684	276	21	is	be	AUX
easat-10684	276	22	stored	store	VERB
easat-10684	276	23	in	in	ADP
easat-10684	276	24	an	an	DET
easat-10684	276	25	array	array	NOUN
easat-10684	276	26	of	of	ADP
easat-10684	276	27	length	length	NOUN
easat-10684	276	28	300	300	NUM
easat-10684	276	29	.	.	PUNCT
easat-10684	277	1	with	with	ADP
easat-10684	277	2	a	a	DET
easat-10684	277	3	2	2	NUM
easat-10684	277	4	-	-	PUNCT
easat-10684	277	5	d	d	NOUN
easat-10684	277	6	feature	feature	NOUN
easat-10684	277	7	vector	vector	NOUN
easat-10684	277	8	matrix	matrix	NOUN
easat-10684	277	9	table	table	NOUN
easat-10684	277	10	format	format	NOUN
easat-10684	277	11	of	of	ADP
easat-10684	277	12	columns	column	NOUN
easat-10684	277	13	and	and	CCONJ
easat-10684	277	14	rows	row	NOUN
easat-10684	277	15	,	,	PUNCT
easat-10684	277	16	a	a	DET
easat-10684	277	17	matrix	matrix	NOUN
easat-10684	277	18	can	can	AUX
easat-10684	277	19	be	be	AUX
easat-10684	277	20	generated	generate	VERB
easat-10684	277	21	for	for	ADP
easat-10684	277	22	each	each	DET
easat-10684	277	23	news	news	NOUN
easat-10684	277	24	sample	sample	NOUN
easat-10684	277	25	sentence	sentence	NOUN
easat-10684	277	26	.	.	PUNCT
easat-10684	278	1	the	the	DET
easat-10684	278	2	columns	column	NOUN
easat-10684	278	3	represent	represent	VERB
easat-10684	278	4	the	the	DET
easat-10684	278	5	numeric	numeric	ADJ
easat-10684	278	6	features	feature	NOUN
easat-10684	278	7	,	,	PUNCT
easat-10684	278	8	and	and	CCONJ
easat-10684	278	9	the	the	DET
easat-10684	278	10	rows	row	NOUN
easat-10684	278	11	contain	contain	VERB
easat-10684	278	12	a	a	DET
easat-10684	278	13	300	300	NUM
easat-10684	278	14	-	-	PUNCT
easat-10684	278	15	dimensional	dimensional	ADJ
easat-10684	278	16	1526	1526	NUM
easat-10684	278	17	edelweiss	edelweiss	PROPN
easat-10684	278	18	applied	apply	VERB
easat-10684	278	19	science	science	NOUN
easat-10684	278	20	and	and	CCONJ
easat-10684	278	21	technology	technology	NOUN
easat-10684	278	22	issn	issn	PROPN
easat-10684	278	23	:	:	PUNCT
easat-10684	278	24	2576	2576	NUM
easat-10684	278	25	-	-	SYM
easat-10684	278	26	8484	8484	NUM
easat-10684	278	27	vol	vol	NOUN
easat-10684	278	28	.	.	PROPN
easat-10684	279	1	9	9	NUM
easat-10684	279	2	,	,	PUNCT
easat-10684	279	3	no	no	INTJ
easat-10684	279	4	.	.	NOUN
easat-10684	279	5	10	10	NUM
easat-10684	279	6	:	:	PUNCT
easat-10684	279	7	1515	1515	NUM
easat-10684	279	8	-	-	SYM
easat-10684	279	9	1540	1540	NUM
easat-10684	279	10	,	,	PUNCT
easat-10684	279	11	2025	2025	NUM
easat-10684	279	12	doi	doi	NOUN
easat-10684	279	13	:	:	PUNCT
easat-10684	279	14	10.55214/2576	10.55214/2576	NUM
easat-10684	279	15	-	-	SYM
easat-10684	279	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	280	1	©	©	PROPN
easat-10684	280	2	2025	2025	NUM
easat-10684	280	3	by	by	ADP
easat-10684	280	4	the	the	DET
easat-10684	280	5	authors	author	NOUN
easat-10684	280	6	;	;	PUNCT
easat-10684	280	7	licensee	licensee	PROPN
easat-10684	280	8	learning	learning	NOUN
easat-10684	280	9	gate	gate	NOUN
easat-10684	280	10	long	long	ADJ
easat-10684	280	11	feature	feature	NOUN
easat-10684	280	12	vector	vector	NOUN
easat-10684	280	13	representing	represent	VERB
easat-10684	280	14	that	that	DET
easat-10684	280	15	particular	particular	ADJ
easat-10684	280	16	sample	sample	NOUN
easat-10684	280	17	.	.	PUNCT
easat-10684	281	1	figure	figure	NOUN
easat-10684	281	2	14	14	NUM
easat-10684	281	3	shows	show	VERB
easat-10684	281	4	the	the	DET
easat-10684	281	5	feature	feature	NOUN
easat-10684	281	6	matrix	matrix	NOUN
easat-10684	281	7	of	of	ADP
easat-10684	281	8	300	300	NUM
easat-10684	281	9	features	feature	NOUN
easat-10684	281	10	by	by	ADP
easat-10684	281	11	4,594	4,594	NUM
easat-10684	281	12	samples	sample	NOUN
easat-10684	281	13	that	that	PRON
easat-10684	281	14	was	be	AUX
easat-10684	281	15	generated	generate	VERB
easat-10684	281	16	for	for	ADP
easat-10684	281	17	the	the	DET
easat-10684	281	18	fake	fake	ADJ
easat-10684	281	19	news	news	NOUN
easat-10684	281	20	dataset	dataset	NOUN
easat-10684	281	21	.	.	PUNCT
easat-10684	282	1	3.1.4.2	3.1.4.2	NUM
easat-10684	282	2	.	.	PUNCT
easat-10684	283	1	sequencing	sequence	VERB
easat-10684	283	2	&	&	CCONJ
easat-10684	283	3	embedding	embed	VERB
easat-10684	283	4	matrix	matrix	NOUN
easat-10684	283	5	sequencing	sequencing	NOUN
easat-10684	283	6	and	and	CCONJ
easat-10684	283	7	generation	generation	NOUN
easat-10684	283	8	of	of	ADP
easat-10684	283	9	the	the	DET
easat-10684	283	10	embedding	embed	VERB
easat-10684	283	11	matrix	matrix	NOUN
easat-10684	283	12	is	be	AUX
easat-10684	283	13	a	a	DET
easat-10684	283	14	technique	technique	NOUN
easat-10684	283	15	used	use	VERB
easat-10684	283	16	to	to	PART
easat-10684	283	17	create	create	VERB
easat-10684	283	18	a	a	DET
easat-10684	283	19	special	special	ADJ
easat-10684	283	20	feature	feature	NOUN
easat-10684	283	21	that	that	PRON
easat-10684	283	22	can	can	AUX
easat-10684	283	23	be	be	AUX
easat-10684	283	24	passed	pass	VERB
easat-10684	283	25	to	to	ADP
easat-10684	283	26	the	the	DET
easat-10684	283	27	embedding	embed	VERB
easat-10684	283	28	layer	layer	NOUN
easat-10684	283	29	of	of	ADP
easat-10684	283	30	the	the	DET
easat-10684	283	31	bi	bi	ADJ
easat-10684	283	32	-	-	ADJ
easat-10684	283	33	lstm	lstm	ADJ
easat-10684	283	34	architecture	architecture	NOUN
easat-10684	283	35	.	.	PUNCT
easat-10684	284	1	the	the	DET
easat-10684	284	2	tensorflow	tensorflow	NOUN
easat-10684	284	3	library	library	NOUN
easat-10684	284	4	will	will	AUX
easat-10684	284	5	be	be	AUX
easat-10684	284	6	used	use	VERB
easat-10684	284	7	to	to	PART
easat-10684	284	8	perform	perform	VERB
easat-10684	284	9	various	various	ADJ
easat-10684	284	10	preprocessing	preprocessing	NOUN
easat-10684	284	11	operations	operation	NOUN
easat-10684	284	12	and	and	CCONJ
easat-10684	284	13	extraction	extraction	NOUN
easat-10684	284	14	of	of	ADP
easat-10684	284	15	numeric	numeric	ADJ
easat-10684	284	16	sequences	sequence	NOUN
easat-10684	284	17	for	for	ADP
easat-10684	284	18	the	the	DET
easat-10684	284	19	embedding	embed	VERB
easat-10684	284	20	layer	layer	NOUN
easat-10684	284	21	.	.	PUNCT
easat-10684	285	1	the	the	DET
easat-10684	285	2	study	study	NOUN
easat-10684	285	3	uses	use	VERB
easat-10684	285	4	the	the	DET
easat-10684	285	5	tensorflow	tensorflow	NOUN
easat-10684	285	6	library	library	NOUN
easat-10684	285	7	for	for	ADP
easat-10684	285	8	preprocessing	preprocesse	VERB
easat-10684	285	9	and	and	CCONJ
easat-10684	285	10	converting	convert	VERB
easat-10684	285	11	fake	fake	ADJ
easat-10684	285	12	news	news	NOUN
easat-10684	285	13	text	text	NOUN
easat-10684	285	14	into	into	ADP
easat-10684	285	15	a	a	DET
easat-10684	285	16	numeric	numeric	ADJ
easat-10684	285	17	sequence	sequence	NOUN
easat-10684	285	18	.	.	PUNCT
easat-10684	286	1	this	this	DET
easat-10684	286	2	preprocessing	preprocessing	NOUN
easat-10684	286	3	and	and	CCONJ
easat-10684	286	4	transformation	transformation	NOUN
easat-10684	286	5	are	be	AUX
easat-10684	286	6	essential	essential	ADJ
easat-10684	286	7	to	to	PART
easat-10684	286	8	have	have	VERB
easat-10684	286	9	a	a	DET
easat-10684	286	10	suitable	suitable	ADJ
easat-10684	286	11	format	format	NOUN
easat-10684	286	12	for	for	ADP
easat-10684	286	13	proper	proper	ADJ
easat-10684	286	14	loading	loading	NOUN
easat-10684	286	15	into	into	ADP
easat-10684	286	16	the	the	DET
easat-10684	286	17	embedding	embed	VERB
easat-10684	286	18	layer	layer	NOUN
easat-10684	286	19	of	of	ADP
easat-10684	286	20	the	the	DET
easat-10684	286	21	bi	bi	ADJ
easat-10684	286	22	-	-	ADJ
easat-10684	286	23	lstm	lstm	ADJ
easat-10684	286	24	architecture	architecture	NOUN
easat-10684	286	25	.	.	PUNCT
easat-10684	287	1	a	a	DET
easat-10684	287	2	maximum	maximum	ADJ
easat-10684	287	3	vocabulary	vocabulary	NOUN
easat-10684	287	4	of	of	ADP
easat-10684	287	5	10,000	10,000	NUM
easat-10684	287	6	is	be	AUX
easat-10684	287	7	generated	generate	VERB
easat-10684	287	8	,	,	PUNCT
easat-10684	287	9	and	and	CCONJ
easat-10684	287	10	each	each	DET
easat-10684	287	11	row	row	NOUN
easat-10684	287	12	(	(	PUNCT
easat-10684	287	13	news	news	NOUN
easat-10684	287	14	samples	sample	NOUN
easat-10684	287	15	)	)	PUNCT
easat-10684	287	16	is	be	AUX
easat-10684	287	17	tokenized	tokenize	VERB
easat-10684	287	18	and	and	CCONJ
easat-10684	287	19	padded	pad	VERB
easat-10684	287	20	to	to	ADP
easat-10684	287	21	500	500	NUM
easat-10684	287	22	long	long	ADV
easat-10684	287	23	;	;	PUNCT
easat-10684	287	24	this	this	PRON
easat-10684	287	25	ensures	ensure	VERB
easat-10684	287	26	that	that	SCONJ
easat-10684	287	27	all	all	DET
easat-10684	287	28	fake	fake	ADJ
easat-10684	287	29	news	news	NOUN
easat-10684	287	30	data	data	NOUN
easat-10684	287	31	samples	sample	NOUN
easat-10684	287	32	are	be	AUX
easat-10684	287	33	in	in	ADP
easat-10684	287	34	an	an	DET
easat-10684	287	35	equal	equal	ADJ
easat-10684	287	36	sequence	sequence	NOUN
easat-10684	287	37	(	(	PUNCT
easat-10684	287	38	word	word	NOUN
easat-10684	287	39	length	length	NOUN
easat-10684	287	40	)	)	PUNCT
easat-10684	287	41	for	for	ADP
easat-10684	287	42	each	each	DET
easat-10684	287	43	text	text	NOUN
easat-10684	287	44	sample	sample	NOUN
easat-10684	287	45	.	.	PUNCT
easat-10684	288	1	finally	finally	ADV
easat-10684	288	2	,	,	PUNCT
easat-10684	288	3	the	the	DET
easat-10684	288	4	texts	text	NOUN
easat-10684	288	5	are	be	AUX
easat-10684	288	6	converted	convert	VERB
easat-10684	288	7	to	to	ADP
easat-10684	288	8	numeric	numeric	ADJ
easat-10684	288	9	representation	representation	NOUN
easat-10684	288	10	using	use	VERB
easat-10684	288	11	the	the	DET
easat-10684	288	12	text_to_sequence	text_to_sequence	PROPN
easat-10684	288	13	function	function	NOUN
easat-10684	288	14	of	of	ADP
easat-10684	288	15	the	the	DET
easat-10684	288	16	tensorflow	tensorflow	NOUN
easat-10684	288	17	library	library	NOUN
easat-10684	288	18	.	.	PUNCT
easat-10684	289	1	the	the	DET
easat-10684	289	2	sequence	sequence	NOUN
easat-10684	289	3	generated	generate	VERB
easat-10684	289	4	is	be	AUX
easat-10684	289	5	used	use	VERB
easat-10684	289	6	to	to	PART
easat-10684	289	7	create	create	VERB
easat-10684	289	8	the	the	DET
easat-10684	289	9	embedding	embed	VERB
easat-10684	289	10	matrix	matrix	NOUN
easat-10684	289	11	.	.	PUNCT
easat-10684	290	1	therefore	therefore	ADV
easat-10684	290	2	,	,	PUNCT
easat-10684	290	3	generating	generate	VERB
easat-10684	290	4	the	the	DET
easat-10684	290	5	embedding	embed	VERB
easat-10684	290	6	matrix	matrix	NOUN
easat-10684	290	7	also	also	ADV
easat-10684	290	8	requires	require	VERB
easat-10684	290	9	the	the	DET
easat-10684	290	10	word2vec	word2vec	ADJ
easat-10684	290	11	representation	representation	NOUN
easat-10684	290	12	and	and	CCONJ
easat-10684	290	13	utilizes	utilize	VERB
easat-10684	290	14	each	each	DET
easat-10684	290	15	word	word	NOUN
easat-10684	290	16	from	from	ADP
easat-10684	290	17	the	the	DET
easat-10684	290	18	generated	generate	VERB
easat-10684	290	19	vocabulary	vocabulary	NOUN
easat-10684	290	20	.	.	PUNCT
easat-10684	291	1	as	as	ADP
easat-10684	291	2	a	a	DET
easat-10684	291	3	result	result	NOUN
easat-10684	291	4	of	of	ADP
easat-10684	291	5	this	this	DET
easat-10684	291	6	process	process	NOUN
easat-10684	291	7	,	,	PUNCT
easat-10684	291	8	an	an	DET
easat-10684	291	9	80,859	80,859	NUM
easat-10684	291	10	by	by	ADP
easat-10684	291	11	500	500	NUM
easat-10684	291	12	(	(	PUNCT
easat-10684	291	13	2	2	NUM
easat-10684	291	14	-	-	PUNCT
easat-10684	291	15	dimensional	dimensional	ADJ
easat-10684	291	16	)	)	PUNCT
easat-10684	291	17	embedding	embed	VERB
easat-10684	291	18	matrix	matrix	NOUN
easat-10684	291	19	is	be	AUX
easat-10684	291	20	produced	produce	VERB
easat-10684	291	21	.	.	PUNCT
easat-10684	292	1	the	the	DET
easat-10684	292	2	embedding	embed	VERB
easat-10684	292	3	matrix	matrix	NOUN
easat-10684	292	4	can	can	AUX
easat-10684	292	5	now	now	ADV
easat-10684	292	6	be	be	AUX
easat-10684	292	7	passed	pass	VERB
easat-10684	292	8	as	as	ADP
easat-10684	292	9	an	an	DET
easat-10684	292	10	embedding	embed	VERB
easat-10684	292	11	parameter	parameter	NOUN
easat-10684	292	12	to	to	ADP
easat-10684	292	13	the	the	DET
easat-10684	292	14	bi	bi	ADJ
easat-10684	292	15	-	-	ADJ
easat-10684	292	16	lstm	lstm	ADJ
easat-10684	292	17	input	input	NOUN
easat-10684	292	18	layer	layer	NOUN
easat-10684	292	19	.	.	PUNCT
easat-10684	293	1	algorithm	algorithm	NOUN
easat-10684	293	2	1	1	NUM
easat-10684	293	3	illustrates	illustrate	VERB
easat-10684	293	4	the	the	DET
easat-10684	293	5	pre	pre	ADJ
easat-10684	293	6	-	-	ADJ
easat-10684	293	7	processing	processing	ADJ
easat-10684	293	8	steps	step	NOUN
easat-10684	293	9	.	.	PUNCT
easat-10684	294	1	3.1.5	3.1.5	X
easat-10684	294	2	.	.	PUNCT
easat-10684	294	3	data	datum	NOUN
easat-10684	294	4	splitting	split	VERB
easat-10684	294	5	data	datum	NOUN
easat-10684	294	6	splitting	splitting	NOUN
easat-10684	294	7	entails	entail	VERB
easat-10684	294	8	dividing	divide	VERB
easat-10684	294	9	the	the	DET
easat-10684	294	10	dataset	dataset	NOUN
easat-10684	294	11	into	into	ADP
easat-10684	294	12	training	training	NOUN
easat-10684	294	13	and	and	CCONJ
easat-10684	294	14	testing	testing	NOUN
easat-10684	294	15	samples	sample	NOUN
easat-10684	294	16	,	,	PUNCT
easat-10684	294	17	which	which	PRON
easat-10684	294	18	results	result	VERB
easat-10684	294	19	in	in	ADP
easat-10684	294	20	having	have	VERB
easat-10684	294	21	separate	separate	ADJ
easat-10684	294	22	data	datum	NOUN
easat-10684	294	23	for	for	ADP
easat-10684	294	24	training	train	VERB
easat-10684	294	25	the	the	DET
easat-10684	294	26	fake	fake	ADJ
easat-10684	294	27	news	news	NOUN
easat-10684	294	28	detection	detection	NOUN
easat-10684	294	29	model	model	NOUN
easat-10684	294	30	and	and	CCONJ
easat-10684	294	31	evaluating	evaluate	VERB
easat-10684	294	32	the	the	DET
easat-10684	294	33	model	model	NOUN
easat-10684	294	34	's	's	PART
easat-10684	294	35	performance	performance	NOUN
easat-10684	294	36	using	use	VERB
easat-10684	294	37	the	the	DET
easat-10684	294	38	remaining	remain	VERB
easat-10684	294	39	samples	sample	NOUN
easat-10684	294	40	(	(	PUNCT
easat-10684	294	41	testing	testing	NOUN
easat-10684	294	42	samples	sample	NOUN
easat-10684	294	43	)	)	PUNCT
easat-10684	294	44	.	.	PUNCT
easat-10684	295	1	70	70	NUM
easat-10684	295	2	%	%	NOUN
easat-10684	295	3	of	of	ADP
easat-10684	295	4	the	the	DET
easat-10684	295	5	fake	fake	ADJ
easat-10684	295	6	news	news	NOUN
easat-10684	295	7	dataset	dataset	NOUN
easat-10684	295	8	is	be	AUX
easat-10684	295	9	allocated	allocate	VERB
easat-10684	295	10	for	for	ADP
easat-10684	295	11	training	training	NOUN
easat-10684	295	12	,	,	PUNCT
easat-10684	295	13	and	and	CCONJ
easat-10684	295	14	the	the	DET
easat-10684	295	15	remaining	remain	VERB
easat-10684	295	16	30	30	NUM
easat-10684	295	17	%	%	NOUN
easat-10684	295	18	is	be	AUX
easat-10684	295	19	assigned	assign	VERB
easat-10684	295	20	to	to	ADP
easat-10684	295	21	model	model	NOUN
easat-10684	295	22	evaluation	evaluation	NOUN
easat-10684	295	23	.	.	PUNCT
easat-10684	296	1	this	this	DET
easat-10684	296	2	study	study	NOUN
easat-10684	296	3	uses	use	VERB
easat-10684	296	4	the	the	DET
easat-10684	296	5	model_selection.train_test_split	model_selection.train_test_split	ADJ
easat-10684	296	6	function	function	NOUN
easat-10684	296	7	from	from	ADP
easat-10684	296	8	the	the	DET
easat-10684	296	9	scikit	scikit	NOUN
easat-10684	296	10	-	-	PUNCT
easat-10684	296	11	learn	learn	NOUN
easat-10684	296	12	module	module	NOUN
easat-10684	296	13	to	to	PART
easat-10684	296	14	split	split	VERB
easat-10684	296	15	the	the	DET
easat-10684	296	16	fake	fake	ADJ
easat-10684	296	17	news	news	NOUN
easat-10684	296	18	dataset	dataset	NOUN
easat-10684	296	19	.	.	PUNCT
easat-10684	297	1	algorithm	algorithm	PROPN
easat-10684	297	2	1	1	NUM
easat-10684	297	3	:	:	PUNCT
easat-10684	297	4	pre	pre	ADJ
easat-10684	297	5	-	-	ADJ
easat-10684	297	6	processing	process	VERB
easat-10684	297	7	1	1	NUM
easat-10684	297	8	.	.	PUNCT
easat-10684	297	9	input	input	NOUN
easat-10684	297	10	:	:	PUNCT
easat-10684	298	1	the	the	DET
easat-10684	298	2	dataset	dataset	NOUN
easat-10684	298	3	2	2	X
easat-10684	298	4	.	.	X
easat-10684	298	5	output	output	NOUN
easat-10684	298	6	:	:	PUNCT
easat-10684	298	7	pre	pre	ADJ
easat-10684	298	8	-	-	ADJ
easat-10684	298	9	processing	processing	ADJ
easat-10684	298	10	to	to	PART
easat-10684	298	11	be	be	AUX
easat-10684	298	12	fed	feed	VERB
easat-10684	298	13	to	to	ADP
easat-10684	298	14	the	the	DET
easat-10684	298	15	model	model	NOUN
easat-10684	298	16	3	3	NUM
easat-10684	298	17	.	.	PUNCT
easat-10684	298	18	step	step	NOUN
easat-10684	298	19	1	1	NUM
easat-10684	298	20	:	:	PUNCT
easat-10684	298	21	text	text	NOUN
easat-10684	298	22	cleaning	clean	VERB
easat-10684	298	23	4	4	NUM
easat-10684	298	24	.	.	PUNCT
easat-10684	299	1	for	for	ADP
easat-10684	299	2	each	each	DET
easat-10684	299	3	data	datum	NOUN
easat-10684	299	4	point	point	NOUN
easat-10684	299	5	,	,	PUNCT
easat-10684	299	6	do	do	VERB
easat-10684	299	7	5	5	NUM
easat-10684	299	8	.	.	PUNCT
easat-10684	299	9	remove	remove	VERB
easat-10684	299	10	urls	url	NOUN
easat-10684	299	11	,	,	PUNCT
easat-10684	299	12	special	special	ADJ
easat-10684	299	13	characters	character	NOUN
easat-10684	299	14	,	,	PUNCT
easat-10684	299	15	numerical	numerical	ADJ
easat-10684	299	16	values	value	NOUN
easat-10684	299	17	,	,	PUNCT
easat-10684	299	18	emoticons	emoticon	NOUN
easat-10684	299	19	,	,	PUNCT
easat-10684	299	20	emojis	emoji	NOUN
easat-10684	299	21	,	,	PUNCT
easat-10684	299	22	mentions	mention	NOUN
easat-10684	299	23	,	,	PUNCT
easat-10684	299	24	and	and	CCONJ
easat-10684	299	25	hashtags	hashtag	NOUN
easat-10684	299	26	.	.	PUNCT
easat-10684	300	1	6	6	X
easat-10684	300	2	.	.	X
easat-10684	300	3	remove	remove	VERB
easat-10684	300	4	non	non	ADJ
easat-10684	300	5	-	-	ADJ
easat-10684	300	6	alphanumeric	alphanumeric	ADJ
easat-10684	300	7	characters	character	NOUN
easat-10684	300	8	7	7	NUM
easat-10684	300	9	.	.	PUNCT
easat-10684	300	10	remove	remove	VERB
easat-10684	300	11	extra	extra	ADJ
easat-10684	300	12	spaces	space	NOUN
easat-10684	300	13	and	and	CCONJ
easat-10684	300	14	trim	trim	VERB
easat-10684	300	15	the	the	DET
easat-10684	300	16	text	text	NOUN
easat-10684	300	17	.	.	PUNCT
easat-10684	301	1	8	8	X
easat-10684	301	2	.	.	X
easat-10684	301	3	convert	convert	VERB
easat-10684	301	4	text	text	NOUN
easat-10684	301	5	to	to	PART
easat-10684	301	6	lowercase	lowercase	VERB
easat-10684	301	7	.	.	PUNCT
easat-10684	302	1	9	9	X
easat-10684	302	2	.	.	X
easat-10684	302	3	end	end	NOUN
easat-10684	302	4	for	for	ADP
easat-10684	302	5	10	10	NUM
easat-10684	302	6	.	.	PUNCT
easat-10684	303	1	step	step	NOUN
easat-10684	303	2	2	2	NUM
easat-10684	303	3	:	:	PUNCT
easat-10684	303	4	tokenisation	tokenisation	NOUN
easat-10684	303	5	11	11	NUM
easat-10684	303	6	.	.	PUNCT
easat-10684	304	1	for	for	ADP
easat-10684	304	2	each	each	DET
easat-10684	304	3	cleaned	clean	VERB
easat-10684	304	4	data	datum	NOUN
easat-10684	304	5	set	set	VERB
easat-10684	304	6	,	,	PUNCT
easat-10684	304	7	do	do	VERB
easat-10684	304	8	12	12	NUM
easat-10684	304	9	.	.	PUNCT
easat-10684	305	1	split	split	VERB
easat-10684	305	2	the	the	DET
easat-10684	305	3	text	text	NOUN
easat-10684	305	4	into	into	ADP
easat-10684	305	5	individual	individual	ADJ
easat-10684	305	6	words	word	NOUN
easat-10684	305	7	(	(	PUNCT
easat-10684	305	8	tokens	token	NOUN
easat-10684	305	9	)	)	PUNCT
easat-10684	305	10	13	13	NUM
easat-10684	305	11	.	.	PUNCT
easat-10684	306	1	end	end	VERB
easat-10684	306	2	for	for	ADP
easat-10684	306	3	14	14	NUM
easat-10684	306	4	.	.	PUNCT
easat-10684	307	1	step	step	NOUN
easat-10684	307	2	3	3	NUM
easat-10684	307	3	:	:	PUNCT
easat-10684	307	4	stop	stop	VERB
easat-10684	307	5	word	word	NOUN
easat-10684	307	6	removal	removal	NOUN
easat-10684	307	7	15	15	NUM
easat-10684	307	8	.	.	PUNCT
easat-10684	308	1	for	for	ADP
easat-10684	308	2	each	each	DET
easat-10684	308	3	token	token	ADJ
easat-10684	308	4	w	w	PROPN
easat-10684	308	5	∈	∈	PROPN
easat-10684	308	6	data	datum	NOUN
easat-10684	308	7	do	do	VERB
easat-10684	308	8	16	16	NUM
easat-10684	308	9	.	.	PUNCT
easat-10684	309	1	if	if	SCONJ
easat-10684	309	2	w	w	NOUN
easat-10684	309	3	is	be	AUX
easat-10684	309	4	a	a	DET
easat-10684	309	5	stop	stop	NOUN
easat-10684	309	6	word	word	NOUN
easat-10684	309	7	,	,	PUNCT
easat-10684	309	8	then	then	ADV
easat-10684	309	9	17	17	NUM
easat-10684	309	10	.	.	PUNCT
easat-10684	310	1	remove	remove	VERB
easat-10684	310	2	w	w	NOUN
easat-10684	310	3	from	from	ADP
easat-10684	310	4	data	datum	NOUN
easat-10684	310	5	18	18	NUM
easat-10684	310	6	.	.	PUNCT
easat-10684	311	1	end	end	VERB
easat-10684	311	2	if	if	SCONJ
easat-10684	311	3	1527	1527	NUM
easat-10684	311	4	edelweiss	edelweiss	PROPN
easat-10684	311	5	applied	apply	VERB
easat-10684	311	6	science	science	NOUN
easat-10684	311	7	and	and	CCONJ
easat-10684	311	8	technology	technology	NOUN
easat-10684	311	9	issn	issn	PROPN
easat-10684	311	10	:	:	PUNCT
easat-10684	311	11	2576	2576	NUM
easat-10684	311	12	-	-	SYM
easat-10684	311	13	8484	8484	NUM
easat-10684	311	14	vol	vol	NOUN
easat-10684	311	15	.	.	PROPN
easat-10684	312	1	9	9	NUM
easat-10684	312	2	,	,	PUNCT
easat-10684	312	3	no	no	INTJ
easat-10684	312	4	.	.	NOUN
easat-10684	312	5	10	10	NUM
easat-10684	312	6	:	:	PUNCT
easat-10684	312	7	1515	1515	NUM
easat-10684	312	8	-	-	SYM
easat-10684	312	9	1540	1540	NUM
easat-10684	312	10	,	,	PUNCT
easat-10684	312	11	2025	2025	NUM
easat-10684	312	12	doi	doi	NOUN
easat-10684	312	13	:	:	PUNCT
easat-10684	312	14	10.55214/2576	10.55214/2576	NUM
easat-10684	312	15	-	-	SYM
easat-10684	312	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	313	1	©	©	PROPN
easat-10684	313	2	2025	2025	NUM
easat-10684	313	3	by	by	ADP
easat-10684	313	4	the	the	DET
easat-10684	313	5	authors	author	NOUN
easat-10684	313	6	;	;	PUNCT
easat-10684	313	7	licensee	licensee	PROPN
easat-10684	313	8	learning	learning	NOUN
easat-10684	313	9	gate	gate	PROPN
easat-10684	313	10	19	19	NUM
easat-10684	313	11	.	.	NOUN
easat-10684	313	12	end	end	VERB
easat-10684	313	13	for	for	ADP
easat-10684	313	14	20	20	NUM
easat-10684	313	15	.	.	PUNCT
easat-10684	314	1	step	step	NOUN
easat-10684	314	2	4	4	NUM
easat-10684	314	3	:	:	PUNCT
easat-10684	314	4	train	train	NOUN
easat-10684	314	5	-	-	PUNCT
easat-10684	314	6	test	test	NOUN
easat-10684	314	7	split	split	NOUN
easat-10684	314	8	21	21	NUM
easat-10684	314	9	.	.	PUNCT
easat-10684	315	1	split	split	VERB
easat-10684	315	2	the	the	DET
easat-10684	315	3	dataset	dataset	NOUN
easat-10684	315	4	into	into	ADP
easat-10684	315	5	a	a	DET
easat-10684	315	6	training	training	NOUN
easat-10684	315	7	set	set	NOUN
easat-10684	315	8	and	and	CCONJ
easat-10684	315	9	a	a	DET
easat-10684	315	10	testing	testing	NOUN
easat-10684	315	11	set	set	NOUN
easat-10684	315	12	.	.	PUNCT
easat-10684	316	1	22	22	NUM
easat-10684	316	2	.	.	PUNCT
easat-10684	317	1	step	step	NOUN
easat-10684	317	2	5	5	NUM
easat-10684	317	3	:	:	PUNCT
easat-10684	317	4	word2vec	word2vec	X
easat-10684	317	5	embedding	embed	VERB
easat-10684	317	6	23	23	NUM
easat-10684	317	7	.	.	PUNCT
easat-10684	318	1	train	train	VERB
easat-10684	318	2	a	a	DET
easat-10684	318	3	word2vec	word2vec	X
easat-10684	318	4	model	model	NOUN
easat-10684	318	5	on	on	ADP
easat-10684	318	6	a	a	DET
easat-10684	318	7	dataset	dataset	NOUN
easat-10684	318	8	.	.	PUNCT
easat-10684	319	1	24	24	NUM
easat-10684	319	2	.	.	X
easat-10684	319	3	vectorise	vectorise	VERB
easat-10684	319	4	the	the	DET
easat-10684	319	5	data	datum	NOUN
easat-10684	319	6	using	use	VERB
easat-10684	319	7	word2vec	word2vec	X
easat-10684	319	8	25	25	NUM
easat-10684	319	9	.	.	PUNCT
easat-10684	320	1	step	step	NOUN
easat-10684	320	2	6	6	NUM
easat-10684	320	3	:	:	PUNCT
easat-10684	320	4	fasttext	fasttext	PROPN
easat-10684	320	5	embedding	embed	VERB
easat-10684	320	6	26	26	NUM
easat-10684	320	7	.	.	PUNCT
easat-10684	321	1	train	train	VERB
easat-10684	321	2	the	the	DET
easat-10684	321	3	fasttext	fasttext	PROPN
easat-10684	321	4	model	model	NOUN
easat-10684	321	5	on	on	ADP
easat-10684	321	6	the	the	DET
easat-10684	321	7	data	datum	NOUN
easat-10684	321	8	.	.	PUNCT
easat-10684	322	1	27	27	NUM
easat-10684	322	2	.	.	PUNCT
easat-10684	323	1	vectorise	vectorise	NOUN
easat-10684	323	2	training	training	NOUN
easat-10684	323	3	and	and	CCONJ
easat-10684	323	4	test	test	NOUN
easat-10684	323	5	data	datum	NOUN
easat-10684	323	6	using	use	VERB
easat-10684	323	7	fasttext	fasttext	PROPN
easat-10684	323	8	28	28	NUM
easat-10684	323	9	.	.	PUNCT
easat-10684	324	1	step	step	NOUN
easat-10684	324	2	7	7	NUM
easat-10684	324	3	:	:	PUNCT
easat-10684	324	4	handle	handle	VERB
easat-10684	324	5	nans	nan	NOUN
easat-10684	324	6	in	in	ADP
easat-10684	324	7	embeddings	embedding	NOUN
easat-10684	324	8	29	29	NUM
easat-10684	324	9	.	.	PUNCT
easat-10684	325	1	for	for	ADP
easat-10684	325	2	each	each	DET
easat-10684	325	3	vector	vector	NOUN
easat-10684	325	4	v	v	PROPN
easat-10684	325	5	∈	∈	PROPN
easat-10684	325	6	ttrain	ttrain	NOUN
easat-10684	325	7	,	,	PUNCT
easat-10684	325	8	ttest	ttest	NOUN
easat-10684	325	9	do	do	VERB
easat-10684	325	10	30	30	NUM
easat-10684	325	11	.	.	PUNCT
easat-10684	326	1	if	if	SCONJ
easat-10684	326	2	any	any	DET
easat-10684	326	3	value	value	NOUN
easat-10684	326	4	in	in	ADP
easat-10684	326	5	v	v	NOUN
easat-10684	326	6	is	be	AUX
easat-10684	326	7	nan	nan	PROPN
easat-10684	326	8	,	,	PUNCT
easat-10684	326	9	then	then	ADV
easat-10684	326	10	31	31	NUM
easat-10684	326	11	.	.	PUNCT
easat-10684	327	1	replace	replace	VERB
easat-10684	327	2	v	v	NOUN
easat-10684	327	3	with	with	ADP
easat-10684	327	4	a	a	DET
easat-10684	327	5	zero	zero	NUM
easat-10684	327	6	vector	vector	NOUN
easat-10684	327	7	.	.	PUNCT
easat-10684	328	1	32	32	NUM
easat-10684	328	2	.	.	PUNCT
easat-10684	329	1	end	end	VERB
easat-10684	329	2	if	if	SCONJ
easat-10684	329	3	33	33	NUM
easat-10684	329	4	.	.	PUNCT
easat-10684	330	1	end	end	VERB
easat-10684	330	2	for	for	ADP
easat-10684	330	3	34	34	NUM
easat-10684	330	4	.	.	PUNCT
easat-10684	331	1	return	return	VERB
easat-10684	331	2	preprocessed	preprocesse	VERB
easat-10684	331	3	data	datum	NOUN
easat-10684	331	4	3.1.6	3.1.6	NUM
easat-10684	331	5	.	.	PUNCT
easat-10684	332	1	bi	bi	ADJ
easat-10684	332	2	-	-	ADJ
easat-10684	332	3	lstm	lstm	ADJ
easat-10684	332	4	training	train	VERB
easat-10684	332	5	the	the	DET
easat-10684	332	6	bidirectional	bidirectional	ADJ
easat-10684	332	7	long	long	ADJ
easat-10684	332	8	short	short	ADJ
easat-10684	332	9	-	-	PUNCT
easat-10684	332	10	term	term	NOUN
easat-10684	332	11	memory	memory	NOUN
easat-10684	332	12	(	(	PUNCT
easat-10684	332	13	bi	bi	NOUN
easat-10684	332	14	-	-	ADJ
easat-10684	332	15	lstm	lstm	ADJ
easat-10684	332	16	)	)	PUNCT
easat-10684	332	17	architecture	architecture	NOUN
easat-10684	332	18	is	be	AUX
easat-10684	332	19	a	a	DET
easat-10684	332	20	recurrent	recurrent	ADJ
easat-10684	332	21	neural	neural	ADJ
easat-10684	332	22	network	network	NOUN
easat-10684	332	23	that	that	PRON
easat-10684	332	24	processes	process	VERB
easat-10684	332	25	text	text	NOUN
easat-10684	332	26	input	input	NOUN
easat-10684	332	27	data	datum	NOUN
easat-10684	332	28	in	in	ADP
easat-10684	332	29	both	both	DET
easat-10684	332	30	directions	direction	NOUN
easat-10684	332	31	.	.	PUNCT
easat-10684	333	1	this	this	DET
easat-10684	333	2	architecture	architecture	NOUN
easat-10684	333	3	will	will	AUX
easat-10684	333	4	be	be	AUX
easat-10684	333	5	used	use	VERB
easat-10684	333	6	to	to	PART
easat-10684	333	7	pre	pre	VERB
easat-10684	333	8	-	-	VERB
easat-10684	333	9	train	train	VERB
easat-10684	333	10	the	the	DET
easat-10684	333	11	fake	fake	ADJ
easat-10684	333	12	news	news	NOUN
easat-10684	333	13	prediction	prediction	NOUN
easat-10684	333	14	model	model	NOUN
easat-10684	333	15	to	to	PART
easat-10684	333	16	extract	extract	VERB
easat-10684	333	17	highly	highly	ADV
easat-10684	333	18	contextualized	contextualized	ADJ
easat-10684	333	19	word	word	NOUN
easat-10684	333	20	embedding	embed	VERB
easat-10684	333	21	vectors	vector	NOUN
easat-10684	333	22	,	,	PUNCT
easat-10684	333	23	which	which	PRON
easat-10684	333	24	will	will	AUX
easat-10684	333	25	be	be	AUX
easat-10684	333	26	used	use	VERB
easat-10684	333	27	as	as	ADP
easat-10684	333	28	input	input	NOUN
easat-10684	333	29	vectors	vector	NOUN
easat-10684	333	30	for	for	ADP
easat-10684	333	31	various	various	ADJ
easat-10684	333	32	machine	machine	NOUN
easat-10684	333	33	learning	learn	VERB
easat-10684	333	34	algorithms	algorithm	NOUN
easat-10684	333	35	during	during	ADP
easat-10684	333	36	training	training	NOUN
easat-10684	333	37	in	in	ADP
easat-10684	333	38	a	a	DET
easat-10684	333	39	transfer	transfer	NOUN
easat-10684	333	40	learning	learning	NOUN
easat-10684	333	41	approach	approach	NOUN
easat-10684	333	42	.	.	PUNCT
easat-10684	334	1	the	the	DET
easat-10684	334	2	bi	bi	ADJ
easat-10684	334	3	-	-	ADJ
easat-10684	334	4	lstm	lstm	ADJ
easat-10684	334	5	architecture	architecture	NOUN
easat-10684	334	6	layers	layer	NOUN
easat-10684	334	7	are	be	AUX
easat-10684	334	8	created	create	VERB
easat-10684	334	9	using	use	VERB
easat-10684	334	10	tensorflow	tensorflow	NOUN
easat-10684	334	11	and	and	CCONJ
easat-10684	334	12	the	the	DET
easat-10684	334	13	keras	keras	PROPN
easat-10684	334	14	module	module	NOUN
easat-10684	334	15	.	.	PUNCT
easat-10684	335	1	a	a	DET
easat-10684	335	2	sequential	sequential	ADJ
easat-10684	335	3	bilstm	bilstm	NOUN
easat-10684	335	4	architecture	architecture	NOUN
easat-10684	335	5	is	be	AUX
easat-10684	335	6	created	create	VERB
easat-10684	335	7	with	with	ADP
easat-10684	335	8	seven	seven	NUM
easat-10684	335	9	layers	layer	NOUN
easat-10684	335	10	to	to	PART
easat-10684	335	11	train	train	VERB
easat-10684	335	12	the	the	DET
easat-10684	335	13	fake	fake	ADJ
easat-10684	335	14	news	news	NOUN
easat-10684	335	15	detection	detection	NOUN
easat-10684	335	16	models	model	NOUN
easat-10684	335	17	.	.	PUNCT
easat-10684	336	1	the	the	DET
easat-10684	336	2	architecture	architecture	NOUN
easat-10684	336	3	consists	consist	VERB
easat-10684	336	4	of	of	ADP
easat-10684	336	5	one	one	NUM
easat-10684	336	6	input	input	NOUN
easat-10684	336	7	layer	layer	NOUN
easat-10684	336	8	,	,	PUNCT
easat-10684	336	9	one	one	NUM
easat-10684	336	10	embedding	embed	VERB
easat-10684	336	11	layer	layer	NOUN
easat-10684	336	12	,	,	PUNCT
easat-10684	336	13	two	two	NUM
easat-10684	336	14	bidirectional	bidirectional	ADJ
easat-10684	336	15	lstm	lstm	ADJ
easat-10684	336	16	layers	layer	NOUN
easat-10684	336	17	,	,	PUNCT
easat-10684	336	18	one	one	NUM
easat-10684	336	19	lstm	lstm	ADJ
easat-10684	336	20	layer	layer	NOUN
easat-10684	336	21	,	,	PUNCT
easat-10684	336	22	one	one	NUM
easat-10684	336	23	fully	fully	ADV
easat-10684	336	24	connected	connected	ADJ
easat-10684	336	25	dense	dense	ADJ
easat-10684	336	26	layer	layer	NOUN
easat-10684	336	27	,	,	PUNCT
easat-10684	336	28	and	and	CCONJ
easat-10684	336	29	finally	finally	ADV
easat-10684	336	30	one	one	NUM
easat-10684	336	31	output	output	NOUN
easat-10684	336	32	layer	layer	NOUN
easat-10684	336	33	.	.	PUNCT
easat-10684	337	1	embedding	embed	VERB
easat-10684	337	2	layer	layer	NOUN
easat-10684	337	3	processes	process	NOUN
easat-10684	337	4	500	500	NUM
easat-10684	337	5	vectors	vector	NOUN
easat-10684	337	6	of	of	ADP
easat-10684	337	7	numeric	numeric	ADJ
easat-10684	337	8	values	value	NOUN
easat-10684	337	9	as	as	ADP
easat-10684	337	10	input	input	NOUN
easat-10684	337	11	,	,	PUNCT
easat-10684	337	12	while	while	SCONJ
easat-10684	337	13	the	the	DET
easat-10684	337	14	bi	bi	NOUN
easat-10684	337	15	-	-	ADJ
easat-10684	337	16	lstm	lstm	ADJ
easat-10684	337	17	,	,	PUNCT
easat-10684	337	18	lstm	lstm	ADJ
easat-10684	337	19	,	,	PUNCT
easat-10684	337	20	and	and	CCONJ
easat-10684	337	21	dense	dense	ADJ
easat-10684	337	22	layers	layer	NOUN
easat-10684	337	23	consist	consist	VERB
easat-10684	337	24	of	of	ADP
easat-10684	337	25	256	256	NUM
easat-10684	337	26	neurons	neuron	NOUN
easat-10684	337	27	,	,	PUNCT
easat-10684	337	28	64	64	NUM
easat-10684	337	29	neurons	neuron	NOUN
easat-10684	337	30	,	,	PUNCT
easat-10684	337	31	and	and	CCONJ
easat-10684	337	32	32	32	NUM
easat-10684	337	33	neurons	neuron	NOUN
easat-10684	337	34	for	for	ADP
easat-10684	337	35	weight	weight	NOUN
easat-10684	337	36	,	,	PUNCT
easat-10684	337	37	bias	bias	NOUN
easat-10684	337	38	,	,	PUNCT
easat-10684	337	39	and	and	CCONJ
easat-10684	337	40	loss	loss	NOUN
easat-10684	337	41	computation	computation	NOUN
easat-10684	337	42	.	.	PUNCT
easat-10684	338	1	finally	finally	ADV
easat-10684	338	2	,	,	PUNCT
easat-10684	338	3	the	the	DET
easat-10684	338	4	output	output	NOUN
easat-10684	338	5	layer	layer	NOUN
easat-10684	338	6	has	have	VERB
easat-10684	338	7	only	only	ADV
easat-10684	338	8	one	one	NUM
easat-10684	338	9	neuron	neuron	NOUN
easat-10684	338	10	denoting	denote	VERB
easat-10684	338	11	either	either	CCONJ
easat-10684	338	12	fake	fake	ADJ
easat-10684	338	13	(	(	PUNCT
easat-10684	338	14	0	0	NUM
easat-10684	338	15	)	)	PUNCT
easat-10684	338	16	or	or	CCONJ
easat-10684	338	17	real	real	ADJ
easat-10684	338	18	(	(	PUNCT
easat-10684	338	19	1	1	X
easat-10684	338	20	)	)	PUNCT
easat-10684	338	21	news	news	NOUN
easat-10684	338	22	output	output	NOUN
easat-10684	338	23	.	.	PUNCT
easat-10684	339	1	3.1.7	3.1.7	X
easat-10684	339	2	.	.	PUNCT
easat-10684	340	1	loading	load	VERB
easat-10684	340	2	pre	pre	ADJ
easat-10684	340	3	-	-	ADJ
easat-10684	340	4	trained	train	VERB
easat-10684	340	5	bi	bi	ADJ
easat-10684	340	6	-	-	ADJ
easat-10684	340	7	lstm	lstm	ADJ
easat-10684	340	8	weight	weight	NOUN
easat-10684	340	9	the	the	DET
easat-10684	340	10	highly	highly	ADV
easat-10684	340	11	contextualized	contextualized	ADJ
easat-10684	340	12	weights	weight	NOUN
easat-10684	340	13	or	or	CCONJ
easat-10684	340	14	features	feature	NOUN
easat-10684	340	15	from	from	ADP
easat-10684	340	16	the	the	DET
easat-10684	340	17	bi	bi	ADJ
easat-10684	340	18	-	-	ADJ
easat-10684	340	19	lstm	lstm	ADJ
easat-10684	340	20	model	model	NOUN
easat-10684	340	21	are	be	AUX
easat-10684	340	22	saved	save	VERB
easat-10684	340	23	and	and	CCONJ
easat-10684	340	24	reloaded	reload	VERB
easat-10684	340	25	(	(	PUNCT
easat-10684	340	26	reused	reuse	VERB
easat-10684	340	27	)	)	PUNCT
easat-10684	340	28	as	as	ADP
easat-10684	340	29	pre	pre	ADJ
easat-10684	340	30	-	-	ADJ
easat-10684	340	31	trained	train	VERB
easat-10684	340	32	weights	weight	NOUN
easat-10684	340	33	.	.	PUNCT
easat-10684	341	1	the	the	DET
easat-10684	341	2	weights	weight	NOUN
easat-10684	341	3	are	be	AUX
easat-10684	341	4	used	use	VERB
easat-10684	341	5	as	as	ADP
easat-10684	341	6	feature	feature	NOUN
easat-10684	341	7	vectors	vector	NOUN
easat-10684	341	8	to	to	PART
easat-10684	341	9	train	train	VERB
easat-10684	341	10	various	various	ADJ
easat-10684	341	11	machine	machine	NOUN
easat-10684	341	12	learning	learn	VERB
easat-10684	341	13	algorithms	algorithm	NOUN
easat-10684	341	14	.	.	PUNCT
easat-10684	342	1	3.1.8	3.1.8	X
easat-10684	342	2	.	.	PUNCT
easat-10684	342	3	machine	machine	NOUN
easat-10684	342	4	learning	learning	NOUN
easat-10684	342	5	algorithms	algorithm	VERB
easat-10684	342	6	five	five	NUM
easat-10684	342	7	machine	machine	NOUN
easat-10684	342	8	learning	learn	VERB
easat-10684	342	9	algorithms	algorithm	NOUN
easat-10684	342	10	,	,	PUNCT
easat-10684	342	11	including	include	VERB
easat-10684	342	12	support	support	NOUN
easat-10684	342	13	vector	vector	NOUN
easat-10684	342	14	machine	machine	NOUN
easat-10684	342	15	(	(	PUNCT
easat-10684	342	16	svm	svm	PROPN
easat-10684	342	17	)	)	PUNCT
easat-10684	342	18	,	,	PUNCT
easat-10684	342	19	k	k	X
easat-10684	342	20	-	-	PUNCT
easat-10684	342	21	nearest	near	ADJ
easat-10684	342	22	neighbour	neighbour	NOUN
easat-10684	342	23	(	(	PUNCT
easat-10684	342	24	knn	knn	PROPN
easat-10684	342	25	)	)	PUNCT
easat-10684	342	26	,	,	PUNCT
easat-10684	342	27	random	random	ADJ
easat-10684	342	28	forest	forest	NOUN
easat-10684	342	29	(	(	PUNCT
easat-10684	342	30	rf	rf	NOUN
easat-10684	342	31	)	)	PUNCT
easat-10684	342	32	,	,	PUNCT
easat-10684	342	33	naïve	naïve	ADJ
easat-10684	342	34	bayes	bayes	PROPN
easat-10684	342	35	(	(	PUNCT
easat-10684	342	36	nb	nb	NOUN
easat-10684	342	37	)	)	PUNCT
easat-10684	342	38	,	,	PUNCT
easat-10684	342	39	and	and	CCONJ
easat-10684	342	40	decision	decision	NOUN
easat-10684	342	41	tree	tree	NOUN
easat-10684	342	42	(	(	PUNCT
easat-10684	342	43	dt	dt	PROPN
easat-10684	342	44	)	)	PUNCT
easat-10684	342	45	,	,	PUNCT
easat-10684	342	46	will	will	AUX
easat-10684	342	47	be	be	AUX
easat-10684	342	48	trained	train	VERB
easat-10684	342	49	to	to	PART
easat-10684	342	50	detect	detect	VERB
easat-10684	342	51	fake	fake	ADJ
easat-10684	342	52	news	news	NOUN
easat-10684	342	53	.	.	PUNCT
easat-10684	343	1	the	the	DET
easat-10684	343	2	pre	pre	ADJ
easat-10684	343	3	-	-	ADJ
easat-10684	343	4	trained	train	VERB
easat-10684	343	5	fake	fake	ADJ
easat-10684	343	6	news	news	NOUN
easat-10684	343	7	weight	weight	NOUN
easat-10684	343	8	(	(	PUNCT
easat-10684	343	9	word	word	NOUN
easat-10684	343	10	embeddings	embedding	NOUN
easat-10684	343	11	from	from	ADP
easat-10684	343	12	bi	bi	NOUN
easat-10684	343	13	-	-	ADJ
easat-10684	343	14	lstm	lstm	ADJ
easat-10684	343	15	)	)	PUNCT
easat-10684	343	16	will	will	AUX
easat-10684	343	17	be	be	AUX
easat-10684	343	18	used	use	VERB
easat-10684	343	19	to	to	PART
easat-10684	343	20	enhance	enhance	VERB
easat-10684	343	21	the	the	DET
easat-10684	343	22	training	training	NOUN
easat-10684	343	23	process	process	NOUN
easat-10684	343	24	of	of	ADP
easat-10684	343	25	the	the	DET
easat-10684	343	26	listed	list	VERB
easat-10684	343	27	machine	machine	NOUN
easat-10684	343	28	learning	learning	NOUN
easat-10684	343	29	algorithms	algorithm	NOUN
easat-10684	343	30	.	.	PUNCT
easat-10684	344	1	3.1.8.1	3.1.8.1	X
easat-10684	344	2	.	.	PUNCT
easat-10684	344	3	naïve	naïve	ADJ
easat-10684	344	4	bayes	bayes	PROPN
easat-10684	344	5	(	(	PUNCT
easat-10684	344	6	nb	nb	NOUN
easat-10684	344	7	)	)	PUNCT
easat-10684	344	8	algorithm	algorithm	PROPN
easat-10684	344	9	naïve	naïve	ADJ
easat-10684	344	10	bayes	bayes	PROPN
easat-10684	344	11	is	be	AUX
easat-10684	344	12	widely	widely	ADV
easat-10684	344	13	used	use	VERB
easat-10684	344	14	to	to	PART
easat-10684	344	15	compute	compute	VERB
easat-10684	344	16	conditional	conditional	ADJ
easat-10684	344	17	probability	probability	NOUN
easat-10684	344	18	,	,	PUNCT
easat-10684	344	19	which	which	PRON
easat-10684	344	20	originates	originate	VERB
easat-10684	344	21	from	from	ADP
easat-10684	344	22	bayes	bayes	PROPN
easat-10684	344	23	'	'	PART
easat-10684	344	24	theorem	theorem	NOUN
easat-10684	344	25	by	by	ADP
easat-10684	344	26	expressing	express	VERB
easat-10684	344	27	the	the	DET
easat-10684	344	28	likelihood	likelihood	NOUN
easat-10684	344	29	of	of	ADP
easat-10684	344	30	an	an	DET
easat-10684	344	31	event	event	NOUN
easat-10684	344	32	occurring	occur	VERB
easat-10684	344	33	based	base	VERB
easat-10684	344	34	on	on	ADP
easat-10684	344	35	the	the	DET
easat-10684	344	36	occurrence	occurrence	NOUN
easat-10684	344	37	of	of	ADP
easat-10684	344	38	another	another	DET
easat-10684	344	39	event	event	NOUN
easat-10684	344	40	[	[	X
easat-10684	344	41	44	44	NUM
easat-10684	344	42	]	]	PUNCT
easat-10684	344	43	.	.	PUNCT
easat-10684	345	1	naïve	naïve	ADJ
easat-10684	345	2	bayes	bayes	PROPN
easat-10684	345	3	functions	function	NOUN
easat-10684	345	4	as	as	ADP
easat-10684	345	5	a	a	DET
easat-10684	345	6	classifier	classifier	NOUN
easat-10684	345	7	within	within	ADP
easat-10684	345	8	the	the	DET
easat-10684	345	9	domain	domain	NOUN
easat-10684	345	10	of	of	ADP
easat-10684	345	11	supervised	supervised	ADJ
easat-10684	345	12	learning	learning	NOUN
easat-10684	345	13	algorithms	algorithm	NOUN
easat-10684	345	14	,	,	PUNCT
easat-10684	345	15	and	and	CCONJ
easat-10684	345	16	it	it	PRON
easat-10684	345	17	performs	perform	VERB
easat-10684	345	18	predictions	prediction	NOUN
easat-10684	345	19	based	base	VERB
easat-10684	345	20	on	on	ADP
easat-10684	345	21	probability	probability	NOUN
easat-10684	345	22	scores	score	NOUN
easat-10684	345	23	from	from	ADP
easat-10684	345	24	different	different	ADJ
easat-10684	345	25	classes	class	NOUN
easat-10684	346	1	[	[	X
easat-10684	346	2	45	45	NUM
easat-10684	346	3	]	]	PUNCT
easat-10684	346	4	.	.	PUNCT
easat-10684	347	1	1528	1528	NUM
easat-10684	347	2	edelweiss	edelweiss	PROPN
easat-10684	347	3	applied	apply	VERB
easat-10684	347	4	science	science	NOUN
easat-10684	347	5	and	and	CCONJ
easat-10684	347	6	technology	technology	NOUN
easat-10684	347	7	issn	issn	PROPN
easat-10684	347	8	:	:	PUNCT
easat-10684	347	9	2576	2576	NUM
easat-10684	347	10	-	-	SYM
easat-10684	347	11	8484	8484	NUM
easat-10684	347	12	vol	vol	NOUN
easat-10684	347	13	.	.	PROPN
easat-10684	348	1	9	9	NUM
easat-10684	348	2	,	,	PUNCT
easat-10684	348	3	no	no	INTJ
easat-10684	348	4	.	.	NOUN
easat-10684	348	5	10	10	NUM
easat-10684	348	6	:	:	PUNCT
easat-10684	348	7	1515	1515	NUM
easat-10684	348	8	-	-	SYM
easat-10684	348	9	1540	1540	NUM
easat-10684	348	10	,	,	PUNCT
easat-10684	348	11	2025	2025	NUM
easat-10684	348	12	doi	doi	NOUN
easat-10684	348	13	:	:	PUNCT
easat-10684	348	14	10.55214/2576	10.55214/2576	NUM
easat-10684	348	15	-	-	SYM
easat-10684	348	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	349	1	©	©	PROPN
easat-10684	349	2	2025	2025	NUM
easat-10684	349	3	by	by	ADP
easat-10684	349	4	the	the	DET
easat-10684	349	5	authors	author	NOUN
easat-10684	349	6	;	;	PUNCT
easat-10684	349	7	licensee	licensee	PROPN
easat-10684	349	8	learning	learning	NOUN
easat-10684	349	9	gate	gate	NOUN
easat-10684	349	10	𝑃(𝐴|𝐵	𝑃(𝐴|𝐵	PROPN
easat-10684	349	11	)	)	PUNCT
easat-10684	349	12	=	=	PUNCT
easat-10684	350	1	𝑃(𝐵|𝐴)⋅𝑃(𝐴	𝑃(𝐵|𝐴)⋅𝑃(𝐴	NUM
easat-10684	350	2	)	)	PUNCT
easat-10684	350	3	𝑃(𝐵	𝑃(𝐵	NOUN
easat-10684	350	4	)	)	PUNCT
easat-10684	350	5	(	(	PUNCT
easat-10684	350	6	1	1	X
easat-10684	350	7	)	)	PUNCT
easat-10684	350	8	𝑃(𝐴|𝐵	𝑃(𝐴|𝐵	ADJ
easat-10684	350	9	):	):	PUNCT
easat-10684	350	10	probability	probability	NOUN
easat-10684	350	11	of	of	ADP
easat-10684	350	12	event	event	NOUN
easat-10684	350	13	a	a	DET
easat-10684	350	14	such	such	ADJ
easat-10684	350	15	that	that	DET
easat-10684	350	16	event	event	NOUN
easat-10684	350	17	b	b	NOUN
easat-10684	350	18	has	have	AUX
easat-10684	350	19	already	already	ADV
easat-10684	350	20	occurred	occur	VERB
easat-10684	350	21	.	.	PUNCT
easat-10684	351	1	3.1.8.2	3.1.8.2	X
easat-10684	351	2	.	.	X
easat-10684	351	3	support	support	NOUN
easat-10684	351	4	vector	vector	NOUN
easat-10684	351	5	machine	machine	NOUN
easat-10684	351	6	(	(	PUNCT
easat-10684	351	7	svm	svm	ADJ
easat-10684	351	8	)	)	PUNCT
easat-10684	351	9	svm	svm	ADJ
easat-10684	351	10	algorithm	algorithm	NOUN
easat-10684	351	11	is	be	AUX
easat-10684	351	12	highly	highly	ADV
easat-10684	351	13	efficient	efficient	ADJ
easat-10684	351	14	for	for	ADP
easat-10684	351	15	addressing	address	VERB
easat-10684	351	16	binary	binary	ADJ
easat-10684	351	17	classification	classification	NOUN
easat-10684	351	18	problems	problem	NOUN
easat-10684	351	19	[	[	X
easat-10684	351	20	3	3	NUM
easat-10684	351	21	]	]	PUNCT
easat-10684	351	22	.	.	PUNCT
easat-10684	352	1	svm	svm	PROPN
easat-10684	352	2	is	be	AUX
easat-10684	352	3	a	a	DET
easat-10684	352	4	supervised	supervised	ADJ
easat-10684	352	5	machine	machine	NOUN
easat-10684	352	6	learning	learn	VERB
easat-10684	352	7	algorithm	algorithm	NOUN
easat-10684	352	8	that	that	PRON
easat-10684	352	9	can	can	AUX
easat-10684	352	10	be	be	AUX
easat-10684	352	11	used	use	VERB
easat-10684	352	12	for	for	ADP
easat-10684	352	13	solving	solve	VERB
easat-10684	352	14	both	both	CCONJ
easat-10684	352	15	classification	classification	NOUN
easat-10684	352	16	and	and	CCONJ
easat-10684	352	17	regression	regression	NOUN
easat-10684	352	18	problems	problem	NOUN
easat-10684	352	19	;	;	PUNCT
easat-10684	352	20	it	it	PRON
easat-10684	352	21	is	be	AUX
easat-10684	352	22	based	base	VERB
easat-10684	352	23	on	on	ADP
easat-10684	352	24	the	the	DET
easat-10684	352	25	idea	idea	NOUN
easat-10684	352	26	of	of	ADP
easat-10684	352	27	finding	find	VERB
easat-10684	352	28	a	a	DET
easat-10684	352	29	hyperplane	hyperplane	NOUN
easat-10684	352	30	that	that	PRON
easat-10684	352	31	best	well	ADV
easat-10684	352	32	separates	separate	VERB
easat-10684	352	33	the	the	DET
easat-10684	352	34	dataset	dataset	NOUN
easat-10684	352	35	into	into	ADP
easat-10684	352	36	two	two	NUM
easat-10684	352	37	classes	class	NOUN
easat-10684	352	38	.	.	PUNCT
easat-10684	353	1	hyperplanes	hyperplane	NOUN
easat-10684	353	2	are	be	AUX
easat-10684	353	3	decision	decision	NOUN
easat-10684	353	4	boundaries	boundary	NOUN
easat-10684	353	5	that	that	PRON
easat-10684	353	6	help	help	VERB
easat-10684	353	7	machine	machine	NOUN
easat-10684	353	8	learning	learn	VERB
easat-10684	353	9	algorithms	algorithm	NOUN
easat-10684	353	10	classify	classify	VERB
easat-10684	353	11	data	datum	NOUN
easat-10684	353	12	points	point	NOUN
easat-10684	353	13	accurately	accurately	ADV
easat-10684	353	14	[	[	X
easat-10684	353	15	3	3	NUM
easat-10684	353	16	]	]	PUNCT
easat-10684	353	17	.	.	PUNCT
easat-10684	354	1	the	the	DET
easat-10684	354	2	algorithm	algorithm	NOUN
easat-10684	354	3	provided	provide	VERB
easat-10684	354	4	in	in	ADP
easat-10684	354	5	figure	figure	NOUN
easat-10684	354	6	x	x	PUNCT
easat-10684	354	7	is	be	AUX
easat-10684	354	8	based	base	VERB
easat-10684	354	9	on	on	ADP
easat-10684	354	10	feature	feature	NOUN
easat-10684	354	11	set	set	VERB
easat-10684	354	12	n	n	CCONJ
easat-10684	354	13	(	(	PUNCT
easat-10684	354	14	0	0	NUM
easat-10684	354	15	-	-	PUNCT
easat-10684	354	16	n	n	CCONJ
easat-10684	354	17	)	)	PUNCT
easat-10684	354	18	and	and	CCONJ
easat-10684	354	19	is	be	AUX
easat-10684	354	20	sorted	sort	VERB
easat-10684	354	21	using	use	VERB
easat-10684	354	22	information	information	NOUN
easat-10684	354	23	gained	gain	VERB
easat-10684	354	24	in	in	ADP
easat-10684	354	25	decreasing	decrease	VERB
easat-10684	354	26	order	order	NOUN
easat-10684	354	27	of	of	ADP
easat-10684	354	28	accuracy	accuracy	NOUN
easat-10684	354	29	.	.	PUNCT
easat-10684	355	1	3.1.8.3	3.1.8.3	NUM
easat-10684	355	2	.	.	PUNCT
easat-10684	356	1	k	k	X
easat-10684	356	2	-	-	PUNCT
easat-10684	356	3	nearest	near	ADJ
easat-10684	356	4	neighbours	neighbour	NOUN
easat-10684	356	5	(	(	PUNCT
easat-10684	356	6	knn	knn	PROPN
easat-10684	356	7	)	)	PUNCT
easat-10684	356	8	the	the	DET
easat-10684	356	9	k	k	PROPN
easat-10684	356	10	-	-	PUNCT
easat-10684	356	11	nn	nn	ADJ
easat-10684	356	12	algorithm	algorithm	NOUN
easat-10684	356	13	operates	operate	VERB
easat-10684	356	14	on	on	ADP
easat-10684	356	15	the	the	DET
easat-10684	356	16	concept	concept	NOUN
easat-10684	356	17	of	of	ADP
easat-10684	356	18	similarity	similarity	NOUN
easat-10684	356	19	measures	measure	NOUN
easat-10684	356	20	between	between	ADP
easat-10684	356	21	new	new	ADJ
easat-10684	356	22	samples	sample	NOUN
easat-10684	356	23	and	and	CCONJ
easat-10684	356	24	existing	exist	VERB
easat-10684	356	25	samples	sample	NOUN
easat-10684	356	26	.	.	PUNCT
easat-10684	357	1	the	the	DET
easat-10684	357	2	algorithm	algorithm	NOUN
easat-10684	357	3	groups	group	NOUN
easat-10684	357	4	the	the	DET
easat-10684	357	5	new	new	ADJ
easat-10684	357	6	sample	sample	NOUN
easat-10684	357	7	point	point	NOUN
easat-10684	357	8	into	into	ADP
easat-10684	357	9	the	the	DET
easat-10684	357	10	category	category	NOUN
easat-10684	357	11	that	that	PRON
easat-10684	357	12	is	be	AUX
easat-10684	357	13	most	most	ADV
easat-10684	357	14	likely	likely	ADJ
easat-10684	357	15	to	to	PART
easat-10684	357	16	be	be	AUX
easat-10684	357	17	one	one	NUM
easat-10684	357	18	of	of	ADP
easat-10684	357	19	the	the	DET
easat-10684	357	20	available	available	ADJ
easat-10684	357	21	categories	category	NOUN
easat-10684	357	22	.	.	PUNCT
easat-10684	358	1	it	it	PRON
easat-10684	358	2	stores	store	VERB
easat-10684	358	3	all	all	DET
easat-10684	358	4	accessible	accessible	ADJ
easat-10684	358	5	data	datum	NOUN
easat-10684	358	6	and	and	CCONJ
easat-10684	358	7	classifies	classify	VERB
easat-10684	358	8	new	new	ADJ
easat-10684	358	9	data	datum	NOUN
easat-10684	358	10	points	point	NOUN
easat-10684	358	11	using	use	VERB
easat-10684	358	12	similarity	similarity	NOUN
easat-10684	358	13	measures	measure	NOUN
easat-10684	358	14	.	.	PUNCT
easat-10684	359	1	3.1.8.4	3.1.8.4	X
easat-10684	359	2	.	.	X
easat-10684	359	3	decision	decision	NOUN
easat-10684	359	4	tree	tree	NOUN
easat-10684	359	5	algorithm	algorithm	NOUN
easat-10684	359	6	khanam	khanam	PROPN
easat-10684	359	7	et	et	PROPN
easat-10684	359	8	al	al	PROPN
easat-10684	359	9	.	.	PUNCT
easat-10684	360	1	[	[	X
easat-10684	360	2	46	46	NUM
easat-10684	360	3	]	]	PUNCT
easat-10684	360	4	introduce	introduce	VERB
easat-10684	360	5	a	a	DET
easat-10684	360	6	decision	decision	NOUN
easat-10684	360	7	tree	tree	NOUN
easat-10684	360	8	as	as	ADP
easat-10684	360	9	a	a	DET
easat-10684	360	10	vital	vital	ADJ
easat-10684	360	11	technique	technique	NOUN
easat-10684	360	12	used	use	VERB
easat-10684	360	13	for	for	ADP
easat-10684	360	14	solving	solve	VERB
easat-10684	360	15	classification	classification	NOUN
easat-10684	360	16	problems	problem	NOUN
easat-10684	360	17	,	,	PUNCT
easat-10684	360	18	which	which	PRON
easat-10684	360	19	operates	operate	VERB
easat-10684	360	20	like	like	ADP
easat-10684	360	21	a	a	DET
easat-10684	360	22	flow	flow	NOUN
easat-10684	360	23	chart	chart	NOUN
easat-10684	360	24	.	.	PUNCT
easat-10684	361	1	each	each	DET
easat-10684	361	2	internal	internal	ADJ
easat-10684	361	3	node	node	NOUN
easat-10684	361	4	of	of	ADP
easat-10684	361	5	a	a	DET
easat-10684	361	6	decision	decision	NOUN
easat-10684	361	7	tree	tree	NOUN
easat-10684	361	8	sets	set	VERB
easat-10684	361	9	a	a	DET
easat-10684	361	10	condition	condition	NOUN
easat-10684	361	11	or	or	CCONJ
easat-10684	361	12	"	"	PUNCT
easat-10684	361	13	test	test	NOUN
easat-10684	361	14	"	"	PUNCT
easat-10684	361	15	on	on	ADP
easat-10684	361	16	an	an	DET
easat-10684	361	17	attribute	attribute	NOUN
easat-10684	361	18	,	,	PUNCT
easat-10684	361	19	guiding	guide	VERB
easat-10684	361	20	the	the	DET
easat-10684	361	21	branching	branching	NOUN
easat-10684	361	22	based	base	VERB
easat-10684	361	23	on	on	ADP
easat-10684	361	24	these	these	DET
easat-10684	361	25	conditions	condition	NOUN
easat-10684	361	26	.	.	PUNCT
easat-10684	362	1	ultimately	ultimately	ADV
easat-10684	362	2	,	,	PUNCT
easat-10684	362	3	the	the	DET
easat-10684	362	4	leaf	leaf	NOUN
easat-10684	362	5	node	node	NOUN
easat-10684	362	6	holds	hold	VERB
easat-10684	362	7	a	a	DET
easat-10684	362	8	class	class	NOUN
easat-10684	362	9	label	label	NOUN
easat-10684	362	10	determined	determine	VERB
easat-10684	362	11	by	by	ADP
easat-10684	362	12	evaluating	evaluate	VERB
easat-10684	362	13	all	all	DET
easat-10684	362	14	attributes	attribute	NOUN
easat-10684	362	15	.	.	PUNCT
easat-10684	363	1	the	the	DET
easat-10684	363	2	distance	distance	NOUN
easat-10684	363	3	from	from	ADP
easat-10684	363	4	the	the	DET
easat-10684	363	5	root	root	NOUN
easat-10684	363	6	to	to	ADP
easat-10684	363	7	a	a	DET
easat-10684	363	8	leaf	leaf	NOUN
easat-10684	363	9	node	node	NOUN
easat-10684	363	10	defines	define	VERB
easat-10684	363	11	the	the	DET
easat-10684	363	12	classification	classification	NOUN
easat-10684	363	13	rule	rule	NOUN
easat-10684	363	14	.	.	PUNCT
easat-10684	364	1	3.1.8.5	3.1.8.5	X
easat-10684	364	2	.	.	PUNCT
easat-10684	364	3	logistic	logistic	ADJ
easat-10684	364	4	regression	regression	NOUN
easat-10684	364	5	logistic	logistic	ADJ
easat-10684	364	6	regression	regression	NOUN
easat-10684	364	7	is	be	AUX
easat-10684	364	8	a	a	DET
easat-10684	364	9	frequently	frequently	ADV
easat-10684	364	10	employed	employ	VERB
easat-10684	364	11	classification	classification	NOUN
easat-10684	364	12	method	method	NOUN
easat-10684	364	13	specifically	specifically	ADV
easat-10684	364	14	designed	design	VERB
easat-10684	364	15	for	for	ADP
easat-10684	364	16	assigning	assign	VERB
easat-10684	364	17	labels	label	NOUN
easat-10684	364	18	to	to	ADP
easat-10684	364	19	observations	observation	NOUN
easat-10684	364	20	within	within	ADP
easat-10684	364	21	a	a	DET
easat-10684	364	22	distinct	distinct	ADJ
easat-10684	364	23	set	set	NOUN
easat-10684	364	24	of	of	ADP
easat-10684	364	25	classes	class	NOUN
easat-10684	364	26	.	.	PUNCT
easat-10684	365	1	in	in	ADP
easat-10684	365	2	addressing	address	VERB
easat-10684	365	3	binary	binary	ADJ
easat-10684	365	4	classification	classification	NOUN
easat-10684	365	5	tasks	task	NOUN
easat-10684	365	6	,	,	PUNCT
easat-10684	365	7	logistic	logistic	ADJ
easat-10684	365	8	regression	regression	NOUN
easat-10684	365	9	has	have	AUX
easat-10684	365	10	proven	prove	VERB
easat-10684	365	11	to	to	PART
easat-10684	365	12	be	be	AUX
easat-10684	365	13	a	a	DET
easat-10684	365	14	successful	successful	ADJ
easat-10684	365	15	choice	choice	NOUN
easat-10684	366	1	[	[	X
easat-10684	366	2	47	47	NUM
easat-10684	366	3	]	]	PUNCT
easat-10684	366	4	.	.	PUNCT
easat-10684	367	1	the	the	DET
easat-10684	367	2	main	main	ADJ
easat-10684	367	3	strength	strength	NOUN
easat-10684	367	4	of	of	ADP
easat-10684	367	5	logistic	logistic	ADJ
easat-10684	367	6	regression	regression	NOUN
easat-10684	367	7	techniques	technique	NOUN
easat-10684	367	8	lies	lie	VERB
easat-10684	367	9	in	in	ADP
easat-10684	367	10	its	its	PRON
easat-10684	367	11	sigmoid	sigmoid	NOUN
easat-10684	367	12	function	function	NOUN
easat-10684	367	13	,	,	PUNCT
easat-10684	367	14	which	which	PRON
easat-10684	367	15	generates	generate	VERB
easat-10684	367	16	a	a	DET
easat-10684	367	17	probability	probability	NOUN
easat-10684	367	18	value	value	NOUN
easat-10684	367	19	that	that	PRON
easat-10684	367	20	is	be	AUX
easat-10684	367	21	subsequently	subsequently	ADV
easat-10684	367	22	linked	link	VERB
easat-10684	367	23	to	to	ADP
easat-10684	367	24	a	a	DET
easat-10684	367	25	class	class	NOUN
easat-10684	367	26	within	within	ADP
easat-10684	367	27	a	a	DET
easat-10684	367	28	discrete	discrete	ADJ
easat-10684	367	29	set	set	NOUN
easat-10684	367	30	for	for	ADP
easat-10684	367	31	comparing	compare	VERB
easat-10684	367	32	two	two	NUM
easat-10684	367	33	or	or	CCONJ
easat-10684	367	34	more	more	ADJ
easat-10684	367	35	classes	class	NOUN
easat-10684	367	36	[	[	X
easat-10684	367	37	47	47	NUM
easat-10684	367	38	]	]	PUNCT
easat-10684	367	39	.	.	PUNCT
easat-10684	368	1	𝑆(𝑥	𝑆(𝑥	X
easat-10684	368	2	)	)	PUNCT
easat-10684	368	3	=	=	SYM
easat-10684	368	4	1	1	NUM
easat-10684	368	5	1+𝑒−𝑥	1+𝑒−𝑥	NUM
easat-10684	368	6	(	(	PUNCT
easat-10684	368	7	2	2	NUM
easat-10684	368	8	)	)	PUNCT
easat-10684	368	9	1	1	NUM
easat-10684	369	1	+	+	CCONJ
easat-10684	369	2	e	e	NOUN
easat-10684	369	3	-	-	NOUN
easat-10684	369	4	x	x	INTJ
easat-10684	369	5	if	if	SCONJ
easat-10684	369	6	𝑆(𝑥	𝑆(𝑥	NUM
easat-10684	369	7	)	)	PUNCT
easat-10684	369	8	>	>	X
easat-10684	369	9	0.5	0.5	NUM
easat-10684	369	10	,	,	PUNCT
easat-10684	369	11	assign	assign	VERB
easat-10684	369	12	label	label	NOUN
easat-10684	369	13	=	=	SYM
easat-10684	369	14	class	class	NOUN
easat-10684	369	15	a	a	NOUN
easat-10684	369	16	,	,	PUNCT
easat-10684	369	17	else	else	ADV
easat-10684	369	18	assign	assign	NOUN
easat-10684	369	19	label	label	NOUN
easat-10684	369	20	=	=	SYM
easat-10684	369	21	class	class	NOUN
easat-10684	369	22	b	b	NOUN
easat-10684	369	23	,	,	PUNCT
easat-10684	369	24	where	where	SCONJ
easat-10684	369	25	0	0	X
easat-10684	369	26	<	<	X
easat-10684	369	27	𝑆(𝑥	𝑆(𝑥	X
easat-10684	369	28	)	)	PUNCT
easat-10684	369	29	<	<	X
easat-10684	369	30	1	1	NUM
easat-10684	369	31	and	and	CCONJ
easat-10684	369	32	(	(	PUNCT
easat-10684	369	33	𝐴	𝐴	PROPN
easat-10684	369	34	,	,	PUNCT
easat-10684	369	35	𝐵	𝐵	PROPN
easat-10684	369	36	)	)	PUNCT
easat-10684	369	37	are	be	AUX
easat-10684	369	38	sets	set	NOUN
easat-10684	369	39	of	of	ADP
easat-10684	369	40	classes	class	NOUN
easat-10684	369	41	[	[	X
easat-10684	369	42	47	47	NUM
easat-10684	369	43	]	]	PUNCT
easat-10684	369	44	.	.	PUNCT
easat-10684	370	1	3.1.9	3.1.9	NUM
easat-10684	370	2	.	.	PUNCT
easat-10684	370	3	model	model	NOUN
easat-10684	370	4	evaluation	evaluation	NOUN
easat-10684	370	5	it	it	PRON
easat-10684	370	6	's	be	AUX
easat-10684	370	7	essential	essential	ADJ
easat-10684	370	8	to	to	PART
easat-10684	370	9	evaluate	evaluate	VERB
easat-10684	370	10	the	the	DET
easat-10684	370	11	performance	performance	NOUN
easat-10684	370	12	of	of	ADP
easat-10684	370	13	the	the	DET
easat-10684	370	14	fake	fake	ADJ
easat-10684	370	15	news	news	NOUN
easat-10684	370	16	prediction	prediction	NOUN
easat-10684	370	17	models	model	NOUN
easat-10684	370	18	using	use	VERB
easat-10684	370	19	the	the	DET
easat-10684	370	20	test	test	NOUN
easat-10684	370	21	sample	sample	NOUN
easat-10684	370	22	data	datum	NOUN
easat-10684	370	23	and	and	CCONJ
easat-10684	370	24	measure	measure	VERB
easat-10684	370	25	their	their	PRON
easat-10684	370	26	performance	performance	NOUN
easat-10684	370	27	using	use	VERB
easat-10684	370	28	standard	standard	ADJ
easat-10684	370	29	performance	performance	NOUN
easat-10684	370	30	metrics	metric	NOUN
easat-10684	370	31	such	such	ADJ
easat-10684	370	32	as	as	ADP
easat-10684	370	33	accuracy	accuracy	NOUN
easat-10684	370	34	,	,	PUNCT
easat-10684	370	35	precision	precision	NOUN
easat-10684	370	36	,	,	PUNCT
easat-10684	370	37	recall	recall	NOUN
easat-10684	370	38	,	,	PUNCT
easat-10684	370	39	and	and	CCONJ
easat-10684	371	1	f	f	X
easat-10684	371	2	-	-	PUNCT
easat-10684	371	3	score	score	NOUN
easat-10684	371	4	.	.	PUNCT
easat-10684	372	1	3.1.9.1	3.1.9.1	X
easat-10684	372	2	.	.	PUNCT
easat-10684	373	1	accuracy	accuracy	NOUN
easat-10684	373	2	the	the	DET
easat-10684	373	3	accuracy	accuracy	NOUN
easat-10684	373	4	metric	metric	NOUN
easat-10684	373	5	is	be	AUX
easat-10684	373	6	used	use	VERB
easat-10684	373	7	when	when	SCONJ
easat-10684	373	8	the	the	DET
easat-10684	373	9	dependent	dependent	ADJ
easat-10684	373	10	or	or	CCONJ
easat-10684	373	11	output	output	NOUN
easat-10684	373	12	variable	variable	NOUN
easat-10684	373	13	is	be	AUX
easat-10684	373	14	categorical	categorical	ADJ
easat-10684	373	15	or	or	CCONJ
easat-10684	373	16	discrete	discrete	ADJ
easat-10684	373	17	.	.	PUNCT
easat-10684	374	1	it	it	PRON
easat-10684	374	2	represents	represent	VERB
easat-10684	374	3	the	the	DET
easat-10684	374	4	fraction	fraction	NOUN
easat-10684	374	5	of	of	ADP
easat-10684	374	6	time	time	NOUN
easat-10684	374	7	a	a	DET
easat-10684	374	8	model	model	NOUN
easat-10684	374	9	correctly	correctly	ADV
easat-10684	374	10	predicts	predict	VERB
easat-10684	374	11	the	the	DET
easat-10684	374	12	total	total	ADJ
easat-10684	374	13	predictions	prediction	NOUN
easat-10684	374	14	it	it	PRON
easat-10684	374	15	makes	make	VERB
easat-10684	374	16	[	[	X
easat-10684	374	17	48	48	NUM
easat-10684	374	18	]	]	PUNCT
easat-10684	374	19	.	.	PUNCT
easat-10684	375	1	accuracy	accuracy	NOUN
easat-10684	375	2	=	=	SYM
easat-10684	375	3	𝑇𝑃+𝑇𝑁	𝑇𝑃+𝑇𝑁	NOUN
easat-10684	375	4	𝑇𝑃+𝐹𝑃+𝐹𝑁+𝑇𝑁	𝑇𝑃+𝐹𝑃+𝐹𝑁+𝑇𝑁	NUM
easat-10684	375	5	(	(	PUNCT
easat-10684	375	6	3	3	NUM
easat-10684	375	7	)	)	PUNCT
easat-10684	375	8	1529	1529	NUM
easat-10684	375	9	edelweiss	edelweiss	PROPN
easat-10684	375	10	applied	apply	VERB
easat-10684	375	11	science	science	NOUN
easat-10684	375	12	and	and	CCONJ
easat-10684	375	13	technology	technology	NOUN
easat-10684	375	14	issn	issn	PROPN
easat-10684	375	15	:	:	PUNCT
easat-10684	375	16	2576	2576	NUM
easat-10684	375	17	-	-	SYM
easat-10684	375	18	8484	8484	NUM
easat-10684	375	19	vol	vol	NOUN
easat-10684	375	20	.	.	PROPN
easat-10684	376	1	9	9	NUM
easat-10684	376	2	,	,	PUNCT
easat-10684	376	3	no	no	INTJ
easat-10684	376	4	.	.	NOUN
easat-10684	376	5	10	10	NUM
easat-10684	376	6	:	:	PUNCT
easat-10684	376	7	1515	1515	NUM
easat-10684	376	8	-	-	SYM
easat-10684	376	9	1540	1540	NUM
easat-10684	376	10	,	,	PUNCT
easat-10684	376	11	2025	2025	NUM
easat-10684	376	12	doi	doi	NOUN
easat-10684	376	13	:	:	PUNCT
easat-10684	376	14	10.55214/2576	10.55214/2576	NUM
easat-10684	376	15	-	-	SYM
easat-10684	376	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	377	1	©	©	PROPN
easat-10684	377	2	2025	2025	NUM
easat-10684	377	3	by	by	ADP
easat-10684	377	4	the	the	DET
easat-10684	377	5	authors	author	NOUN
easat-10684	377	6	;	;	PUNCT
easat-10684	377	7	licensee	licensee	PROPN
easat-10684	377	8	learning	learn	VERB
easat-10684	377	9	gate	gate	NOUN
easat-10684	377	10	3.1.9.2	3.1.9.2	PROPN
easat-10684	377	11	.	.	PUNCT
easat-10684	378	1	precision	precision	VERB
easat-10684	378	2	the	the	DET
easat-10684	378	3	precision	precision	NOUN
easat-10684	378	4	of	of	ADP
easat-10684	378	5	text	text	NOUN
easat-10684	378	6	classification	classification	NOUN
easat-10684	378	7	indicates	indicate	VERB
easat-10684	378	8	the	the	DET
easat-10684	378	9	level	level	NOUN
easat-10684	378	10	of	of	ADP
easat-10684	378	11	exact	exact	ADJ
easat-10684	378	12	prediction	prediction	NOUN
easat-10684	378	13	carried	carry	VERB
easat-10684	378	14	out	out	ADP
easat-10684	378	15	by	by	ADP
easat-10684	378	16	the	the	DET
easat-10684	378	17	model	model	NOUN
easat-10684	379	1	[	[	X
easat-10684	379	2	48	48	NUM
easat-10684	379	3	]	]	PUNCT
easat-10684	379	4	.	.	PUNCT
easat-10684	380	1	technically	technically	ADV
easat-10684	380	2	,	,	PUNCT
easat-10684	380	3	given	give	VERB
easat-10684	380	4	all	all	DET
easat-10684	380	5	the	the	DET
easat-10684	380	6	positive	positive	ADJ
easat-10684	380	7	cases	case	NOUN
easat-10684	380	8	(	(	PUNCT
easat-10684	380	9	thus	thus	ADV
easat-10684	380	10	the	the	DET
easat-10684	380	11	relevant	relevant	ADJ
easat-10684	380	12	classes	class	NOUN
easat-10684	380	13	)	)	PUNCT
easat-10684	380	14	,	,	PUNCT
easat-10684	380	15	to	to	ADP
easat-10684	380	16	how	how	SCONJ
easat-10684	380	17	many	many	ADJ
easat-10684	380	18	classes	class	NOUN
easat-10684	380	19	are	be	AUX
easat-10684	380	20	correctly	correctly	ADV
easat-10684	380	21	classified	classify	VERB
easat-10684	380	22	by	by	ADP
easat-10684	380	23	the	the	DET
easat-10684	380	24	model	model	NOUN
easat-10684	380	25	[	[	X
easat-10684	380	26	48	48	NUM
easat-10684	380	27	]	]	PUNCT
easat-10684	380	28	.	.	PUNCT
easat-10684	381	1	precision	precision	NOUN
easat-10684	381	2	=	=	SYM
easat-10684	381	3	𝑇𝑃	𝑇𝑃	PROPN
easat-10684	381	4	𝑇𝑃+𝐹𝑃	𝑇𝑃+𝐹𝑃	NUM
easat-10684	381	5	(	(	PUNCT
easat-10684	381	6	4	4	NUM
easat-10684	381	7	)	)	PUNCT
easat-10684	381	8	3.1.9.3	3.1.9.3	NUM
easat-10684	381	9	.	.	PUNCT
easat-10684	381	10	recall	recall	VERB
easat-10684	381	11	the	the	DET
easat-10684	381	12	recall	recall	NOUN
easat-10684	381	13	and	and	CCONJ
easat-10684	381	14	precision	precision	NOUN
easat-10684	381	15	metrics	metric	NOUN
easat-10684	381	16	are	be	AUX
easat-10684	381	17	complementary	complementary	ADJ
easat-10684	381	18	to	to	ADP
easat-10684	381	19	each	each	DET
easat-10684	381	20	other	other	ADJ
easat-10684	381	21	.	.	PUNCT
easat-10684	382	1	recall	recall	PROPN
easat-10684	382	2	indicates	indicate	VERB
easat-10684	382	3	how	how	SCONJ
easat-10684	382	4	well	well	ADV
easat-10684	382	5	a	a	DET
easat-10684	382	6	model	model	NOUN
easat-10684	382	7	can	can	AUX
easat-10684	382	8	identify	identify	VERB
easat-10684	382	9	positive	positive	ADJ
easat-10684	382	10	instances	instance	NOUN
easat-10684	382	11	in	in	ADP
easat-10684	382	12	a	a	DET
easat-10684	382	13	problem	problem	NOUN
easat-10684	382	14	statement	statement	NOUN
easat-10684	382	15	.	.	PUNCT
easat-10684	383	1	it	it	PRON
easat-10684	383	2	is	be	AUX
easat-10684	383	3	calculated	calculate	VERB
easat-10684	383	4	based	base	VERB
easat-10684	383	5	on	on	ADP
easat-10684	383	6	all	all	DET
easat-10684	383	7	the	the	DET
easat-10684	383	8	positive	positive	ADJ
easat-10684	383	9	predictions	prediction	NOUN
easat-10684	383	10	made	make	VERB
easat-10684	383	11	by	by	ADP
easat-10684	383	12	the	the	DET
easat-10684	383	13	model	model	NOUN
easat-10684	383	14	and	and	CCONJ
easat-10684	383	15	the	the	DET
easat-10684	383	16	proportion	proportion	NOUN
easat-10684	383	17	of	of	ADP
easat-10684	383	18	these	these	DET
easat-10684	383	19	predictions	prediction	NOUN
easat-10684	383	20	that	that	PRON
easat-10684	383	21	are	be	AUX
easat-10684	383	22	truly	truly	ADV
easat-10684	383	23	positive	positive	ADJ
easat-10684	384	1	[	[	X
easat-10684	384	2	48	48	NUM
easat-10684	384	3	]	]	PUNCT
easat-10684	384	4	.	.	PUNCT
easat-10684	385	1	recall	recall	NOUN
easat-10684	385	2	=	=	SYM
easat-10684	385	3	𝑇𝑃	𝑇𝑃	PROPN
easat-10684	385	4	𝑇𝑃+𝐹𝑁	𝑇𝑃+𝐹𝑁	NOUN
easat-10684	385	5	(	(	PUNCT
easat-10684	385	6	5	5	NUM
easat-10684	385	7	)	)	PUNCT
easat-10684	385	8	3.1.9.4	3.1.9.4	NUM
easat-10684	385	9	.	.	PUNCT
easat-10684	385	10	f1	f1	NOUN
easat-10684	385	11	-	-	PUNCT
easat-10684	385	12	score	score	NOUN
easat-10684	385	13	the	the	DET
easat-10684	385	14	f1	f1	NOUN
easat-10684	385	15	-	-	PUNCT
easat-10684	385	16	score	score	NOUN
easat-10684	385	17	metric	metric	ADJ
easat-10684	385	18	combines	combine	NOUN
easat-10684	385	19	precision	precision	NOUN
easat-10684	385	20	and	and	CCONJ
easat-10684	385	21	recall	recall	VERB
easat-10684	385	22	into	into	ADP
easat-10684	385	23	a	a	DET
easat-10684	385	24	single	single	ADJ
easat-10684	385	25	measure	measure	NOUN
easat-10684	385	26	;	;	PUNCT
easat-10684	385	27	it	it	PRON
easat-10684	385	28	also	also	ADV
easat-10684	385	29	captures	capture	VERB
easat-10684	385	30	the	the	DET
easat-10684	385	31	tradeoff	tradeoff	NOUN
easat-10684	385	32	between	between	ADP
easat-10684	385	33	these	these	DET
easat-10684	385	34	metrics	metric	NOUN
easat-10684	385	35	,	,	PUNCT
easat-10684	385	36	representing	represent	VERB
easat-10684	385	37	both	both	CCONJ
easat-10684	385	38	completeness	completeness	NOUN
easat-10684	385	39	and	and	CCONJ
easat-10684	385	40	accuracy	accuracy	NOUN
easat-10684	385	41	[	[	X
easat-10684	385	42	48	48	NUM
easat-10684	385	43	]	]	PUNCT
easat-10684	385	44	.	.	PUNCT
easat-10684	386	1	f1	f1	NOUN
easat-10684	386	2	-	-	PUNCT
easat-10684	386	3	score	score	NOUN
easat-10684	386	4	=	=	PUNCT
easat-10684	386	5	2×𝑃𝑖×𝑅𝑖	2×𝑃𝑖×𝑅𝑖	PROPN
easat-10684	386	6	𝑃𝑖+𝑅𝑖	𝑃𝑖+𝑅𝑖	NOUN
easat-10684	386	7	(	(	PUNCT
easat-10684	386	8	6	6	NUM
easat-10684	386	9	)	)	PUNCT
easat-10684	386	10	3.2	3.2	NUM
easat-10684	386	11	.	.	PUNCT
easat-10684	387	1	system	system	NOUN
easat-10684	387	2	analysis	analysis	NOUN
easat-10684	387	3	and	and	CCONJ
easat-10684	387	4	design	design	VERB
easat-10684	387	5	this	this	DET
easat-10684	387	6	subsection	subsection	NOUN
easat-10684	387	7	introduces	introduce	VERB
easat-10684	387	8	various	various	ADJ
easat-10684	387	9	conceptual	conceptual	ADJ
easat-10684	387	10	diagrams	diagram	NOUN
easat-10684	387	11	to	to	PART
easat-10684	387	12	illustrate	illustrate	VERB
easat-10684	387	13	how	how	SCONJ
easat-10684	387	14	the	the	DET
easat-10684	387	15	proposed	propose	VERB
easat-10684	387	16	fake	fake	ADJ
easat-10684	387	17	news	news	NOUN
easat-10684	387	18	detection	detection	NOUN
easat-10684	387	19	model	model	NOUN
easat-10684	387	20	will	will	AUX
easat-10684	387	21	be	be	AUX
easat-10684	387	22	designed	design	VERB
easat-10684	387	23	and	and	CCONJ
easat-10684	387	24	developed	develop	VERB
easat-10684	387	25	.	.	PUNCT
easat-10684	388	1	the	the	DET
easat-10684	388	2	system	system	NOUN
easat-10684	388	3	framework	framework	NOUN
easat-10684	388	4	and	and	CCONJ
easat-10684	388	5	flow	flow	NOUN
easat-10684	388	6	chart	chart	ADJ
easat-10684	388	7	diagram	diagram	NOUN
easat-10684	388	8	are	be	AUX
easat-10684	388	9	designed	design	VERB
easat-10684	388	10	using	use	VERB
easat-10684	388	11	the	the	DET
easat-10684	388	12	microsoft	microsoft	PROPN
easat-10684	388	13	visio	visio	PROPN
easat-10684	388	14	application	application	NOUN
easat-10684	388	15	.	.	PUNCT
easat-10684	389	1	finally	finally	ADV
easat-10684	389	2	,	,	PUNCT
easat-10684	389	3	the	the	DET
easat-10684	389	4	bi	bi	NOUN
easat-10684	389	5	-	-	PUNCT
easat-10684	389	6	lstm-2	lstm-2	NOUN
easat-10684	389	7	-	-	PUNCT
easat-10684	389	8	ml	ml	NOUN
easat-10684	389	9	algorithm	algorithm	NOUN
easat-10684	389	10	will	will	AUX
easat-10684	389	11	also	also	ADV
easat-10684	389	12	be	be	AUX
easat-10684	389	13	illustrated	illustrate	VERB
easat-10684	389	14	in	in	ADP
easat-10684	389	15	this	this	DET
easat-10684	389	16	subsection	subsection	NOUN
easat-10684	389	17	.	.	PUNCT
easat-10684	390	1	algorithm	algorithm	PROPN
easat-10684	390	2	2	2	NUM
easat-10684	390	3	:	:	PUNCT
easat-10684	390	4	bi	bi	NOUN
easat-10684	390	5	-	-	ADJ
easat-10684	390	6	lstm-2	lstm-2	NOUN
easat-10684	390	7	-	-	PUNCT
easat-10684	390	8	ml	ml	NOUN
easat-10684	390	9	transfer	transfer	NOUN
easat-10684	390	10	learning	learning	NOUN
easat-10684	390	11	-	-	PUNCT
easat-10684	390	12	based	base	VERB
easat-10684	390	13	model	model	NOUN
easat-10684	390	14	input	input	NOUN
easat-10684	390	15	:	:	PUNCT
easat-10684	390	16	fake	fake	ADJ
easat-10684	390	17	news	news	NOUN
easat-10684	390	18	dataset	dataset	VERB
easat-10684	390	19	output	output	NOUN
easat-10684	390	20	:	:	PUNCT
easat-10684	390	21	evaluation	evaluation	NOUN
easat-10684	390	22	results	result	NOUN
easat-10684	390	23	r	r	NOUN
easat-10684	390	24	1	1	NUM
easat-10684	390	25	2	2	NUM
easat-10684	390	26	3	3	NUM
easat-10684	390	27	𝟑.	𝟑.	ADP
easat-10684	390	28	𝟐	𝟐	NUM
easat-10684	390	29	𝑹𝒆𝒎𝒐𝒗𝒆	𝑹𝒆𝒎𝒐𝒗𝒆	NOUN
easat-10684	390	30	𝒑𝒖𝒏𝒄𝒕𝒖𝒂𝒕𝒊𝒐𝒏	𝒑𝒖𝒏𝒄𝒕𝒖𝒂𝒕𝒊𝒐𝒏	NOUN
easat-10684	390	31	𝒎𝒂𝒓𝒌𝒔	𝒎𝒂𝒓𝒌𝒔	VERB
easat-10684	390	32	4	4	NUM
easat-10684	390	33	5	5	NUM
easat-10684	390	34	6	6	NUM
easat-10684	390	35	𝑆𝑡𝑎𝑟𝑡	𝑆𝑡𝑎𝑟𝑡	PROPN
easat-10684	390	36	𝐿𝑜𝑎𝑑	𝐿𝑜𝑎𝑑	PROPN
easat-10684	390	37	𝑡ℎ𝑒	𝑡ℎ𝑒	NUM
easat-10684	390	38	𝑙𝑎𝑏𝑒𝑙𝑒𝑑	𝑙𝑎𝑏𝑒𝑙𝑒𝑑	NOUN
easat-10684	390	39	𝑓𝑎𝑘𝑒	𝑓𝑎𝑘𝑒	PROPN
easat-10684	390	40	𝑛𝑒𝑤𝑠	𝑛𝑒𝑤𝑠	NOUN
easat-10684	390	41	𝑑𝑎𝑡𝑎𝑠𝑒𝑡	𝑑𝑎𝑡𝑎𝑠𝑒𝑡	NOUN
easat-10684	390	42	𝐷𝑎𝑡𝑎	𝐷𝑎𝑡𝑎	PROPN
easat-10684	390	43	𝐶𝑙𝑒𝑎𝑛𝑖𝑛𝑔	𝐶𝑙𝑒𝑎𝑛𝑖𝑛𝑔	PROPN
easat-10684	390	44	3.1	3.1	NUM
easat-10684	390	45	𝑇𝑜𝑘𝑒𝑛𝑖𝑧𝑒	𝑇𝑜𝑘𝑒𝑛𝑖𝑧𝑒	PROPN
easat-10684	390	46	𝑡ℎ𝑒	𝑡ℎ𝑒	VERB
easat-10684	390	47	𝑡𝑒𝑥𝑡	𝑡𝑒𝑥𝑡	NOUN
easat-10684	390	48	3.2	3.2	NUM
easat-10684	390	49	𝑅𝑒𝑚𝑜𝑣𝑒	𝑅𝑒𝑚𝑜𝑣𝑒	PROPN
easat-10684	390	50	𝑝𝑢𝑛𝑐𝑡𝑢𝑎𝑡𝑖𝑜𝑛	𝑝𝑢𝑛𝑐𝑡𝑢𝑎𝑡𝑖𝑜𝑛	NOUN
easat-10684	390	51	𝑚𝑎𝑟𝑘𝑠	𝑚𝑎𝑟𝑘𝑠	NOUN
easat-10684	390	52	3.3	3.3	NUM
easat-10684	390	53	𝑅𝑒𝑚𝑜𝑣𝑒	𝑅𝑒𝑚𝑜𝑣𝑒	PROPN
easat-10684	390	54	𝑠𝑡𝑜𝑝	𝑠𝑡𝑜𝑝	ADJ
easat-10684	390	55	𝑤𝑜𝑟𝑑𝑠	𝑤𝑜𝑟𝑑𝑠	NOUN
easat-10684	390	56	3.4	3.4	NUM
easat-10684	390	57	apply	apply	VERB
easat-10684	390	58	stemming	stem	VERB
easat-10684	390	59	using	use	VERB
easat-10684	390	60	porterstemmer	porterstemmer	NOUN
easat-10684	390	61	3.5	3.5	NUM
easat-10684	390	62	convert	convert	VERB
easat-10684	390	63	all	all	DET
easat-10684	390	64	text	text	NOUN
easat-10684	390	65	to	to	PART
easat-10684	390	66	lowercase	lowercase	VERB
easat-10684	390	67	.	.	PUNCT
easat-10684	391	1	replace	replace	VERB
easat-10684	391	2	label	label	NOUN
easat-10684	391	3	‘	'	PUNCT
easat-10684	391	4	fake	fake	ADJ
easat-10684	391	5	’	'	PUNCT
easat-10684	391	6	with	with	ADP
easat-10684	391	7	0	0	NUM
easat-10684	391	8	and	and	CCONJ
easat-10684	391	9	‘	'	PUNCT
easat-10684	391	10	real	real	ADJ
easat-10684	391	11	’	'	PUNCT
easat-10684	391	12	with	with	ADP
easat-10684	391	13	1	1	NUM
easat-10684	391	14	apply	apply	VERB
easat-10684	391	15	word	word	NOUN
easat-10684	391	16	embeddings	embedding	NOUN
easat-10684	391	17	using	use	VERB
easat-10684	391	18	word2vec	word2vec	X
easat-10684	391	19	(	(	PUNCT
easat-10684	391	20	)	)	PUNCT
easat-10684	391	21	𝐴𝑝𝑝𝑙𝑦	𝐴𝑝𝑝𝑙𝑦	PROPN
easat-10684	391	22	𝑡𝑒𝑛𝑠𝑜𝑟𝑡𝑜𝑘𝑒𝑛𝑖𝑧𝑎𝑡𝑖𝑜𝑛	𝑡𝑒𝑛𝑠𝑜𝑟𝑡𝑜𝑘𝑒𝑛𝑖𝑧𝑎𝑡𝑖𝑜𝑛	NOUN
easat-10684	391	23	(	(	PUNCT
easat-10684	391	24	)	)	PUNCT
easat-10684	391	25	𝑓𝑜𝑟	𝑓𝑜𝑟	ADV
easat-10684	391	26	𝑓𝑒𝑎𝑡𝑢𝑟𝑒	𝑓𝑒𝑎𝑡𝑢𝑟𝑒	VERB
easat-10684	391	27	𝑒𝑥𝑡𝑟𝑎𝑐𝑡𝑖𝑜𝑛	𝑒𝑥𝑡𝑟𝑎𝑐𝑡𝑖𝑜𝑛	NOUN
easat-10684	391	28	7	7	NUM
easat-10684	391	29	8	8	NUM
easat-10684	391	30	9	9	NUM
easat-10684	391	31	10	10	NUM
easat-10684	391	32	11	11	NUM
easat-10684	391	33	12	12	NUM
easat-10684	391	34	13	13	NUM
easat-10684	392	1	𝑈𝑠𝑒	𝑈𝑠𝑒	PROPN
easat-10684	392	2	𝑝𝑎𝑑𝑑𝑖𝑛𝑔𝑠𝑒𝑞𝑢𝑒𝑛𝑐𝑒	𝑝𝑎𝑑𝑑𝑖𝑛𝑔𝑠𝑒𝑞𝑢𝑒𝑛𝑐𝑒	NOUN
easat-10684	392	3	(	(	PUNCT
easat-10684	392	4	)	)	PUNCT
easat-10684	392	5	𝑡𝑜	𝑡𝑜	PROPN
easat-10684	392	6	𝑒𝑛𝑠𝑢𝑟𝑒	𝑒𝑛𝑠𝑢𝑟𝑒	VERB
easat-10684	392	7	𝑒𝑢𝑎𝑙	𝑒𝑢𝑎𝑙	NOUN
easat-10684	392	8	𝑙𝑒𝑛𝑔𝑡ℎ	𝑙𝑒𝑛𝑔𝑡ℎ	PROPN
easat-10684	392	9	𝑝𝑎𝑑	𝑝𝑎𝑑	PROPN
easat-10684	392	10	𝑠𝑒𝑞𝑢𝑒𝑛𝑐𝑒𝑠	𝑠𝑒𝑞𝑢𝑒𝑛𝑐𝑒𝑠	NOUN
easat-10684	392	11	𝑆𝑝𝑙𝑖𝑡	𝑆𝑝𝑙𝑖𝑡	PROPN
easat-10684	392	12	𝑑𝑎𝑡𝑎𝑠𝑒𝑡	𝑑𝑎𝑡𝑎𝑠𝑒𝑡	PROPN
easat-10684	392	13	𝑤𝑖𝑡ℎ	𝑤𝑖𝑡ℎ	NOUN
easat-10684	392	14	𝑡𝑟𝑎𝑖𝑛𝑠𝑖𝑧𝑒	𝑡𝑟𝑎𝑖𝑛𝑠𝑖𝑧𝑒	PROPN
easat-10684	392	15	=	=	SYM
easat-10684	392	16	70	70	NUM
easat-10684	392	17	𝑎𝑛𝑑	𝑎𝑛𝑑	ADJ
easat-10684	392	18	𝑡𝑒𝑠𝑡𝑠𝑖𝑧𝑒	𝑡𝑒𝑠𝑡𝑠𝑖𝑧𝑒	NOUN
easat-10684	392	19	=	=	SYM
easat-10684	392	20	30	30	NUM
easat-10684	392	21	build	build	VERB
easat-10684	392	22	and	and	CCONJ
easat-10684	392	23	train	train	VERB
easat-10684	392	24	a	a	DET
easat-10684	392	25	bi	bi	ADJ
easat-10684	392	26	-	-	ADJ
easat-10684	392	27	lstm	lstm	ADJ
easat-10684	392	28	model	model	NOUN
easat-10684	392	29	using	use	VERB
easat-10684	392	30	learning.train	learning.train	PROPN
easat-10684	392	31	(	(	PUNCT
easat-10684	392	32	)	)	PUNCT
easat-10684	392	33	use	use	VERB
easat-10684	392	34	the	the	DET
easat-10684	392	35	trained	train	VERB
easat-10684	392	36	bi	bi	ADJ
easat-10684	392	37	-	-	ADJ
easat-10684	392	38	lstm	lstm	ADJ
easat-10684	392	39	model	model	NOUN
easat-10684	392	40	as	as	SCONJ
easat-10684	392	41	a	a	DET
easat-10684	392	42	feature	feature	NOUN
easat-10684	392	43	extractor	extractor	NOUN
easat-10684	392	44	extract	extract	VERB
easat-10684	392	45	bi	bi	ADJ
easat-10684	392	46	-	-	ADJ
easat-10684	392	47	lstm	lstm	ADJ
easat-10684	392	48	features	feature	NOUN
easat-10684	392	49	.	.	PUNCT
easat-10684	393	1	train	train	NOUN
easat-10684	393	2	multiple	multiple	ADJ
easat-10684	393	3	ml	ml	ADP
easat-10684	393	4	classifiers	classifier	NOUN
easat-10684	393	5	on	on	ADP
easat-10684	393	6	extracted	extract	VERB
easat-10684	393	7	features	feature	NOUN
easat-10684	393	8	(	(	PUNCT
easat-10684	393	9	knn	knn	PROPN
easat-10684	393	10	,	,	PUNCT
easat-10684	393	11	gnb	gnb	PROPN
easat-10684	393	12	,	,	PUNCT
easat-10684	393	13	svm	svm	NOUN
easat-10684	393	14	and	and	CCONJ
easat-10684	393	15	dt	dt	NOUN
easat-10684	393	16	)	)	PUNCT
easat-10684	393	17	evaluate	evaluate	VERB
easat-10684	393	18	each	each	DET
easat-10684	393	19	ml	ml	NOUN
easat-10684	393	20	model	model	NOUN
easat-10684	393	21	1530	1530	NUM
easat-10684	393	22	edelweiss	edelweiss	PROPN
easat-10684	393	23	applied	apply	VERB
easat-10684	393	24	science	science	NOUN
easat-10684	393	25	and	and	CCONJ
easat-10684	393	26	technology	technology	NOUN
easat-10684	393	27	issn	issn	PROPN
easat-10684	393	28	:	:	PUNCT
easat-10684	393	29	2576	2576	NUM
easat-10684	393	30	-	-	SYM
easat-10684	393	31	8484	8484	NUM
easat-10684	393	32	vol	vol	NOUN
easat-10684	393	33	.	.	PROPN
easat-10684	394	1	9	9	NUM
easat-10684	394	2	,	,	PUNCT
easat-10684	394	3	no	no	INTJ
easat-10684	394	4	.	.	NOUN
easat-10684	394	5	10	10	NUM
easat-10684	394	6	:	:	PUNCT
easat-10684	394	7	1515	1515	NUM
easat-10684	394	8	-	-	SYM
easat-10684	394	9	1540	1540	NUM
easat-10684	394	10	,	,	PUNCT
easat-10684	394	11	2025	2025	NUM
easat-10684	394	12	doi	doi	NOUN
easat-10684	394	13	:	:	PUNCT
easat-10684	394	14	10.55214/2576	10.55214/2576	NUM
easat-10684	394	15	-	-	SYM
easat-10684	394	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	395	1	©	©	PROPN
easat-10684	395	2	2025	2025	NUM
easat-10684	395	3	by	by	ADP
easat-10684	395	4	the	the	DET
easat-10684	395	5	authors	author	NOUN
easat-10684	395	6	;	;	PUNCT
easat-10684	395	7	licensee	licensee	PROPN
easat-10684	395	8	learning	learning	NOUN
easat-10684	395	9	gate	gate	NOUN
easat-10684	395	10	14	14	NUM
easat-10684	395	11	15	15	NUM
easat-10684	395	12	store	store	NOUN
easat-10684	395	13	results	result	NOUN
easat-10684	395	14	in	in	ADP
easat-10684	395	15	r	r	NOUN
easat-10684	395	16	return	return	NOUN
easat-10684	395	17	r	r	NOUN
easat-10684	395	18	𝐸𝑁𝐷	𝐸𝑁𝐷	PROPN
easat-10684	395	19	algorithm	algorithm	NOUN
easat-10684	395	20	2	2	NUM
easat-10684	395	21	illustrates	illustrate	VERB
easat-10684	395	22	the	the	DET
easat-10684	395	23	step	step	NOUN
easat-10684	395	24	-	-	PUNCT
easat-10684	395	25	by	by	ADP
easat-10684	395	26	-	-	PUNCT
easat-10684	395	27	step	step	NOUN
easat-10684	395	28	approach	approach	NOUN
easat-10684	395	29	for	for	ADP
easat-10684	395	30	developing	develop	VERB
easat-10684	395	31	the	the	DET
easat-10684	395	32	proposed	propose	VERB
easat-10684	395	33	bi	bi	NOUN
easat-10684	395	34	-	-	PUNCT
easat-10684	395	35	lstm-2	lstm-2	NOUN
easat-10684	395	36	-	-	PUNCT
easat-10684	395	37	ml	ml	ADV
easat-10684	395	38	fake	fake	ADJ
easat-10684	395	39	news	news	NOUN
easat-10684	395	40	detection	detection	NOUN
easat-10684	395	41	model	model	NOUN
easat-10684	395	42	.	.	PUNCT
easat-10684	396	1	the	the	DET
easat-10684	396	2	algorithm	algorithm	NOUN
easat-10684	396	3	is	be	AUX
easat-10684	396	4	divided	divide	VERB
easat-10684	396	5	into	into	ADP
easat-10684	396	6	nine	nine	NUM
easat-10684	396	7	major	major	ADJ
easat-10684	396	8	steps	step	NOUN
easat-10684	396	9	,	,	PUNCT
easat-10684	396	10	each	each	PRON
easat-10684	396	11	with	with	ADP
easat-10684	396	12	associated	associated	ADJ
easat-10684	396	13	subtasks	subtask	NOUN
easat-10684	396	14	.	.	PUNCT
easat-10684	397	1	these	these	DET
easat-10684	397	2	steps	step	NOUN
easat-10684	397	3	include	include	VERB
easat-10684	397	4	data	datum	NOUN
easat-10684	397	5	loading	loading	NOUN
easat-10684	397	6	,	,	PUNCT
easat-10684	397	7	data	datum	NOUN
easat-10684	397	8	cleaning	cleaning	NOUN
easat-10684	397	9	,	,	PUNCT
easat-10684	397	10	data	datum	NOUN
easat-10684	397	11	transformation	transformation	NOUN
easat-10684	397	12	,	,	PUNCT
easat-10684	397	13	feature	feature	NOUN
easat-10684	397	14	extraction	extraction	NOUN
easat-10684	397	15	,	,	PUNCT
easat-10684	397	16	data	datum	NOUN
easat-10684	397	17	splitting	splitting	NOUN
easat-10684	397	18	,	,	PUNCT
easat-10684	397	19	bi	bi	ADJ
easat-10684	397	20	-	-	ADJ
easat-10684	397	21	lstm	lstm	ADJ
easat-10684	397	22	deep	deep	ADJ
easat-10684	397	23	learning	learning	NOUN
easat-10684	397	24	training	training	NOUN
easat-10684	397	25	,	,	PUNCT
easat-10684	397	26	machine	machine	NOUN
easat-10684	397	27	learning	learning	NOUN
easat-10684	397	28	using	use	VERB
easat-10684	397	29	transfer	transfer	NOUN
easat-10684	397	30	learning	learning	NOUN
easat-10684	397	31	,	,	PUNCT
easat-10684	397	32	and	and	CCONJ
easat-10684	397	33	model	model	NOUN
easat-10684	397	34	evaluation	evaluation	NOUN
easat-10684	397	35	.	.	PUNCT
easat-10684	398	1	4	4	X
easat-10684	398	2	.	.	X
easat-10684	398	3	implementation	implementation	NOUN
easat-10684	398	4	the	the	DET
easat-10684	398	5	pre	pre	ADJ
easat-10684	398	6	-	-	ADJ
easat-10684	398	7	trained	train	VERB
easat-10684	398	8	bi	bi	ADJ
easat-10684	398	9	-	-	ADJ
easat-10684	398	10	lstm	lstm	ADJ
easat-10684	398	11	model	model	NOUN
easat-10684	398	12	implementation	implementation	NOUN
easat-10684	398	13	,	,	PUNCT
easat-10684	398	14	along	along	ADP
easat-10684	398	15	with	with	ADP
easat-10684	398	16	the	the	DET
easat-10684	398	17	bi	bi	ADJ
easat-10684	398	18	-	-	ADJ
easat-10684	398	19	lstm	lstm	ADJ
easat-10684	398	20	transfer	transfer	NOUN
easat-10684	398	21	learning	learning	NOUN
easat-10684	398	22	for	for	ADP
easat-10684	398	23	machine	machine	NOUN
easat-10684	398	24	learning	learning	NOUN
easat-10684	398	25	training	training	NOUN
easat-10684	398	26	,	,	PUNCT
easat-10684	398	27	will	will	AUX
easat-10684	398	28	be	be	AUX
easat-10684	398	29	comprehensively	comprehensively	ADV
easat-10684	398	30	discussed	discuss	VERB
easat-10684	398	31	in	in	ADP
easat-10684	398	32	this	this	DET
easat-10684	398	33	sub	sub	NOUN
easat-10684	398	34	-	-	NOUN
easat-10684	398	35	section	section	NOUN
easat-10684	398	36	.	.	PUNCT
easat-10684	399	1	marjory	marjory	ADJ
easat-10684	399	2	python	python	NOUN
easat-10684	399	3	libraries	library	NOUN
easat-10684	399	4	such	such	ADJ
easat-10684	399	5	as	as	ADP
easat-10684	399	6	sklearn	sklearn	ADJ
easat-10684	399	7	,	,	PUNCT
easat-10684	399	8	tensorflow	tensorflow	NOUN
easat-10684	399	9	,	,	PUNCT
easat-10684	399	10	and	and	CCONJ
easat-10684	399	11	keras	keras	PROPN
easat-10684	399	12	will	will	AUX
easat-10684	399	13	be	be	AUX
easat-10684	399	14	used	use	VERB
easat-10684	399	15	for	for	ADP
easat-10684	399	16	training	train	VERB
easat-10684	399	17	the	the	DET
easat-10684	399	18	deep	deep	ADJ
easat-10684	399	19	/	/	SYM
easat-10684	399	20	machine	machine	NOUN
easat-10684	399	21	learning	learning	NOUN
easat-10684	399	22	models	model	NOUN
easat-10684	399	23	.	.	PUNCT
easat-10684	400	1	firstly	firstly	ADV
easat-10684	400	2	,	,	PUNCT
easat-10684	400	3	a	a	DET
easat-10684	400	4	sequential	sequential	ADJ
easat-10684	400	5	-	-	PUNCT
easat-10684	400	6	based	base	VERB
easat-10684	400	7	bi	bi	ADJ
easat-10684	400	8	-	-	ADJ
easat-10684	400	9	lstm	lstm	ADJ
easat-10684	400	10	model	model	NOUN
easat-10684	400	11	is	be	AUX
easat-10684	400	12	built	build	VERB
easat-10684	400	13	,	,	PUNCT
easat-10684	400	14	compiled	compile	VERB
easat-10684	400	15	,	,	PUNCT
easat-10684	400	16	and	and	CCONJ
easat-10684	400	17	summarized	summarize	VERB
easat-10684	400	18	before	before	SCONJ
easat-10684	400	19	the	the	DET
easat-10684	400	20	fake	fake	ADJ
easat-10684	400	21	news	news	NOUN
easat-10684	400	22	training	training	NOUN
easat-10684	400	23	process	process	NOUN
easat-10684	400	24	is	be	AUX
easat-10684	400	25	started	start	VERB
easat-10684	400	26	.	.	PUNCT
easat-10684	401	1	4.1	4.1	NUM
easat-10684	401	2	.	.	PUNCT
easat-10684	401	3	deep	deep	ADJ
easat-10684	401	4	learning	learn	VERB
easat-10684	401	5	implementation	implementation	NOUN
easat-10684	401	6	(	(	PUNCT
easat-10684	401	7	bi	bi	NOUN
easat-10684	401	8	-	-	ADJ
easat-10684	401	9	lstm	lstm	ADJ
easat-10684	401	10	)	)	PUNCT
easat-10684	401	11	figure	figure	NOUN
easat-10684	401	12	6	6	NUM
easat-10684	401	13	.	.	PUNCT
easat-10684	402	1	bi	bi	ADJ
easat-10684	402	2	-	-	ADJ
easat-10684	402	3	lstm	lstm	ADJ
easat-10684	402	4	sequential	sequential	ADJ
easat-10684	402	5	model	model	NOUN
easat-10684	402	6	architecture	architecture	NOUN
easat-10684	402	7	.	.	PUNCT
easat-10684	403	1	based	base	VERB
easat-10684	403	2	on	on	ADP
easat-10684	403	3	figure	figure	NOUN
easat-10684	403	4	6	6	NUM
easat-10684	403	5	,	,	PUNCT
easat-10684	403	6	the	the	DET
easat-10684	403	7	bi	bi	ADJ
easat-10684	403	8	-	-	ADJ
easat-10684	403	9	lstm	lstm	ADJ
easat-10684	403	10	sequential	sequential	ADJ
easat-10684	403	11	model	model	NOUN
easat-10684	403	12	instance	instance	NOUN
easat-10684	403	13	is	be	AUX
easat-10684	403	14	created	create	VERB
easat-10684	403	15	,	,	PUNCT
easat-10684	403	16	along	along	ADP
easat-10684	403	17	with	with	ADP
easat-10684	403	18	other	other	ADJ
easat-10684	403	19	essential	essential	ADJ
easat-10684	403	20	layers	layer	NOUN
easat-10684	403	21	,	,	PUNCT
easat-10684	403	22	using	use	VERB
easat-10684	403	23	the	the	DET
easat-10684	403	24	tensorflow	tensorflow	NOUN
easat-10684	403	25	keras	keras	PROPN
easat-10684	403	26	api	api	PROPN
easat-10684	403	27	.	.	PUNCT
easat-10684	404	1	additional	additional	ADJ
easat-10684	404	2	layers	layer	NOUN
easat-10684	404	3	are	be	AUX
easat-10684	404	4	added	add	VERB
easat-10684	404	5	to	to	ADP
easat-10684	404	6	the	the	DET
easat-10684	404	7	bi	bi	ADJ
easat-10684	404	8	-	-	ADJ
easat-10684	404	9	lstm	lstm	ADJ
easat-10684	404	10	model	model	NOUN
easat-10684	404	11	instance	instance	NOUN
easat-10684	404	12	using	use	VERB
easat-10684	404	13	the	the	DET
easat-10684	404	14	model.add	model.add	NOUN
easat-10684	404	15	function	function	NOUN
easat-10684	404	16	;	;	PUNCT
easat-10684	404	17	the	the	DET
easat-10684	404	18	layers	layer	NOUN
easat-10684	404	19	have	have	AUX
easat-10684	404	20	been	be	AUX
easat-10684	404	21	added	add	VERB
easat-10684	404	22	,	,	PUNCT
easat-10684	404	23	including	include	VERB
easat-10684	404	24	one	one	NUM
easat-10684	404	25	(	(	PUNCT
easat-10684	404	26	1	1	NUM
easat-10684	404	27	)	)	PUNCT
easat-10684	404	28	embedding	embed	VERB
easat-10684	404	29	layer	layer	NOUN
easat-10684	404	30	,	,	PUNCT
easat-10684	404	31	two	two	NUM
easat-10684	404	32	(	(	PUNCT
easat-10684	404	33	2	2	NUM
easat-10684	404	34	)	)	PUNCT
easat-10684	404	35	bidirectional	bidirectional	ADJ
easat-10684	404	36	lstm	lstm	ADJ
easat-10684	404	37	layers	layer	NOUN
easat-10684	404	38	,	,	PUNCT
easat-10684	404	39	one	one	NUM
easat-10684	404	40	(	(	PUNCT
easat-10684	404	41	1	1	NUM
easat-10684	404	42	)	)	PUNCT
easat-10684	404	43	lstm	lstm	ADJ
easat-10684	404	44	layer	layer	NOUN
easat-10684	404	45	,	,	PUNCT
easat-10684	404	46	one	one	NUM
easat-10684	404	47	(	(	PUNCT
easat-10684	404	48	1	1	NUM
easat-10684	404	49	)	)	PUNCT
easat-10684	404	50	fully	fully	ADV
easat-10684	404	51	connected	connect	VERB
easat-10684	404	52	dense	dense	ADJ
easat-10684	404	53	layer	layer	NOUN
easat-10684	404	54	,	,	PUNCT
easat-10684	404	55	and	and	CCONJ
easat-10684	404	56	lastly	lastly	ADV
easat-10684	404	57	,	,	PUNCT
easat-10684	404	58	one	one	NUM
easat-10684	404	59	(	(	PUNCT
easat-10684	404	60	1	1	NUM
easat-10684	404	61	)	)	PUNCT
easat-10684	404	62	output	output	NOUN
easat-10684	404	63	layer	layer	NOUN
easat-10684	404	64	.	.	PUNCT
easat-10684	405	1	the	the	DET
easat-10684	405	2	embedding	embed	VERB
easat-10684	405	3	layer	layer	NOUN
easat-10684	405	4	contains	contain	VERB
easat-10684	405	5	parameters	parameter	NOUN
easat-10684	405	6	such	such	ADJ
easat-10684	405	7	as	as	ADP
easat-10684	405	8	the	the	DET
easat-10684	405	9	vocabulary	vocabulary	ADJ
easat-10684	405	10	size	size	NOUN
easat-10684	405	11	(	(	PUNCT
easat-10684	405	12	vocabulary_size	vocabulary_size	PROPN
easat-10684	405	13	)	)	PUNCT
easat-10684	405	14	,	,	PUNCT
easat-10684	405	15	output	output	NOUN
easat-10684	405	16	dimension	dimension	NOUN
easat-10684	405	17	(	(	PUNCT
easat-10684	405	18	output_dim	output_dim	PROPN
easat-10684	405	19	)	)	PUNCT
easat-10684	405	20	,	,	PUNCT
easat-10684	405	21	sequence	sequence	NOUN
easat-10684	405	22	length	length	NOUN
easat-10684	405	23	(	(	PUNCT
easat-10684	405	24	input_length	input_length	NOUN
easat-10684	405	25	)	)	PUNCT
easat-10684	405	26	,	,	PUNCT
easat-10684	405	27	feature	feature	NOUN
easat-10684	405	28	matrix	matrix	NOUN
easat-10684	405	29	(	(	PUNCT
easat-10684	405	30	weights	weight	NOUN
easat-10684	405	31	)	)	PUNCT
easat-10684	405	32	,	,	PUNCT
easat-10684	405	33	and	and	CCONJ
easat-10684	405	34	finally	finally	ADV
easat-10684	405	35	,	,	PUNCT
easat-10684	405	36	the	the	DET
easat-10684	405	37	trainable	trainable	ADJ
easat-10684	405	38	parameter	parameter	NOUN
easat-10684	405	39	has	have	AUX
easat-10684	405	40	been	be	AUX
easat-10684	405	41	set	set	VERB
easat-10684	405	42	to	to	AUX
easat-10684	405	43	false	false	ADJ
easat-10684	405	44	(	(	PUNCT
easat-10684	405	45	trainable	trainable	ADJ
easat-10684	405	46	=	=	ADJ
easat-10684	405	47	false	false	ADJ
easat-10684	405	48	)	)	PUNCT
easat-10684	405	49	.	.	PUNCT
easat-10684	406	1	the	the	DET
easat-10684	406	2	input	input	NOUN
easat-10684	406	3	sequence	sequence	NOUN
easat-10684	406	4	(	(	PUNCT
easat-10684	406	5	embedding	embed	VERB
easat-10684	406	6	size	size	NOUN
easat-10684	406	7	)	)	PUNCT
easat-10684	406	8	is	be	AUX
easat-10684	406	9	set	set	VERB
easat-10684	406	10	to	to	ADP
easat-10684	406	11	500	500	NUM
easat-10684	406	12	features	feature	NOUN
easat-10684	406	13	per	per	ADP
easat-10684	406	14	sequence	sequence	NOUN
easat-10684	406	15	,	,	PUNCT
easat-10684	406	16	and	and	CCONJ
easat-10684	406	17	the	the	DET
easat-10684	406	18	two	two	NUM
easat-10684	406	19	layers	layer	NOUN
easat-10684	406	20	of	of	ADP
easat-10684	406	21	bidirectional	bidirectional	ADJ
easat-10684	406	22	lstm	lstm	NOUN
easat-10684	406	23	contain	contain	VERB
easat-10684	406	24	128	128	NUM
easat-10684	406	25	neurons	neuron	NOUN
easat-10684	406	26	with	with	ADP
easat-10684	406	27	a	a	DET
easat-10684	406	28	20	20	NUM
easat-10684	406	29	%	%	NOUN
easat-10684	406	30	dropout	dropout	NOUN
easat-10684	406	31	rate	rate	NOUN
easat-10684	406	32	.	.	PUNCT
easat-10684	407	1	the	the	DET
easat-10684	407	2	lstm	lstm	NOUN
easat-10684	407	3	layer	layer	NOUN
easat-10684	407	4	contains	contain	VERB
easat-10684	407	5	only	only	ADV
easat-10684	407	6	64	64	NUM
easat-10684	407	7	neurons	neuron	NOUN
easat-10684	407	8	with	with	ADP
easat-10684	407	9	a	a	DET
easat-10684	407	10	10	10	NUM
easat-10684	407	11	%	%	NOUN
easat-10684	407	12	dropout	dropout	NOUN
easat-10684	407	13	rate	rate	NOUN
easat-10684	407	14	for	for	ADP
easat-10684	407	15	each	each	DET
easat-10684	407	16	training	training	NOUN
easat-10684	407	17	iteration	iteration	NOUN
easat-10684	407	18	,	,	PUNCT
easat-10684	407	19	while	while	SCONJ
easat-10684	407	20	the	the	DET
easat-10684	407	21	dense	dense	ADJ
easat-10684	407	22	layer	layer	NOUN
easat-10684	407	23	contains	contain	VERB
easat-10684	407	24	32	32	NUM
easat-10684	407	25	neurons	neuron	NOUN
easat-10684	407	26	and	and	CCONJ
easat-10684	407	27	a	a	DET
easat-10684	407	28	relu	relu	NOUN
easat-10684	407	29	activation	activation	NOUN
easat-10684	407	30	function	function	VERB
easat-10684	407	31	to	to	PART
easat-10684	407	32	introduce	introduce	VERB
easat-10684	407	33	nonlinearity	nonlinearity	NOUN
easat-10684	407	34	.	.	PUNCT
easat-10684	408	1	the	the	DET
easat-10684	408	2	final	final	ADJ
easat-10684	408	3	layer	layer	NOUN
easat-10684	408	4	contains	contain	VERB
easat-10684	408	5	a	a	DET
easat-10684	408	6	single	single	ADJ
easat-10684	408	7	neuron	neuron	NOUN
easat-10684	408	8	outputting	output	VERB
easat-10684	408	9	the	the	DET
easat-10684	408	10	value	value	NOUN
easat-10684	408	11	of	of	ADP
easat-10684	408	12	either	either	CCONJ
easat-10684	408	13	‘	'	PUNCT
easat-10684	408	14	0	0	NUM
easat-10684	408	15	’	'	PUNCT
easat-10684	408	16	or	or	CCONJ
easat-10684	408	17	‘	'	PUNCT
easat-10684	408	18	1	1	NUM
easat-10684	408	19	’	'	PUNCT
easat-10684	408	20	with	with	ADP
easat-10684	408	21	a	a	DET
easat-10684	408	22	sigmoid	sigmoid	NOUN
easat-10684	408	23	activation	activation	NOUN
easat-10684	408	24	function	function	NOUN
easat-10684	408	25	,	,	PUNCT
easat-10684	408	26	which	which	PRON
easat-10684	408	27	is	be	AUX
easat-10684	408	28	used	use	VERB
easat-10684	408	29	for	for	ADP
easat-10684	408	30	binary	binary	ADJ
easat-10684	408	31	classification	classification	NOUN
easat-10684	408	32	tasks	task	NOUN
easat-10684	408	33	.	.	PUNCT
easat-10684	409	1	furthermore	furthermore	ADV
easat-10684	409	2	,	,	PUNCT
easat-10684	409	3	the	the	DET
easat-10684	409	4	bi	bi	ADJ
easat-10684	409	5	-	-	ADJ
easat-10684	409	6	lstm	lstm	ADJ
easat-10684	409	7	model	model	NOUN
easat-10684	409	8	architecture	architecture	NOUN
easat-10684	409	9	is	be	AUX
easat-10684	409	10	compiled	compile	VERB
easat-10684	409	11	and	and	CCONJ
easat-10684	409	12	summarized	summarize	VERB
easat-10684	409	13	before	before	ADP
easat-10684	409	14	training	training	NOUN
easat-10684	409	15	.	.	PUNCT
easat-10684	410	1	the	the	DET
easat-10684	410	2	bilstm	bilstm	NOUN
easat-10684	410	3	architecture	architecture	NOUN
easat-10684	410	4	is	be	AUX
easat-10684	410	5	compiled	compile	VERB
easat-10684	410	6	using	use	VERB
easat-10684	410	7	adam	adam	PROPN
easat-10684	410	8	as	as	ADP
easat-10684	410	9	the	the	DET
easat-10684	410	10	optimization	optimization	NOUN
easat-10684	410	11	algorithm	algorithm	NOUN
easat-10684	410	12	,	,	PUNCT
easat-10684	410	13	binary_crossentropy	binary_crossentropy	VERB
easat-10684	410	14	as	as	SCONJ
easat-10684	410	15	the	the	DET
easat-10684	410	16	loss	loss	NOUN
easat-10684	410	17	1531	1531	NUM
easat-10684	410	18	edelweiss	edelweiss	PROPN
easat-10684	410	19	applied	apply	VERB
easat-10684	410	20	science	science	NOUN
easat-10684	410	21	and	and	CCONJ
easat-10684	410	22	technology	technology	NOUN
easat-10684	410	23	issn	issn	PROPN
easat-10684	410	24	:	:	PUNCT
easat-10684	410	25	2576	2576	NUM
easat-10684	410	26	-	-	SYM
easat-10684	410	27	8484	8484	NUM
easat-10684	410	28	vol	vol	NOUN
easat-10684	410	29	.	.	PROPN
easat-10684	411	1	9	9	NUM
easat-10684	411	2	,	,	PUNCT
easat-10684	411	3	no	no	INTJ
easat-10684	411	4	.	.	NOUN
easat-10684	411	5	10	10	NUM
easat-10684	411	6	:	:	PUNCT
easat-10684	411	7	1515	1515	NUM
easat-10684	411	8	-	-	SYM
easat-10684	411	9	1540	1540	NUM
easat-10684	411	10	,	,	PUNCT
easat-10684	411	11	2025	2025	NUM
easat-10684	411	12	doi	doi	NOUN
easat-10684	411	13	:	:	PUNCT
easat-10684	411	14	10.55214/2576	10.55214/2576	NUM
easat-10684	411	15	-	-	SYM
easat-10684	411	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	412	1	©	©	PROPN
easat-10684	412	2	2025	2025	NUM
easat-10684	412	3	by	by	ADP
easat-10684	412	4	the	the	DET
easat-10684	412	5	authors	author	NOUN
easat-10684	412	6	;	;	PUNCT
easat-10684	412	7	licensee	licensee	PROPN
easat-10684	412	8	learning	learn	VERB
easat-10684	412	9	gate	gate	NOUN
easat-10684	412	10	function	function	NOUN
easat-10684	412	11	,	,	PUNCT
easat-10684	412	12	and	and	CCONJ
easat-10684	412	13	accuracy	accuracy	NOUN
easat-10684	412	14	as	as	SCONJ
easat-10684	412	15	the	the	DET
easat-10684	412	16	performance	performance	NOUN
easat-10684	412	17	metric	metric	ADJ
easat-10684	412	18	.	.	PUNCT
easat-10684	413	1	the	the	DET
easat-10684	413	2	training	training	NOUN
easat-10684	413	3	of	of	ADP
easat-10684	413	4	the	the	DET
easat-10684	413	5	bi	bi	ADJ
easat-10684	413	6	-	-	ADJ
easat-10684	413	7	lstm	lstm	ADJ
easat-10684	413	8	model	model	NOUN
easat-10684	413	9	is	be	AUX
easat-10684	413	10	initiated	initiate	VERB
easat-10684	413	11	using	use	VERB
easat-10684	413	12	the	the	DET
easat-10684	413	13	fit	fit	ADJ
easat-10684	413	14	function	function	NOUN
easat-10684	413	15	of	of	ADP
easat-10684	413	16	the	the	DET
easat-10684	413	17	bi	bi	ADJ
easat-10684	413	18	-	-	ADJ
easat-10684	413	19	lstm	lstm	ADJ
easat-10684	413	20	model	model	NOUN
easat-10684	413	21	object	object	NOUN
easat-10684	413	22	.	.	PUNCT
easat-10684	414	1	the	the	DET
easat-10684	414	2	fit	fit	ADJ
easat-10684	414	3	function	function	NOUN
easat-10684	414	4	uses	use	VERB
easat-10684	414	5	parameters	parameter	NOUN
easat-10684	414	6	that	that	PRON
easat-10684	414	7	define	define	VERB
easat-10684	414	8	how	how	SCONJ
easat-10684	414	9	the	the	DET
easat-10684	414	10	training	training	NOUN
easat-10684	414	11	process	process	NOUN
easat-10684	414	12	is	be	AUX
easat-10684	414	13	conducted	conduct	VERB
easat-10684	414	14	;	;	PUNCT
easat-10684	414	15	this	this	PRON
easat-10684	414	16	includes	include	VERB
easat-10684	414	17	x_train	x_train	X
easat-10684	415	1	(	(	PUNCT
easat-10684	415	2	predictor	predictor	NOUN
easat-10684	415	3	features	feature	VERB
easat-10684	415	4	)	)	PUNCT
easat-10684	415	5	,	,	PUNCT
easat-10684	415	6	y_train	y_train	X
easat-10684	415	7	(	(	PUNCT
easat-10684	415	8	response	response	NOUN
easat-10684	415	9	features	feature	NOUN
easat-10684	415	10	)	)	PUNCT
easat-10684	415	11	,	,	PUNCT
easat-10684	415	12	validation_split	validation_split	NOUN
easat-10684	415	13	of	of	ADP
easat-10684	415	14	20	20	NUM
easat-10684	415	15	%	%	NOUN
easat-10684	415	16	(	(	PUNCT
easat-10684	415	17	data	datum	NOUN
easat-10684	415	18	sample	sample	NOUN
easat-10684	415	19	for	for	ADP
easat-10684	415	20	training	train	VERB
easat-10684	415	21	validation	validation	NOUN
easat-10684	415	22	)	)	PUNCT
easat-10684	415	23	,	,	PUNCT
easat-10684	415	24	and	and	CCONJ
easat-10684	415	25	an	an	DET
easat-10684	415	26	epoch	epoch	NOUN
easat-10684	415	27	set	set	NOUN
easat-10684	415	28	to	to	ADP
easat-10684	415	29	12	12	NUM
easat-10684	415	30	.	.	PUNCT
easat-10684	416	1	the	the	DET
easat-10684	416	2	training	training	NOUN
easat-10684	416	3	process	process	NOUN
easat-10684	416	4	summarizes	summarize	VERB
easat-10684	416	5	training	training	NOUN
easat-10684	416	6	loss	loss	NOUN
easat-10684	416	7	,	,	PUNCT
easat-10684	416	8	accuracy	accuracy	NOUN
easat-10684	416	9	,	,	PUNCT
easat-10684	416	10	val_loss	val_loss	ADJ
easat-10684	416	11	,	,	PUNCT
easat-10684	416	12	and	and	CCONJ
easat-10684	416	13	val_accuracy	val_accuracy	NOUN
easat-10684	416	14	for	for	ADP
easat-10684	416	15	each	each	DET
easat-10684	416	16	iteration	iteration	NOUN
easat-10684	416	17	.	.	PUNCT
easat-10684	417	1	finally	finally	ADV
easat-10684	417	2	,	,	PUNCT
easat-10684	417	3	the	the	DET
easat-10684	417	4	last	last	ADJ
easat-10684	417	5	accuracy	accuracy	NOUN
easat-10684	417	6	and	and	CCONJ
easat-10684	417	7	loss	loss	NOUN
easat-10684	417	8	on	on	ADP
easat-10684	417	9	the	the	DET
easat-10684	417	10	12th	12th	ADJ
easat-10684	417	11	iteration	iteration	NOUN
easat-10684	417	12	represent	represent	VERB
easat-10684	417	13	the	the	DET
easat-10684	417	14	highest	high	ADJ
easat-10684	417	15	accuracy	accuracy	NOUN
easat-10684	417	16	and	and	CCONJ
easat-10684	417	17	minimum	minimum	NOUN
easat-10684	417	18	loss	loss	NOUN
easat-10684	417	19	achieved	achieve	VERB
easat-10684	417	20	by	by	ADP
easat-10684	417	21	the	the	DET
easat-10684	417	22	model	model	NOUN
easat-10684	417	23	.	.	PUNCT
easat-10684	418	1	4.2	4.2	NUM
easat-10684	418	2	.	.	PUNCT
easat-10684	418	3	machine	machine	NOUN
easat-10684	418	4	learning	learn	VERB
easat-10684	418	5	implementation	implementation	NOUN
easat-10684	418	6	this	this	DET
easat-10684	418	7	subsection	subsection	NOUN
easat-10684	418	8	details	detail	VERB
easat-10684	418	9	the	the	DET
easat-10684	418	10	training	training	NOUN
easat-10684	418	11	process	process	NOUN
easat-10684	418	12	of	of	ADP
easat-10684	418	13	various	various	ADJ
easat-10684	418	14	machine	machine	NOUN
easat-10684	418	15	learning	learning	NOUN
easat-10684	418	16	models	model	NOUN
easat-10684	418	17	,	,	PUNCT
easat-10684	418	18	including	include	VERB
easat-10684	418	19	support	support	NOUN
easat-10684	418	20	vector	vector	NOUN
easat-10684	418	21	machine	machine	NOUN
easat-10684	418	22	(	(	PUNCT
easat-10684	418	23	svm	svm	PROPN
easat-10684	418	24	)	)	PUNCT
easat-10684	418	25	,	,	PUNCT
easat-10684	418	26	decision	decision	NOUN
easat-10684	418	27	tree	tree	NOUN
easat-10684	418	28	(	(	PUNCT
easat-10684	418	29	dt	dt	PROPN
easat-10684	418	30	)	)	PUNCT
easat-10684	418	31	,	,	PUNCT
easat-10684	418	32	naïve	naïve	ADJ
easat-10684	418	33	bayes	bayes	PROPN
easat-10684	418	34	(	(	PUNCT
easat-10684	418	35	nb	nb	PROPN
easat-10684	418	36	)	)	PUNCT
easat-10684	418	37	,	,	PUNCT
easat-10684	418	38	logistic	logistic	ADJ
easat-10684	418	39	regression	regression	NOUN
easat-10684	418	40	(	(	PUNCT
easat-10684	418	41	lr	lr	NOUN
easat-10684	418	42	)	)	PUNCT
easat-10684	418	43	,	,	PUNCT
easat-10684	418	44	and	and	CCONJ
easat-10684	418	45	knearest	knearest	PROPN
easat-10684	418	46	neighbour	neighbour	PROPN
easat-10684	418	47	(	(	PUNCT
easat-10684	418	48	knn	knn	PROPN
easat-10684	418	49	)	)	PUNCT
easat-10684	418	50	.	.	PUNCT
easat-10684	419	1	additionally	additionally	ADV
easat-10684	419	2	,	,	PUNCT
easat-10684	419	3	transfer	transfer	VERB
easat-10684	419	4	learning	learning	NOUN
easat-10684	419	5	-	-	PUNCT
easat-10684	419	6	based	base	VERB
easat-10684	419	7	models	model	NOUN
easat-10684	419	8	are	be	AUX
easat-10684	419	9	retrained	retrain	VERB
easat-10684	419	10	using	use	VERB
easat-10684	419	11	bi	bi	NOUN
easat-10684	419	12	-	-	ADJ
easat-10684	419	13	lstm	lstm	ADJ
easat-10684	419	14	pre	pre	ADJ
easat-10684	419	15	-	-	ADJ
easat-10684	419	16	trained	train	VERB
easat-10684	419	17	weights	weight	NOUN
easat-10684	419	18	to	to	PART
easat-10684	419	19	enhance	enhance	VERB
easat-10684	419	20	performance	performance	NOUN
easat-10684	419	21	.	.	PUNCT
easat-10684	420	1	to	to	PART
easat-10684	420	2	train	train	VERB
easat-10684	420	3	these	these	DET
easat-10684	420	4	machine	machine	NOUN
easat-10684	420	5	learning	learning	NOUN
easat-10684	420	6	models	model	NOUN
easat-10684	420	7	,	,	PUNCT
easat-10684	420	8	the	the	DET
easat-10684	420	9	fit	fit	ADJ
easat-10684	420	10	function	function	NOUN
easat-10684	420	11	is	be	AUX
easat-10684	420	12	called	call	VERB
easat-10684	420	13	,	,	PUNCT
easat-10684	420	14	passing	pass	VERB
easat-10684	420	15	the	the	DET
easat-10684	420	16	‘	'	PUNCT
easat-10684	420	17	x_train	x_train	NOUN
easat-10684	420	18	’	'	PUNCT
easat-10684	420	19	and	and	CCONJ
easat-10684	420	20	‘	'	PUNCT
easat-10684	420	21	y_train	y_train	X
easat-10684	420	22	’	'	PUNCT
easat-10684	420	23	data	datum	NOUN
easat-10684	420	24	samples	sample	NOUN
easat-10684	420	25	as	as	ADP
easat-10684	420	26	parameters	parameter	NOUN
easat-10684	420	27	.	.	PUNCT
easat-10684	421	1	once	once	ADV
easat-10684	421	2	trained	train	VERB
easat-10684	421	3	,	,	PUNCT
easat-10684	421	4	the	the	DET
easat-10684	421	5	models	model	NOUN
easat-10684	421	6	for	for	ADP
easat-10684	421	7	fake	fake	ADJ
easat-10684	421	8	news	news	NOUN
easat-10684	421	9	detection	detection	NOUN
easat-10684	421	10	can	can	AUX
easat-10684	421	11	be	be	AUX
easat-10684	421	12	evaluated	evaluate	VERB
easat-10684	421	13	and	and	CCONJ
easat-10684	421	14	used	use	VERB
easat-10684	421	15	for	for	ADP
easat-10684	421	16	predictions	prediction	NOUN
easat-10684	421	17	.	.	PUNCT
easat-10684	422	1	5	5	X
easat-10684	422	2	.	.	X
easat-10684	422	3	result	result	NOUN
easat-10684	422	4	and	and	CCONJ
easat-10684	422	5	discussion	discussion	NOUN
easat-10684	422	6	5.1	5.1	NUM
easat-10684	422	7	.	.	PUNCT
easat-10684	423	1	deep	deep	ADJ
easat-10684	423	2	learning	learning	NOUN
easat-10684	423	3	performance	performance	NOUN
easat-10684	423	4	figure	figure	NOUN
easat-10684	423	5	7	7	NUM
easat-10684	423	6	shows	show	VERB
easat-10684	423	7	the	the	DET
easat-10684	423	8	training	training	NOUN
easat-10684	423	9	performance	performance	NOUN
easat-10684	423	10	measure	measure	NOUN
easat-10684	423	11	of	of	ADP
easat-10684	423	12	the	the	DET
easat-10684	423	13	bi	bi	ADJ
easat-10684	423	14	-	-	ADJ
easat-10684	423	15	lstm	lstm	ADJ
easat-10684	423	16	fake	fake	ADJ
easat-10684	423	17	news	news	NOUN
easat-10684	423	18	detection	detection	NOUN
easat-10684	423	19	model	model	NOUN
easat-10684	423	20	using	use	VERB
easat-10684	423	21	line	line	NOUN
easat-10684	423	22	chart	chart	NOUN
easat-10684	423	23	and	and	CCONJ
easat-10684	423	24	a	a	DET
easat-10684	423	25	confusion	confusion	NOUN
easat-10684	423	26	matrix	matrix	NOUN
easat-10684	423	27	.	.	PUNCT
easat-10684	424	1	the	the	DET
easat-10684	424	2	line	line	NOUN
easat-10684	424	3	chart	chart	NOUN
easat-10684	424	4	illustrates	illustrate	VERB
easat-10684	424	5	the	the	DET
easat-10684	424	6	change	change	NOUN
easat-10684	424	7	in	in	ADP
easat-10684	424	8	performance	performance	NOUN
easat-10684	424	9	based	base	VERB
easat-10684	424	10	on	on	ADP
easat-10684	424	11	accuracy	accuracy	NOUN
easat-10684	424	12	and	and	CCONJ
easat-10684	424	13	loss	loss	NOUN
easat-10684	424	14	for	for	ADP
easat-10684	424	15	every	every	DET
easat-10684	424	16	epoch	epoch	NOUN
easat-10684	424	17	.	.	PUNCT
easat-10684	425	1	the	the	DET
easat-10684	425	2	plot	plot	NOUN
easat-10684	425	3	shows	show	VERB
easat-10684	425	4	a	a	DET
easat-10684	425	5	smooth	smooth	ADJ
easat-10684	425	6	trend	trend	NOUN
easat-10684	425	7	in	in	ADP
easat-10684	425	8	accuracy	accuracy	NOUN
easat-10684	425	9	increase	increase	NOUN
easat-10684	425	10	for	for	ADP
easat-10684	425	11	both	both	PRON
easat-10684	425	12	training	training	NOUN
easat-10684	425	13	and	and	CCONJ
easat-10684	425	14	validation	validation	NOUN
easat-10684	425	15	samples	sample	NOUN
easat-10684	425	16	,	,	PUNCT
easat-10684	425	17	but	but	CCONJ
easat-10684	425	18	initially	initially	ADV
easat-10684	425	19	,	,	PUNCT
easat-10684	425	20	the	the	DET
easat-10684	425	21	training	training	NOUN
easat-10684	425	22	accuracy	accuracy	NOUN
easat-10684	425	23	was	be	AUX
easat-10684	425	24	low	low	ADJ
easat-10684	425	25	for	for	ADP
easat-10684	425	26	the	the	DET
easat-10684	425	27	first	first	ADJ
easat-10684	425	28	epoch	epoch	NOUN
easat-10684	425	29	,	,	PUNCT
easat-10684	425	30	and	and	CCONJ
easat-10684	425	31	later	later	ADV
easat-10684	425	32	performed	perform	VERB
easat-10684	425	33	efficiently	efficiently	ADV
easat-10684	425	34	on	on	ADP
easat-10684	425	35	the	the	DET
easat-10684	425	36	training	training	NOUN
easat-10684	425	37	sample	sample	NOUN
easat-10684	425	38	.	.	PUNCT
easat-10684	426	1	for	for	ADP
easat-10684	426	2	every	every	DET
easat-10684	426	3	epoch	epoch	NOUN
easat-10684	426	4	,	,	PUNCT
easat-10684	426	5	the	the	DET
easat-10684	426	6	model	model	NOUN
easat-10684	426	7	significantly	significantly	ADV
easat-10684	426	8	drops	drop	VERB
easat-10684	426	9	in	in	ADP
easat-10684	426	10	loss	loss	NOUN
easat-10684	426	11	,	,	PUNCT
easat-10684	426	12	which	which	PRON
easat-10684	426	13	shows	show	VERB
easat-10684	426	14	a	a	DET
easat-10684	426	15	smooth	smooth	ADJ
easat-10684	426	16	learning	learning	NOUN
easat-10684	426	17	curve	curve	NOUN
easat-10684	426	18	.	.	PUNCT
easat-10684	427	1	1532	1532	NUM
easat-10684	427	2	edelweiss	edelweiss	PROPN
easat-10684	427	3	applied	apply	VERB
easat-10684	427	4	science	science	NOUN
easat-10684	427	5	and	and	CCONJ
easat-10684	427	6	technology	technology	NOUN
easat-10684	427	7	issn	issn	PROPN
easat-10684	427	8	:	:	PUNCT
easat-10684	427	9	2576	2576	NUM
easat-10684	427	10	-	-	SYM
easat-10684	427	11	8484	8484	NUM
easat-10684	427	12	vol	vol	NOUN
easat-10684	427	13	.	.	PROPN
easat-10684	428	1	9	9	NUM
easat-10684	428	2	,	,	PUNCT
easat-10684	428	3	no	no	INTJ
easat-10684	428	4	.	.	NOUN
easat-10684	428	5	10	10	NUM
easat-10684	428	6	:	:	PUNCT
easat-10684	428	7	1515	1515	NUM
easat-10684	428	8	-	-	SYM
easat-10684	428	9	1540	1540	NUM
easat-10684	428	10	,	,	PUNCT
easat-10684	428	11	2025	2025	NUM
easat-10684	428	12	doi	doi	NOUN
easat-10684	428	13	:	:	PUNCT
easat-10684	428	14	10.55214/2576	10.55214/2576	NUM
easat-10684	428	15	-	-	SYM
easat-10684	428	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	429	1	©	©	PROPN
easat-10684	429	2	2025	2025	NUM
easat-10684	429	3	by	by	ADP
easat-10684	429	4	the	the	DET
easat-10684	429	5	authors	author	NOUN
easat-10684	429	6	;	;	PUNCT
easat-10684	429	7	licensee	licensee	PROPN
easat-10684	429	8	learning	learn	VERB
easat-10684	429	9	gate	gate	NOUN
easat-10684	429	10	figure	figure	NOUN
easat-10684	429	11	7	7	NUM
easat-10684	429	12	.	.	PUNCT
easat-10684	429	13	bi	bi	ADJ
easat-10684	429	14	-	-	ADJ
easat-10684	429	15	lstm	lstm	ADJ
easat-10684	429	16	accuracy	accuracy	NOUN
easat-10684	429	17	&	&	CCONJ
easat-10684	429	18	loss	loss	NOUN
easat-10684	429	19	.	.	PUNCT
easat-10684	430	1	figure	figure	NOUN
easat-10684	430	2	8	8	NUM
easat-10684	430	3	shows	show	VERB
easat-10684	430	4	the	the	DET
easat-10684	430	5	confusion	confusion	NOUN
easat-10684	430	6	matrix	matrix	NOUN
easat-10684	430	7	illustrating	illustrate	VERB
easat-10684	430	8	the	the	DET
easat-10684	430	9	interception	interception	NOUN
easat-10684	430	10	of	of	ADP
easat-10684	430	11	the	the	DET
easat-10684	430	12	actual	actual	ADJ
easat-10684	430	13	output	output	NOUN
easat-10684	430	14	to	to	PART
easat-10684	430	15	be	be	AUX
easat-10684	430	16	predicted	predict	VERB
easat-10684	430	17	(	(	PUNCT
easat-10684	430	18	actual	actual	ADJ
easat-10684	430	19	value	value	NOUN
easat-10684	430	20	)	)	PUNCT
easat-10684	430	21	and	and	CCONJ
easat-10684	430	22	the	the	DET
easat-10684	430	23	predicted	predict	VERB
easat-10684	430	24	value	value	NOUN
easat-10684	430	25	(	(	PUNCT
easat-10684	430	26	prediction	prediction	NOUN
easat-10684	430	27	value	value	NOUN
easat-10684	430	28	)	)	PUNCT
easat-10684	430	29	.	.	PUNCT
easat-10684	431	1	the	the	DET
easat-10684	431	2	confusion	confusion	NOUN
easat-10684	431	3	matrix	matrix	NOUN
easat-10684	431	4	indicates	indicate	VERB
easat-10684	431	5	that	that	SCONJ
easat-10684	431	6	real	real	ADJ
easat-10684	431	7	news	news	NOUN
easat-10684	431	8	was	be	AUX
easat-10684	431	9	misclassified	misclassifie	VERB
easat-10684	431	10	44	44	NUM
easat-10684	431	11	times	time	NOUN
easat-10684	431	12	as	as	ADP
easat-10684	431	13	fake	fake	ADJ
easat-10684	431	14	news	news	NOUN
easat-10684	431	15	,	,	PUNCT
easat-10684	431	16	and	and	CCONJ
easat-10684	431	17	fake	fake	ADJ
easat-10684	431	18	news	news	NOUN
easat-10684	431	19	was	be	AUX
easat-10684	431	20	misclassified	misclassifie	VERB
easat-10684	431	21	46	46	NUM
easat-10684	431	22	times	time	NOUN
easat-10684	431	23	as	as	ADP
easat-10684	431	24	real	real	ADJ
easat-10684	431	25	news	news	NOUN
easat-10684	431	26	.	.	PUNCT
easat-10684	432	1	1533	1533	NUM
easat-10684	432	2	edelweiss	edelweiss	PROPN
easat-10684	432	3	applied	apply	VERB
easat-10684	432	4	science	science	NOUN
easat-10684	432	5	and	and	CCONJ
easat-10684	432	6	technology	technology	NOUN
easat-10684	432	7	issn	issn	PROPN
easat-10684	432	8	:	:	PUNCT
easat-10684	432	9	2576	2576	NUM
easat-10684	432	10	-	-	SYM
easat-10684	432	11	8484	8484	NUM
easat-10684	432	12	vol	vol	NOUN
easat-10684	432	13	.	.	PROPN
easat-10684	433	1	9	9	NUM
easat-10684	433	2	,	,	PUNCT
easat-10684	433	3	no	no	INTJ
easat-10684	433	4	.	.	NOUN
easat-10684	433	5	10	10	NUM
easat-10684	433	6	:	:	PUNCT
easat-10684	433	7	1515	1515	NUM
easat-10684	433	8	-	-	SYM
easat-10684	433	9	1540	1540	NUM
easat-10684	433	10	,	,	PUNCT
easat-10684	433	11	2025	2025	NUM
easat-10684	433	12	doi	doi	NOUN
easat-10684	433	13	:	:	PUNCT
easat-10684	433	14	10.55214/2576	10.55214/2576	NUM
easat-10684	433	15	-	-	SYM
easat-10684	433	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	434	1	©	©	PROPN
easat-10684	434	2	2025	2025	NUM
easat-10684	434	3	by	by	ADP
easat-10684	434	4	the	the	DET
easat-10684	434	5	authors	author	NOUN
easat-10684	434	6	;	;	PUNCT
easat-10684	434	7	licensee	licensee	PROPN
easat-10684	434	8	learning	learn	VERB
easat-10684	434	9	gate	gate	PROPN
easat-10684	434	10	figure	figure	NOUN
easat-10684	434	11	8	8	NUM
easat-10684	434	12	.	.	PUNCT
easat-10684	435	1	bi	bi	ADJ
easat-10684	435	2	-	-	ADJ
easat-10684	435	3	lstm	lstm	ADJ
easat-10684	435	4	confusion	confusion	NOUN
easat-10684	435	5	matrix	matrix	NOUN
easat-10684	435	6	plot	plot	NOUN
easat-10684	435	7	.	.	PUNCT
easat-10684	436	1	figure	figure	NOUN
easat-10684	436	2	9	9	NUM
easat-10684	436	3	shows	show	VERB
easat-10684	436	4	the	the	DET
easat-10684	436	5	bi	bi	ADJ
easat-10684	436	6	-	-	ADJ
easat-10684	436	7	lstm	lstm	ADJ
easat-10684	436	8	model	model	NOUN
easat-10684	436	9	's	's	PART
easat-10684	436	10	performance	performance	NOUN
easat-10684	436	11	metrics	metric	NOUN
easat-10684	436	12	,	,	PUNCT
easat-10684	436	13	which	which	PRON
easat-10684	436	14	achieved	achieve	VERB
easat-10684	436	15	an	an	DET
easat-10684	436	16	accuracy	accuracy	NOUN
easat-10684	436	17	of	of	ADP
easat-10684	436	18	92.16	92.16	NUM
easat-10684	436	19	%	%	NOUN
easat-10684	436	20	,	,	PUNCT
easat-10684	436	21	a	a	DET
easat-10684	436	22	precision	precision	NOUN
easat-10684	436	23	of	of	ADP
easat-10684	436	24	92.18	92.18	NUM
easat-10684	436	25	%	%	NOUN
easat-10684	436	26	,	,	PUNCT
easat-10684	436	27	a	a	DET
easat-10684	436	28	recall	recall	NOUN
easat-10684	436	29	of	of	ADP
easat-10684	436	30	91.85	91.85	NUM
easat-10684	436	31	%	%	NOUN
easat-10684	436	32	,	,	PUNCT
easat-10684	436	33	and	and	CCONJ
easat-10684	436	34	an	an	DET
easat-10684	436	35	f1	f1	NOUN
easat-10684	436	36	-	-	PUNCT
easat-10684	436	37	score	score	NOUN
easat-10684	436	38	of	of	ADP
easat-10684	436	39	92.02	92.02	NUM
easat-10684	436	40	%	%	NOUN
easat-10684	436	41	.	.	PUNCT
easat-10684	437	1	figure	figure	NOUN
easat-10684	437	2	9	9	NUM
easat-10684	437	3	.	.	PUNCT
easat-10684	438	1	bi	bi	ADJ
easat-10684	438	2	-	-	ADJ
easat-10684	438	3	lstm	lstm	ADJ
easat-10684	438	4	performance	performance	NOUN
easat-10684	438	5	summary	summary	NOUN
easat-10684	438	6	.	.	PUNCT
easat-10684	439	1	5.2	5.2	NUM
easat-10684	439	2	.	.	PUNCT
easat-10684	440	1	machinese	machinese	PROPN
easat-10684	440	2	learning	learn	VERB
easat-10684	440	3	performance	performance	NOUN
easat-10684	440	4	(	(	PUNCT
easat-10684	440	5	bi	bi	NOUN
easat-10684	440	6	-	-	ADJ
easat-10684	440	7	lsmt-2	lsmt-2	PROPN
easat-10684	440	8	-	-	PUNCT
easat-10684	440	9	ml	ml	NOUN
easat-10684	440	10	)	)	PUNCT
easat-10684	440	11	this	this	DET
easat-10684	440	12	subsection	subsection	NOUN
easat-10684	440	13	comprehensively	comprehensively	ADV
easat-10684	440	14	discusses	discuss	VERB
easat-10684	440	15	the	the	DET
easat-10684	440	16	result	result	NOUN
easat-10684	440	17	summary	summary	NOUN
easat-10684	440	18	of	of	ADP
easat-10684	440	19	the	the	DET
easat-10684	440	20	machine	machine	NOUN
easat-10684	440	21	learning	learn	VERB
easat-10684	440	22	training	training	NOUN
easat-10684	440	23	process	process	NOUN
easat-10684	440	24	using	use	VERB
easat-10684	440	25	the	the	DET
easat-10684	440	26	bi	bi	ADJ
easat-10684	440	27	-	-	ADJ
easat-10684	440	28	lstm	lstm	ADJ
easat-10684	440	29	trained	train	VERB
easat-10684	440	30	weights	weight	NOUN
easat-10684	440	31	.	.	PUNCT
easat-10684	441	1	table	table	NOUN
easat-10684	441	2	2	2	NUM
easat-10684	441	3	shows	show	VERB
easat-10684	441	4	the	the	DET
easat-10684	441	5	accuracy	accuracy	NOUN
easat-10684	441	6	,	,	PUNCT
easat-10684	441	7	precision	precision	NOUN
easat-10684	441	8	,	,	PUNCT
easat-10684	441	9	recall	recall	NOUN
easat-10684	441	10	,	,	PUNCT
easat-10684	441	11	and	and	CCONJ
easat-10684	441	12	f1	f1	NOUN
easat-10684	441	13	-	-	PUNCT
easat-10684	441	14	score	score	NOUN
easat-10684	441	15	for	for	ADP
easat-10684	441	16	each	each	DET
easat-10684	441	17	benchmark	benchmark	NOUN
easat-10684	441	18	machine	machine	NOUN
easat-10684	441	19	learning	learn	VERB
easat-10684	441	20	algorithm	algorithm	NOUN
easat-10684	441	21	trained	train	VERB
easat-10684	441	22	without	without	ADP
easat-10684	441	23	the	the	DET
easat-10684	441	24	bi	bi	NOUN
easat-10684	441	25	-	-	ADJ
easat-10684	441	26	lstm	lstm	ADJ
easat-10684	441	27	pre	pre	ADJ
easat-10684	441	28	-	-	ADJ
easat-10684	441	29	trained	train	VERB
easat-10684	441	30	weight	weight	NOUN
easat-10684	441	31	.	.	PUNCT
easat-10684	442	1	based	base	VERB
easat-10684	442	2	on	on	ADP
easat-10684	442	3	table	table	NOUN
easat-10684	442	4	2	2	NUM
easat-10684	442	5	,	,	PUNCT
easat-10684	442	6	the	the	DET
easat-10684	442	7	support	support	NOUN
easat-10684	442	8	vector	vector	NOUN
easat-10684	442	9	machine	machine	NOUN
easat-10684	442	10	(	(	PUNCT
easat-10684	442	11	svm	svm	PROPN
easat-10684	442	12	)	)	PUNCT
easat-10684	442	13	achieved	achieve	VERB
easat-10684	442	14	88.1	88.1	NUM
easat-10684	442	15	%	%	NOUN
easat-10684	442	16	accuracy	accuracy	NOUN
easat-10684	442	17	,	,	PUNCT
easat-10684	442	18	which	which	PRON
easat-10684	442	19	is	be	AUX
easat-10684	442	20	the	the	DET
easat-10684	442	21	best	good	ADJ
easat-10684	442	22	accuracy	accuracy	NOUN
easat-10684	442	23	observed	observe	VERB
easat-10684	442	24	.	.	PUNCT
easat-10684	443	1	logistic	logistic	ADJ
easat-10684	443	2	regression	regression	NOUN
easat-10684	443	3	follows	follow	VERB
easat-10684	443	4	the	the	DET
easat-10684	443	5	svm	svm	ADJ
easat-10684	443	6	performance	performance	NOUN
easat-10684	443	7	with	with	ADP
easat-10684	443	8	87.5	87.5	NUM
easat-10684	443	9	%	%	NOUN
easat-10684	443	10	accuracy	accuracy	NOUN
easat-10684	443	11	,	,	PUNCT
easat-10684	443	12	followed	follow	VERB
easat-10684	443	13	by	by	ADP
easat-10684	443	14	the	the	DET
easat-10684	443	15	k	k	PROPN
easat-10684	443	16	-	-	PUNCT
easat-10684	443	17	nearest	near	ADJ
easat-10684	443	18	neighbour	neighbour	NOUN
easat-10684	443	19	with	with	ADP
easat-10684	443	20	85.8	85.8	NUM
easat-10684	443	21	%	%	NOUN
easat-10684	443	22	,	,	PUNCT
easat-10684	443	23	the	the	DET
easat-10684	443	24	decision	decision	NOUN
easat-10684	443	25	tree	tree	NOUN
easat-10684	443	26	with	with	ADP
easat-10684	443	27	81.8	81.8	NUM
easat-10684	443	28	%	%	NOUN
easat-10684	443	29	,	,	PUNCT
easat-10684	443	30	and	and	CCONJ
easat-10684	443	31	finally	finally	ADV
easat-10684	443	32	,	,	PUNCT
easat-10684	443	33	the	the	DET
easat-10684	443	34	naïve	naïve	ADJ
easat-10684	443	35	bayes	bayes	NOUN
easat-10684	443	36	algorithm	algorithm	VERB
easat-10684	443	37	with	with	ADP
easat-10684	443	38	the	the	DET
easat-10684	443	39	lowest	low	ADJ
easat-10684	443	40	performance	performance	NOUN
easat-10684	443	41	of	of	ADP
easat-10684	443	42	79.4	79.4	NUM
easat-10684	443	43	%	%	NOUN
easat-10684	443	44	accuracy	accuracy	NOUN
easat-10684	443	45	.	.	PUNCT
easat-10684	444	1	the	the	DET
easat-10684	444	2	result	result	NOUN
easat-10684	444	3	summary	summary	NOUN
easat-10684	444	4	in	in	ADP
easat-10684	444	5	table	table	NOUN
easat-10684	444	6	2	2	NUM
easat-10684	444	7	serves	serve	VERB
easat-10684	444	8	as	as	ADP
easat-10684	444	9	a	a	DET
easat-10684	444	10	benchmark	benchmark	NOUN
easat-10684	444	11	,	,	PUNCT
easat-10684	444	12	which	which	PRON
easat-10684	444	13	will	will	AUX
easat-10684	444	14	be	be	AUX
easat-10684	444	15	used	use	VERB
easat-10684	444	16	later	later	ADV
easat-10684	444	17	for	for	ADP
easat-10684	444	18	comparing	compare	VERB
easat-10684	444	19	and	and	CCONJ
easat-10684	444	20	validating	validate	VERB
easat-10684	444	21	the	the	DET
easat-10684	444	22	impact	impact	NOUN
easat-10684	444	23	of	of	ADP
easat-10684	444	24	transfer	transfer	NOUN
easat-10684	444	25	learning	learning	NOUN
easat-10684	444	26	on	on	ADP
easat-10684	444	27	the	the	DET
easat-10684	444	28	machine	machine	NOUN
easat-10684	444	29	learning	learn	VERB
easat-10684	444	30	algorithms	algorithm	NOUN
easat-10684	444	31	.	.	PUNCT
easat-10684	445	1	table	table	NOUN
easat-10684	445	2	2	2	NUM
easat-10684	445	3	is	be	AUX
easat-10684	445	4	graphically	graphically	ADV
easat-10684	445	5	represented	represent	VERB
easat-10684	445	6	using	use	VERB
easat-10684	445	7	the	the	DET
easat-10684	445	8	bar	bar	NOUN
easat-10684	445	9	chart	chart	NOUN
easat-10684	445	10	visual	visual	ADJ
easat-10684	445	11	displayed	display	VERB
easat-10684	445	12	in	in	ADP
easat-10684	445	13	figure	figure	NOUN
easat-10684	445	14	10	10	NUM
easat-10684	445	15	.	.	PUNCT
easat-10684	446	1	1534	1534	NUM
easat-10684	446	2	edelweiss	edelweiss	PROPN
easat-10684	446	3	applied	apply	VERB
easat-10684	446	4	science	science	NOUN
easat-10684	446	5	and	and	CCONJ
easat-10684	446	6	technology	technology	NOUN
easat-10684	446	7	issn	issn	PROPN
easat-10684	446	8	:	:	PUNCT
easat-10684	446	9	2576	2576	NUM
easat-10684	446	10	-	-	SYM
easat-10684	446	11	8484	8484	NUM
easat-10684	446	12	vol	vol	NOUN
easat-10684	446	13	.	.	PROPN
easat-10684	447	1	9	9	NUM
easat-10684	447	2	,	,	PUNCT
easat-10684	447	3	no	no	INTJ
easat-10684	447	4	.	.	NOUN
easat-10684	447	5	10	10	NUM
easat-10684	447	6	:	:	PUNCT
easat-10684	447	7	1515	1515	NUM
easat-10684	447	8	-	-	SYM
easat-10684	447	9	1540	1540	NUM
easat-10684	447	10	,	,	PUNCT
easat-10684	447	11	2025	2025	NUM
easat-10684	447	12	doi	doi	NOUN
easat-10684	447	13	:	:	PUNCT
easat-10684	447	14	10.55214/2576	10.55214/2576	NUM
easat-10684	447	15	-	-	SYM
easat-10684	447	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	448	1	©	©	PROPN
easat-10684	448	2	2025	2025	NUM
easat-10684	448	3	by	by	ADP
easat-10684	448	4	the	the	DET
easat-10684	448	5	authors	author	NOUN
easat-10684	448	6	;	;	PUNCT
easat-10684	448	7	licensee	licensee	PROPN
easat-10684	448	8	learning	learn	VERB
easat-10684	448	9	gate	gate	NOUN
easat-10684	448	10	table	table	NOUN
easat-10684	448	11	2	2	NUM
easat-10684	448	12	.	.	PUNCT
easat-10684	448	13	machine	machine	NOUN
easat-10684	448	14	learning	learn	VERB
easat-10684	448	15	performance	performance	NOUN
easat-10684	448	16	(	(	PUNCT
easat-10684	448	17	without	without	ADP
easat-10684	448	18	bi	bi	NOUN
easat-10684	448	19	-	-	NOUN
easat-10684	448	20	lstm	lstm	ADJ
easat-10684	448	21	)	)	PUNCT
easat-10684	448	22	.	.	PUNCT
easat-10684	449	1	s	s	X
easat-10684	449	2	/	/	SYM
easat-10684	449	3	n	n	CCONJ
easat-10684	449	4	model	model	NOUN
easat-10684	449	5	accuracy	accuracy	NOUN
easat-10684	449	6	precision	precision	NOUN
easat-10684	449	7	recall	recall	VERB
easat-10684	449	8	f1	f1	NOUN
easat-10684	449	9	-	-	PUNCT
easat-10684	449	10	score	score	NOUN
easat-10684	449	11	0	0	NUM
easat-10684	449	12	logistic	logistic	ADJ
easat-10684	449	13	regression	regression	NOUN
easat-10684	449	14	0.875272	0.875272	NUM
easat-10684	449	15	0.904412	0.904412	NUM
easat-10684	449	16	0.851801	0.851801	NUM
easat-10684	449	17	0.877318	0.877318	NUM
easat-10684	449	18	1	1	NUM
easat-10684	449	19	k	k	NOUN
easat-10684	449	20	nearest	near	ADJ
easat-10684	449	21	neighbour	neighbour	NOUN
easat-10684	449	22	0.858593	0.858593	NUM
easat-10684	449	23	0.861454	0.861454	NUM
easat-10684	449	24	0.869806	0.869806	NUM
easat-10684	449	25	0.86561	0.86561	NUM
easat-10684	449	26	2	2	NUM
easat-10684	449	27	naive	naive	ADJ
easat-10684	449	28	bayes	bayes	NOUN
easat-10684	449	29	0.794054	0.794054	NUM
easat-10684	449	30	0.834862	0.834862	NUM
easat-10684	449	31	0.756233	0.756233	NUM
easat-10684	449	32	0.793605	0.793605	NUM
easat-10684	449	33	3	3	NUM
easat-10684	449	34	svm	svm	VERB
easat-10684	449	35	0.881073	0.881073	NUM
easat-10684	449	36	0.918919	0.918919	NUM
easat-10684	449	37	0.847645	0.847645	NUM
easat-10684	449	38	0.881844	0.881844	NUM
easat-10684	449	39	4	4	NUM
easat-10684	449	40	decision	decision	NOUN
easat-10684	449	41	tree	tree	NOUN
easat-10684	449	42	0.818709	0.818709	NUM
easat-10684	449	43	0.834278	0.834278	NUM
easat-10684	449	44	0.815789	0.815789	NUM
easat-10684	449	45	0.82493	0.82493	NUM
easat-10684	449	46	the	the	DET
easat-10684	449	47	bar	bar	NOUN
easat-10684	449	48	chart	chart	NOUN
easat-10684	449	49	in	in	ADP
easat-10684	449	50	figure	figure	NOUN
easat-10684	449	51	10	10	NUM
easat-10684	449	52	graphically	graphically	ADV
easat-10684	449	53	represents	represent	VERB
easat-10684	449	54	the	the	DET
easat-10684	449	55	performance	performance	NOUN
easat-10684	449	56	of	of	ADP
easat-10684	449	57	various	various	ADJ
easat-10684	449	58	machine	machine	NOUN
easat-10684	449	59	learning	learn	VERB
easat-10684	449	60	algorithm	algorithm	NOUN
easat-10684	449	61	models	model	NOUN
easat-10684	449	62	without	without	ADP
easat-10684	449	63	the	the	DET
easat-10684	449	64	bi	bi	ADJ
easat-10684	449	65	-	-	ADJ
easat-10684	449	66	lstm	lstm	ADJ
easat-10684	449	67	weight	weight	NOUN
easat-10684	449	68	.	.	PUNCT
easat-10684	450	1	the	the	DET
easat-10684	450	2	accuracy	accuracy	NOUN
easat-10684	450	3	,	,	PUNCT
easat-10684	450	4	precision	precision	NOUN
easat-10684	450	5	,	,	PUNCT
easat-10684	450	6	recall	recall	NOUN
easat-10684	450	7	,	,	PUNCT
easat-10684	450	8	and	and	CCONJ
easat-10684	450	9	f1	f1	NOUN
easat-10684	450	10	-	-	PUNCT
easat-10684	450	11	score	score	NOUN
easat-10684	450	12	are	be	AUX
easat-10684	450	13	displayed	display	VERB
easat-10684	450	14	in	in	ADP
easat-10684	450	15	bars	bar	NOUN
easat-10684	450	16	for	for	ADP
easat-10684	450	17	every	every	DET
easat-10684	450	18	model	model	NOUN
easat-10684	450	19	.	.	PUNCT
easat-10684	451	1	the	the	DET
easat-10684	451	2	bar	bar	NOUN
easat-10684	451	3	chart	chart	NOUN
easat-10684	451	4	shows	show	VERB
easat-10684	451	5	a	a	DET
easat-10684	451	6	graphical	graphical	ADJ
easat-10684	451	7	representation	representation	NOUN
easat-10684	451	8	of	of	ADP
easat-10684	451	9	table	table	NOUN
easat-10684	451	10	2	2	NUM
easat-10684	451	11	,	,	PUNCT
easat-10684	451	12	and	and	CCONJ
easat-10684	451	13	it	it	PRON
easat-10684	451	14	can	can	AUX
easat-10684	451	15	be	be	AUX
easat-10684	451	16	seen	see	VERB
easat-10684	451	17	that	that	SCONJ
easat-10684	451	18	svm	svm	PROPN
easat-10684	451	19	performs	perform	VERB
easat-10684	451	20	best	good	ADJ
easat-10684	451	21	,	,	PUNCT
easat-10684	451	22	and	and	CCONJ
easat-10684	451	23	the	the	DET
easat-10684	451	24	decision	decision	NOUN
easat-10684	451	25	tree	tree	NOUN
easat-10684	451	26	has	have	VERB
easat-10684	451	27	the	the	DET
easat-10684	451	28	lowest	low	ADJ
easat-10684	451	29	performance	performance	NOUN
easat-10684	451	30	(	(	PUNCT
easat-10684	451	31	below	below	ADP
easat-10684	451	32	80	80	NUM
easat-10684	451	33	%	%	NOUN
easat-10684	451	34	)	)	PUNCT
easat-10684	451	35	.	.	PUNCT
easat-10684	452	1	figure	figure	VERB
easat-10684	452	2	10	10	NUM
easat-10684	452	3	.	.	PUNCT
easat-10684	453	1	machine	machine	NOUN
easat-10684	453	2	learning	learn	VERB
easat-10684	453	3	algorithm	algorithm	PROPN
easat-10684	453	4	performance	performance	NOUN
easat-10684	453	5	comparison	comparison	NOUN
easat-10684	453	6	.	.	PUNCT
easat-10684	454	1	the	the	DET
easat-10684	454	2	bar	bar	NOUN
easat-10684	454	3	chart	chart	NOUN
easat-10684	454	4	in	in	ADP
easat-10684	454	5	figure	figure	NOUN
easat-10684	454	6	11	11	NUM
easat-10684	454	7	illustrates	illustrate	VERB
easat-10684	454	8	the	the	DET
easat-10684	454	9	performance	performance	NOUN
easat-10684	454	10	of	of	ADP
easat-10684	454	11	various	various	ADJ
easat-10684	454	12	machine	machine	NOUN
easat-10684	454	13	learning	learn	VERB
easat-10684	454	14	algorithms	algorithm	NOUN
easat-10684	454	15	in	in	ADP
easat-10684	454	16	terms	term	NOUN
easat-10684	454	17	of	of	ADP
easat-10684	454	18	accuracy	accuracy	NOUN
easat-10684	454	19	,	,	PUNCT
easat-10684	454	20	precision	precision	NOUN
easat-10684	454	21	,	,	PUNCT
easat-10684	454	22	and	and	CCONJ
easat-10684	454	23	recall	recall	NOUN
easat-10684	454	24	.	.	PUNCT
easat-10684	455	1	figure	figure	VERB
easat-10684	455	2	13	13	NUM
easat-10684	455	3	presents	present	VERB
easat-10684	455	4	the	the	DET
easat-10684	455	5	performance	performance	NOUN
easat-10684	455	6	of	of	ADP
easat-10684	455	7	these	these	DET
easat-10684	455	8	algorithms	algorithm	NOUN
easat-10684	455	9	in	in	ADP
easat-10684	455	10	terms	term	NOUN
easat-10684	455	11	of	of	ADP
easat-10684	455	12	accuracy	accuracy	NOUN
easat-10684	455	13	,	,	PUNCT
easat-10684	455	14	precision	precision	NOUN
easat-10684	455	15	,	,	PUNCT
easat-10684	455	16	recall	recall	NOUN
easat-10684	455	17	,	,	PUNCT
easat-10684	455	18	and	and	CCONJ
easat-10684	455	19	f1	f1	NOUN
easat-10684	455	20	-	-	PUNCT
easat-10684	455	21	score	score	NOUN
easat-10684	455	22	,	,	PUNCT
easat-10684	455	23	utilizing	utilize	VERB
easat-10684	455	24	bi	bi	ADJ
easat-10684	455	25	-	-	ADJ
easat-10684	455	26	lstm	lstm	ADJ
easat-10684	455	27	transfer	transfer	NOUN
easat-10684	455	28	learning	learn	VERB
easat-10684	455	29	to	to	PART
easat-10684	455	30	enhance	enhance	VERB
easat-10684	455	31	model	model	NOUN
easat-10684	455	32	performance	performance	NOUN
easat-10684	455	33	.	.	PUNCT
easat-10684	456	1	the	the	DET
easat-10684	456	2	bar	bar	NOUN
easat-10684	456	3	chart	chart	NOUN
easat-10684	456	4	provides	provide	VERB
easat-10684	456	5	a	a	DET
easat-10684	456	6	graphical	graphical	ADJ
easat-10684	456	7	representation	representation	NOUN
easat-10684	456	8	of	of	ADP
easat-10684	456	9	table	table	NOUN
easat-10684	456	10	2	2	NUM
easat-10684	456	11	,	,	PUNCT
easat-10684	456	12	indicating	indicate	VERB
easat-10684	456	13	that	that	SCONJ
easat-10684	456	14	logistic	logistic	ADJ
easat-10684	456	15	regression	regression	NOUN
easat-10684	456	16	(	(	PUNCT
easat-10684	456	17	lr	lr	NOUN
easat-10684	456	18	)	)	PUNCT
easat-10684	456	19	performs	perform	VERB
easat-10684	456	20	the	the	DET
easat-10684	456	21	best	good	ADJ
easat-10684	456	22	,	,	PUNCT
easat-10684	456	23	while	while	SCONJ
easat-10684	456	24	the	the	DET
easat-10684	456	25	decision	decision	NOUN
easat-10684	456	26	tree	tree	NOUN
easat-10684	456	27	algorithm	algorithm	NOUN
easat-10684	456	28	has	have	VERB
easat-10684	456	29	the	the	DET
easat-10684	456	30	lowest	low	ADJ
easat-10684	456	31	accuracy	accuracy	NOUN
easat-10684	456	32	performance	performance	NOUN
easat-10684	456	33	at	at	ADP
easat-10684	456	34	91.2	91.2	NUM
easat-10684	456	35	%	%	NOUN
easat-10684	456	36	.	.	PUNCT
easat-10684	457	1	1535	1535	NUM
easat-10684	457	2	edelweiss	edelweiss	PROPN
easat-10684	457	3	applied	apply	VERB
easat-10684	457	4	science	science	NOUN
easat-10684	457	5	and	and	CCONJ
easat-10684	457	6	technology	technology	NOUN
easat-10684	457	7	issn	issn	PROPN
easat-10684	457	8	:	:	PUNCT
easat-10684	457	9	2576	2576	NUM
easat-10684	457	10	-	-	SYM
easat-10684	457	11	8484	8484	NUM
easat-10684	457	12	vol	vol	NOUN
easat-10684	457	13	.	.	PROPN
easat-10684	458	1	9	9	NUM
easat-10684	458	2	,	,	PUNCT
easat-10684	458	3	no	no	INTJ
easat-10684	458	4	.	.	NOUN
easat-10684	458	5	10	10	NUM
easat-10684	458	6	:	:	PUNCT
easat-10684	458	7	1515	1515	NUM
easat-10684	458	8	-	-	SYM
easat-10684	458	9	1540	1540	NUM
easat-10684	458	10	,	,	PUNCT
easat-10684	458	11	2025	2025	NUM
easat-10684	458	12	doi	doi	NOUN
easat-10684	458	13	:	:	PUNCT
easat-10684	458	14	10.55214/2576	10.55214/2576	NUM
easat-10684	458	15	-	-	SYM
easat-10684	458	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	459	1	©	©	PROPN
easat-10684	459	2	2025	2025	NUM
easat-10684	459	3	by	by	ADP
easat-10684	459	4	the	the	DET
easat-10684	459	5	authors	author	NOUN
easat-10684	459	6	;	;	PUNCT
easat-10684	459	7	licensee	licensee	PROPN
easat-10684	459	8	learning	learn	VERB
easat-10684	459	9	gate	gate	PROPN
easat-10684	459	10	figure	figure	NOUN
easat-10684	459	11	11	11	NUM
easat-10684	459	12	.	.	PUNCT
easat-10684	460	1	machine	machine	NOUN
easat-10684	460	2	learning	learn	VERB
easat-10684	460	3	algorithm	algorithm	NOUN
easat-10684	460	4	with	with	ADP
easat-10684	460	5	bi	bi	ADJ
easat-10684	460	6	-	-	ADJ
easat-10684	460	7	lstm	lstm	ADJ
easat-10684	460	8	transfer	transfer	NOUN
easat-10684	460	9	learning	learning	NOUN
easat-10684	460	10	.	.	PUNCT
easat-10684	461	1	table	table	NOUN
easat-10684	461	2	3	3	NUM
easat-10684	461	3	shows	show	VERB
easat-10684	461	4	the	the	DET
easat-10684	461	5	performance	performance	NOUN
easat-10684	461	6	increase	increase	NOUN
easat-10684	461	7	in	in	ADP
easat-10684	461	8	accuracy	accuracy	NOUN
easat-10684	461	9	,	,	PUNCT
easat-10684	461	10	precision	precision	NOUN
easat-10684	461	11	,	,	PUNCT
easat-10684	461	12	recall	recall	NOUN
easat-10684	461	13	,	,	PUNCT
easat-10684	461	14	and	and	CCONJ
easat-10684	461	15	f1	f1	NOUN
easat-10684	461	16	-	-	PUNCT
easat-10684	461	17	score	score	NOUN
easat-10684	461	18	,	,	PUNCT
easat-10684	461	19	using	use	VERB
easat-10684	461	20	the	the	DET
easat-10684	461	21	concept	concept	NOUN
easat-10684	461	22	of	of	ADP
easat-10684	461	23	transfer	transfer	NOUN
easat-10684	461	24	learning	learn	VERB
easat-10684	461	25	to	to	PART
easat-10684	461	26	improve	improve	VERB
easat-10684	461	27	the	the	DET
easat-10684	461	28	performance	performance	NOUN
easat-10684	461	29	of	of	ADP
easat-10684	461	30	the	the	DET
easat-10684	461	31	trained	train	VERB
easat-10684	461	32	machine	machine	NOUN
easat-10684	461	33	learning	learn	VERB
easat-10684	461	34	algorithm	algorithm	NOUN
easat-10684	461	35	with	with	ADP
easat-10684	461	36	the	the	DET
easat-10684	461	37	bi	bi	ADJ
easat-10684	461	38	-	-	ADJ
easat-10684	461	39	lstm	lstm	ADJ
easat-10684	461	40	pre	pre	ADJ
easat-10684	461	41	-	-	ADJ
easat-10684	461	42	trained	train	VERB
easat-10684	461	43	model	model	NOUN
easat-10684	461	44	.	.	PUNCT
easat-10684	462	1	based	base	VERB
easat-10684	462	2	on	on	ADP
easat-10684	462	3	table	table	NOUN
easat-10684	462	4	3	3	NUM
easat-10684	462	5	,	,	PUNCT
easat-10684	462	6	it	it	PRON
easat-10684	462	7	is	be	AUX
easat-10684	462	8	revealed	reveal	VERB
easat-10684	462	9	that	that	SCONJ
easat-10684	462	10	logistic	logistic	ADJ
easat-10684	462	11	regression	regression	NOUN
easat-10684	462	12	,	,	PUNCT
easat-10684	462	13	naïve	naïve	ADJ
easat-10684	462	14	bayes	bayes	NOUN
easat-10684	462	15	,	,	PUNCT
easat-10684	462	16	and	and	CCONJ
easat-10684	462	17	support	support	VERB
easat-10684	462	18	vector	vector	NOUN
easat-10684	462	19	machine	machine	NOUN
easat-10684	462	20	attain	attain	VERB
easat-10684	462	21	accuracies	accuracy	NOUN
easat-10684	462	22	of	of	ADP
easat-10684	462	23	92.2	92.2	NUM
easat-10684	462	24	%	%	NOUN
easat-10684	462	25	,	,	PUNCT
easat-10684	462	26	92.1	92.1	NUM
easat-10684	462	27	%	%	NOUN
easat-10684	462	28	,	,	PUNCT
easat-10684	462	29	and	and	CCONJ
easat-10684	462	30	92.0	92.0	NUM
easat-10684	462	31	%	%	NOUN
easat-10684	462	32	,	,	PUNCT
easat-10684	462	33	respectively	respectively	ADV
easat-10684	462	34	,	,	PUNCT
easat-10684	462	35	while	while	SCONJ
easat-10684	462	36	nearest	near	ADJ
easat-10684	462	37	neighbour	neighbour	NOUN
easat-10684	462	38	and	and	CCONJ
easat-10684	462	39	decision	decision	NOUN
easat-10684	462	40	tree	tree	NOUN
easat-10684	462	41	obtain	obtain	VERB
easat-10684	462	42	accuracies	accuracy	NOUN
easat-10684	462	43	of	of	ADP
easat-10684	462	44	91.6	91.6	NUM
easat-10684	462	45	%	%	NOUN
easat-10684	462	46	and	and	CCONJ
easat-10684	462	47	91.2	91.2	NUM
easat-10684	462	48	%	%	NOUN
easat-10684	462	49	.	.	PUNCT
easat-10684	463	1	table	table	NOUN
easat-10684	463	2	3	3	NUM
easat-10684	463	3	.	.	PUNCT
easat-10684	463	4	machine	machine	NOUN
easat-10684	463	5	learning	learning	NOUN
easat-10684	463	6	performance	performance	NOUN
easat-10684	463	7	(	(	PUNCT
easat-10684	463	8	bi	bi	NOUN
easat-10684	463	9	-	-	ADJ
easat-10684	463	10	lstm-2	lstm-2	NOUN
easat-10684	463	11	-	-	PUNCT
easat-10684	463	12	ml	ml	NOUN
easat-10684	463	13	)	)	PUNCT
easat-10684	463	14	.	.	PUNCT
easat-10684	464	1	s	s	X
easat-10684	464	2	/	/	SYM
easat-10684	464	3	n	n	CCONJ
easat-10684	464	4	model	model	NOUN
easat-10684	464	5	accuracy	accuracy	NOUN
easat-10684	464	6	precision	precision	NOUN
easat-10684	464	7	recall	recall	VERB
easat-10684	464	8	f1	f1	NOUN
easat-10684	464	9	-	-	PUNCT
easat-10684	464	10	score	score	NOUN
easat-10684	464	11	1	1	NUM
easat-10684	464	12	logistic	logistic	ADJ
easat-10684	464	13	regression	regression	NOUN
easat-10684	464	14	0.921671	0.921671	NUM
easat-10684	464	15	0.921847	0.921847	NUM
easat-10684	464	16	0.918584	0.918584	NUM
easat-10684	464	17	0.920213	0.920213	NUM
easat-10684	464	18	2	2	NUM
easat-10684	464	19	k	k	ADV
easat-10684	464	20	-	-	PUNCT
easat-10684	464	21	nearest	near	ADJ
easat-10684	464	22	neighbor	neighbor	NOUN
easat-10684	464	23	0.916449	0.916449	PROPN
easat-10684	464	24	0.916519	0.916519	NUM
easat-10684	464	25	0.913274	0.913274	NUM
easat-10684	464	26	0.914894	0.914894	NUM
easat-10684	464	27	3	3	NUM
easat-10684	464	28	naive	naive	ADJ
easat-10684	464	29	bayes	bayes	NOUN
easat-10684	464	30	0.921671	0.921671	NUM
easat-10684	465	1	0.921847	0.921847	NUM
easat-10684	465	2	0.918584	0.918584	NUM
easat-10684	465	3	0.920213	0.920213	NUM
easat-10684	465	4	4	4	NUM
easat-10684	465	5	support	support	NOUN
easat-10684	465	6	vector	vector	NOUN
easat-10684	465	7	machine	machine	NOUN
easat-10684	465	8	0.920801	0.920801	NUM
easat-10684	465	9	0.926259	0.926259	NUM
easat-10684	465	10	0.911504	0.911504	NUM
easat-10684	465	11	0.918822	0.918822	NUM
easat-10684	465	12	5	5	NUM
easat-10684	465	13	decision	decision	NOUN
easat-10684	465	14	tree	tree	NOUN
easat-10684	465	15	0.912097	0.912097	NUM
easat-10684	465	16	0.911348	0.911348	NUM
easat-10684	465	17	0.909735	0.909735	NUM
easat-10684	465	18	0.91054	0.91054	NUM
easat-10684	465	19	furthermore	furthermore	ADV
easat-10684	465	20	,	,	PUNCT
easat-10684	465	21	table	table	NOUN
easat-10684	465	22	4	4	NUM
easat-10684	465	23	shows	show	VERB
easat-10684	465	24	the	the	DET
easat-10684	465	25	performance	performance	NOUN
easat-10684	465	26	comparison	comparison	NOUN
easat-10684	465	27	between	between	ADP
easat-10684	465	28	the	the	DET
easat-10684	465	29	benchmark	benchmark	ADJ
easat-10684	465	30	accuracy	accuracy	NOUN
easat-10684	465	31	of	of	ADP
easat-10684	465	32	various	various	ADJ
easat-10684	465	33	machine	machine	NOUN
easat-10684	465	34	learning	learn	VERB
easat-10684	465	35	fake	fake	ADJ
easat-10684	465	36	news	news	NOUN
easat-10684	465	37	detection	detection	NOUN
easat-10684	465	38	models	model	NOUN
easat-10684	465	39	and	and	CCONJ
easat-10684	465	40	the	the	DET
easat-10684	465	41	proposed	propose	VERB
easat-10684	465	42	bi	bi	ADJ
easat-10684	465	43	-	-	ADJ
easat-10684	465	44	lstm	lstm	ADJ
easat-10684	465	45	transfer	transfer	NOUN
easat-10684	465	46	learning-2ml	learning-2ml	NOUN
easat-10684	465	47	fake	fake	ADJ
easat-10684	465	48	news	news	NOUN
easat-10684	465	49	detection	detection	NOUN
easat-10684	465	50	models	model	NOUN
easat-10684	465	51	.	.	PUNCT
easat-10684	466	1	it	it	PRON
easat-10684	466	2	is	be	AUX
easat-10684	466	3	clearly	clearly	ADV
easat-10684	466	4	shown	show	VERB
easat-10684	466	5	that	that	SCONJ
easat-10684	466	6	the	the	DET
easat-10684	466	7	machine	machine	NOUN
easat-10684	466	8	learning	learn	VERB
easat-10684	466	9	algorithm	algorithm	NOUN
easat-10684	466	10	trained	train	VERB
easat-10684	466	11	using	use	VERB
easat-10684	466	12	the	the	DET
easat-10684	466	13	bi	bi	ADJ
easat-10684	466	14	-	-	ADJ
easat-10684	466	15	lstm	lstm	ADJ
easat-10684	466	16	transfer	transfer	NOUN
easat-10684	466	17	weight	weight	NOUN
easat-10684	466	18	performs	perform	VERB
easat-10684	466	19	significantly	significantly	ADV
easat-10684	466	20	better	well	ADV
easat-10684	466	21	than	than	ADP
easat-10684	466	22	machine	machine	NOUN
easat-10684	466	23	learning	learn	VERB
easat-10684	466	24	fake	fake	ADJ
easat-10684	466	25	news	news	NOUN
easat-10684	466	26	detection	detection	NOUN
easat-10684	466	27	without	without	ADP
easat-10684	466	28	the	the	DET
easat-10684	466	29	bi	bi	ADJ
easat-10684	466	30	-	-	ADJ
easat-10684	466	31	lstm	lstm	ADJ
easat-10684	466	32	enhancement	enhancement	NOUN
easat-10684	466	33	.	.	PUNCT
easat-10684	467	1	it	it	PRON
easat-10684	467	2	is	be	AUX
easat-10684	467	3	also	also	ADV
easat-10684	467	4	revealed	reveal	VERB
easat-10684	467	5	that	that	SCONJ
easat-10684	467	6	support	support	NOUN
easat-10684	467	7	vector	vector	NOUN
easat-10684	467	8	machine	machine	NOUN
easat-10684	467	9	improves	improve	VERB
easat-10684	467	10	in	in	ADP
easat-10684	467	11	performance	performance	NOUN
easat-10684	467	12	from	from	ADP
easat-10684	467	13	87.4	87.4	NUM
easat-10684	467	14	%	%	NOUN
easat-10684	467	15	to	to	ADP
easat-10684	467	16	92.0	92.0	NUM
easat-10684	467	17	%	%	NOUN
easat-10684	467	18	.	.	PUNCT
easat-10684	468	1	in	in	ADP
easat-10684	468	2	comparison	comparison	NOUN
easat-10684	468	3	,	,	PUNCT
easat-10684	468	4	logistic	logistic	ADJ
easat-10684	468	5	regression	regression	NOUN
easat-10684	468	6	achieved	achieve	VERB
easat-10684	468	7	92.2	92.2	NUM
easat-10684	468	8	%	%	NOUN
easat-10684	468	9	from	from	ADP
easat-10684	468	10	86.0	86.0	NUM
easat-10684	468	11	%	%	NOUN
easat-10684	468	12	,	,	PUNCT
easat-10684	468	13	knearest	knearest	NOUN
easat-10684	468	14	neighbour	neighbour	NOUN
easat-10684	468	15	from	from	ADP
easat-10684	468	16	85.4	85.4	NUM
easat-10684	468	17	%	%	NOUN
easat-10684	468	18	to	to	ADP
easat-10684	468	19	91.6	91.6	NUM
easat-10684	468	20	%	%	NOUN
easat-10684	468	21	,	,	PUNCT
easat-10684	468	22	naïve	naïve	ADJ
easat-10684	468	23	bayes	bayes	PROPN
easat-10684	468	24	obtained	obtain	VERB
easat-10684	468	25	an	an	DET
easat-10684	468	26	accuracy	accuracy	NOUN
easat-10684	468	27	increase	increase	NOUN
easat-10684	468	28	from	from	ADP
easat-10684	468	29	79.0	79.0	NUM
easat-10684	468	30	%	%	NOUN
easat-10684	468	31	to	to	ADP
easat-10684	468	32	92.1	92.1	NUM
easat-10684	468	33	%	%	NOUN
easat-10684	468	34	,	,	PUNCT
easat-10684	468	35	and	and	CCONJ
easat-10684	468	36	finally	finally	ADV
easat-10684	468	37	,	,	PUNCT
easat-10684	468	38	decision	decision	NOUN
easat-10684	468	39	tree	tree	NOUN
easat-10684	468	40	attained	attain	VERB
easat-10684	468	41	91.2	91.2	NUM
easat-10684	468	42	%	%	NOUN
easat-10684	468	43	from	from	ADP
easat-10684	468	44	78.5	78.5	NUM
easat-10684	468	45	%	%	NOUN
easat-10684	468	46	.	.	PUNCT
easat-10684	469	1	this	this	DET
easat-10684	469	2	performance	performance	NOUN
easat-10684	469	3	difference	difference	NOUN
easat-10684	469	4	can	can	AUX
easat-10684	469	5	be	be	AUX
easat-10684	469	6	seen	see	VERB
easat-10684	469	7	in	in	ADP
easat-10684	469	8	the	the	DET
easat-10684	469	9	bar	bar	NOUN
easat-10684	469	10	chart	chart	NOUN
easat-10684	469	11	in	in	ADP
easat-10684	469	12	figure	figure	NOUN
easat-10684	469	13	12	12	NUM
easat-10684	469	14	.	.	PUNCT
easat-10684	469	15	table	table	NOUN
easat-10684	469	16	4	4	NUM
easat-10684	469	17	.	.	PUNCT
easat-10684	470	1	ml	ml	PROPN
easat-10684	470	2	vs	vs	ADP
easat-10684	470	3	bi	bi	PROPN
easat-10684	470	4	-	-	PUNCT
easat-10684	470	5	lstm-2	lstm-2	NOUN
easat-10684	470	6	-	-	PUNCT
easat-10684	470	7	ml	ml	NOUN
easat-10684	470	8	.	.	PUNCT
easat-10684	471	1	s	s	X
easat-10684	471	2	/	/	SYM
easat-10684	471	3	n	n	PRON
easat-10684	471	4	model	model	NOUN
easat-10684	471	5	proposed	propose	VERB
easat-10684	471	6	bi	bi	ADJ
easat-10684	471	7	-	-	ADJ
easat-10684	471	8	lstm	lstm	ADJ
easat-10684	471	9	transfer	transfer	NOUN
easat-10684	471	10	learning-2	learning-2	NUM
easat-10684	471	11	-	-	PUNCT
easat-10684	471	12	ml	ml	NOUN
easat-10684	471	13	machine	machine	NOUN
easat-10684	471	14	learning	learning	NOUN
easat-10684	471	15	(	(	PUNCT
easat-10684	471	16	benchmark	benchmark	NOUN
easat-10684	471	17	)	)	PUNCT
easat-10684	471	18	1	1	NUM
easat-10684	471	19	logistic	logistic	ADJ
easat-10684	471	20	regression	regression	NOUN
easat-10684	471	21	92.16	92.16	NUM
easat-10684	471	22	%	%	NOUN
easat-10684	471	23	86.00	86.00	NUM
easat-10684	471	24	%	%	NOUN
easat-10684	471	25	2	2	NUM
easat-10684	471	26	k	k	NOUN
easat-10684	471	27	-	-	PUNCT
easat-10684	471	28	nearest	near	ADJ
easat-10684	471	29	neighbor	neighbor	NOUN
easat-10684	471	30	91.64	91.64	NUM
easat-10684	471	31	%	%	NOUN
easat-10684	471	32	85.49	85.49	NUM
easat-10684	471	33	%	%	NOUN
easat-10684	471	34	3	3	NUM
easat-10684	471	35	naive	naive	ADJ
easat-10684	471	36	bayes	baye	NOUN
easat-10684	471	37	92.16	92.16	NUM
easat-10684	471	38	%	%	NOUN
easat-10684	471	39	79.04	79.04	NUM
easat-10684	471	40	%	%	NOUN
easat-10684	471	41	4	4	NUM
easat-10684	471	42	svm	svm	NOUN
easat-10684	471	43	92.08	92.08	NUM
easat-10684	471	44	%	%	NOUN
easat-10684	471	45	87.45	87.45	NUM
easat-10684	471	46	%	%	NOUN
easat-10684	471	47	5	5	NUM
easat-10684	471	48	decision	decision	NOUN
easat-10684	471	49	tree	tree	NOUN
easat-10684	471	50	91.20	91.20	NUM
easat-10684	471	51	%	%	NOUN
easat-10684	471	52	78.53	78.53	NUM
easat-10684	471	53	%	%	NOUN
easat-10684	471	54	1536	1536	NUM
easat-10684	471	55	edelweiss	edelweiss	PROPN
easat-10684	471	56	applied	apply	VERB
easat-10684	471	57	science	science	NOUN
easat-10684	471	58	and	and	CCONJ
easat-10684	471	59	technology	technology	NOUN
easat-10684	471	60	issn	issn	PROPN
easat-10684	471	61	:	:	PUNCT
easat-10684	471	62	2576	2576	NUM
easat-10684	471	63	-	-	SYM
easat-10684	471	64	8484	8484	NUM
easat-10684	471	65	vol	vol	NOUN
easat-10684	471	66	.	.	PROPN
easat-10684	472	1	9	9	NUM
easat-10684	472	2	,	,	PUNCT
easat-10684	472	3	no	no	INTJ
easat-10684	472	4	.	.	NOUN
easat-10684	472	5	10	10	NUM
easat-10684	472	6	:	:	PUNCT
easat-10684	472	7	1515	1515	NUM
easat-10684	472	8	-	-	SYM
easat-10684	472	9	1540	1540	NUM
easat-10684	472	10	,	,	PUNCT
easat-10684	472	11	2025	2025	NUM
easat-10684	472	12	doi	doi	NOUN
easat-10684	472	13	:	:	PUNCT
easat-10684	472	14	10.55214/2576	10.55214/2576	NUM
easat-10684	472	15	-	-	SYM
easat-10684	472	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	473	1	©	©	PROPN
easat-10684	473	2	2025	2025	NUM
easat-10684	473	3	by	by	ADP
easat-10684	473	4	the	the	DET
easat-10684	473	5	authors	author	NOUN
easat-10684	473	6	;	;	PUNCT
easat-10684	473	7	licensee	licensee	PROPN
easat-10684	473	8	learning	learn	VERB
easat-10684	473	9	gate	gate	PROPN
easat-10684	473	10	figure	figure	NOUN
easat-10684	473	11	12	12	NUM
easat-10684	473	12	shows	show	VERB
easat-10684	473	13	the	the	DET
easat-10684	473	14	performance	performance	NOUN
easat-10684	473	15	comparison	comparison	NOUN
easat-10684	473	16	between	between	ADP
easat-10684	473	17	the	the	DET
easat-10684	473	18	bi	bi	ADJ
easat-10684	473	19	-	-	ADJ
easat-10684	473	20	lstm	lstm	ADJ
easat-10684	473	21	transfer	transfer	NOUN
easat-10684	473	22	learning-2	learning-2	NOUN
easat-10684	473	23	-	-	PUNCT
easat-10684	473	24	ml	ml	NOUN
easat-10684	473	25	and	and	CCONJ
easat-10684	473	26	the	the	DET
easat-10684	473	27	machine	machine	NOUN
easat-10684	473	28	learning	learning	NOUN
easat-10684	473	29	(	(	PUNCT
easat-10684	473	30	benchmarked	benchmarke	VERB
easat-10684	473	31	)	)	PUNCT
easat-10684	473	32	models	model	NOUN
easat-10684	473	33	'	'	PART
easat-10684	473	34	accuracy	accuracy	NOUN
easat-10684	473	35	.	.	PUNCT
easat-10684	474	1	regarding	regard	VERB
easat-10684	474	2	the	the	DET
easat-10684	474	3	figure	figure	NOUN
easat-10684	474	4	,	,	PUNCT
easat-10684	474	5	it	it	PRON
easat-10684	474	6	is	be	AUX
easat-10684	474	7	revealed	reveal	VERB
easat-10684	474	8	that	that	SCONJ
easat-10684	474	9	the	the	DET
easat-10684	474	10	proposed	propose	VERB
easat-10684	474	11	bi	bi	ADJ
easat-10684	474	12	-	-	ADJ
easat-10684	474	13	lstm	lstm	ADJ
easat-10684	474	14	transfer	transfer	NOUN
easat-10684	474	15	learning-2	learning-2	NOUN
easat-10684	474	16	-	-	PUNCT
easat-10684	474	17	ml	ml	NOUN
easat-10684	474	18	performs	perform	NOUN
easat-10684	474	19	efficiently	efficiently	ADV
easat-10684	474	20	,	,	PUNCT
easat-10684	474	21	with	with	ADP
easat-10684	474	22	an	an	DET
easat-10684	474	23	increase	increase	NOUN
easat-10684	474	24	of	of	ADP
easat-10684	474	25	8.54	8.54	NUM
easat-10684	474	26	%	%	NOUN
easat-10684	474	27	in	in	ADP
easat-10684	474	28	average	average	ADJ
easat-10684	474	29	accuracy	accuracy	NOUN
easat-10684	474	30	,	,	PUNCT
easat-10684	474	31	surpassing	surpass	VERB
easat-10684	474	32	the	the	DET
easat-10684	474	33	benchmark	benchmark	ADJ
easat-10684	474	34	machine	machine	NOUN
easat-10684	474	35	learning	learn	VERB
easat-10684	474	36	algorithms	algorithm	NOUN
easat-10684	474	37	.	.	PUNCT
easat-10684	475	1	table	table	NOUN
easat-10684	475	2	5	5	NUM
easat-10684	475	3	presents	present	VERB
easat-10684	475	4	a	a	DET
easat-10684	475	5	comparison	comparison	NOUN
easat-10684	475	6	of	of	ADP
easat-10684	475	7	the	the	DET
easat-10684	475	8	proposed	propose	VERB
easat-10684	475	9	algorithm	algorithm	NOUN
easat-10684	475	10	with	with	ADP
easat-10684	475	11	the	the	DET
easat-10684	475	12	state	state	NOUN
easat-10684	475	13	-	-	PUNCT
easat-10684	475	14	of	of	ADP
easat-10684	475	15	-	-	PUNCT
easat-10684	475	16	the	the	DET
easat-10684	475	17	-	-	PUNCT
easat-10684	475	18	art	art	NOUN
easat-10684	475	19	studies	study	NOUN
easat-10684	475	20	.	.	PUNCT
easat-10684	476	1	figure	figure	NOUN
easat-10684	476	2	12	12	NUM
easat-10684	476	3	.	.	PUNCT
easat-10684	477	1	ml	ml	X
easat-10684	477	2	vs	vs	ADP
easat-10684	477	3	bi	bi	PROPN
easat-10684	477	4	-	-	PUNCT
easat-10684	477	5	lstm-2	lstm-2	NOUN
easat-10684	477	6	-	-	PUNCT
easat-10684	477	7	ml	ml	NOUN
easat-10684	477	8	.	.	PUNCT
easat-10684	477	9	table	table	NOUN
easat-10684	477	10	5	5	NUM
easat-10684	477	11	.	.	PUNCT
easat-10684	478	1	proposed	propose	VERB
easat-10684	478	2	bi	bi	PROPN
easat-10684	478	3	-	-	PUNCT
easat-10684	478	4	lstm-2	lstm-2	NOUN
easat-10684	478	5	-	-	PUNCT
easat-10684	478	6	ml	ml	NOUN
easat-10684	478	7	vs	vs	ADP
easat-10684	478	8	benchmark	benchmark	NOUN
easat-10684	478	9	models	model	NOUN
easat-10684	478	10	.	.	PUNCT
easat-10684	479	1	s	s	X
easat-10684	479	2	/	/	SYM
easat-10684	479	3	n	n	CCONJ
easat-10684	479	4	model	model	NOUN
easat-10684	479	5	machine	machine	NOUN
easat-10684	479	6	learning	learning	NOUN
easat-10684	479	7	(	(	PUNCT
easat-10684	479	8	benchmark	benchmark	PROPN
easat-10684	479	9	)	)	PUNCT
easat-10684	479	10	pandey	pandey	PROPN
easat-10684	479	11	et	et	PROPN
easat-10684	479	12	al	al	PROPN
easat-10684	479	13	.	.	PUNCT
easat-10684	480	1	[	[	X
easat-10684	480	2	6	6	NUM
easat-10684	480	3	]	]	PUNCT
easat-10684	480	4	jaybhaye	jaybhaye	NOUN
easat-10684	480	5	[	[	X
easat-10684	480	6	13	13	NUM
easat-10684	480	7	]	]	PUNCT
easat-10684	480	8	proposed	propose	VERB
easat-10684	480	9	bi	bi	ADJ
easat-10684	480	10	-	-	ADJ
easat-10684	480	11	lstm	lstm	ADJ
easat-10684	480	12	transfer	transfer	NOUN
easat-10684	480	13	learning-2	learning-2	NOUN
easat-10684	480	14	-	-	PUNCT
easat-10684	480	15	ml	ml	ADP
easat-10684	480	16	1	1	NUM
easat-10684	480	17	logistic	logistic	ADJ
easat-10684	480	18	regression	regression	NOUN
easat-10684	480	19	86.00	86.00	NUM
easat-10684	480	20	%	%	NOUN
easat-10684	480	21	90.46	90.46	NUM
easat-10684	480	22	%	%	NOUN
easat-10684	480	23	91.22	91.22	NUM
easat-10684	480	24	%	%	NOUN
easat-10684	480	25	92.16	92.16	NUM
easat-10684	480	26	%	%	NOUN
easat-10684	480	27	2	2	NUM
easat-10684	480	28	k	k	NOUN
easat-10684	480	29	-	-	PUNCT
easat-10684	480	30	nearest	near	ADJ
easat-10684	480	31	neighbor	neighbor	NOUN
easat-10684	480	32	85.49	85.49	NUM
easat-10684	480	33	%	%	NOUN
easat-10684	480	34	89.98	89.98	NUM
easat-10684	480	35	%	%	NOUN
easat-10684	480	36	74.68	74.68	NUM
easat-10684	480	37	%	%	NOUN
easat-10684	480	38	91.64	91.64	NUM
easat-10684	480	39	%	%	NOUN
easat-10684	480	40	3	3	NUM
easat-10684	480	41	naive	naive	ADJ
easat-10684	480	42	bayes	bayes	NOUN
easat-10684	480	43	79.04	79.04	NUM
easat-10684	480	44	%	%	NOUN
easat-10684	480	45	86.89	86.89	NUM
easat-10684	480	46	%	%	NOUN
easat-10684	480	47	79.41	79.41	NUM
easat-10684	480	48	%	%	NOUN
easat-10684	480	49	92.16	92.16	NUM
easat-10684	480	50	%	%	NOUN
easat-10684	480	51	4	4	NUM
easat-10684	480	52	svm	svm	NOUN
easat-10684	480	53	87.45	87.45	NUM
easat-10684	480	54	%	%	NOUN
easat-10684	480	55	73.33	73.33	NUM
easat-10684	480	56	%	%	NOUN
easat-10684	480	57	*	*	PUNCT
easat-10684	480	58	*	*	PUNCT
easat-10684	480	59	*	*	PUNCT
easat-10684	480	60	*	*	PUNCT
easat-10684	480	61	*	*	PUNCT
easat-10684	480	62	92.08	92.08	NUM
easat-10684	480	63	%	%	NOUN
easat-10684	480	64	5	5	NUM
easat-10684	480	65	decision	decision	NOUN
easat-10684	480	66	tree	tree	NOUN
easat-10684	480	67	78.53	78.53	NUM
easat-10684	480	68	%	%	NOUN
easat-10684	480	69	89.33	89.33	NUM
easat-10684	480	70	%	%	NOUN
easat-10684	480	71	89.00	89.00	NUM
easat-10684	480	72	%	%	NOUN
easat-10684	480	73	91.20	91.20	NUM
easat-10684	480	74	%	%	NOUN
easat-10684	480	75	table	table	NOUN
easat-10684	480	76	5	5	NUM
easat-10684	480	77	and	and	CCONJ
easat-10684	480	78	figure	figure	VERB
easat-10684	480	79	13	13	NUM
easat-10684	480	80	show	show	VERB
easat-10684	480	81	the	the	DET
easat-10684	480	82	results	result	NOUN
easat-10684	480	83	and	and	CCONJ
easat-10684	480	84	performance	performance	NOUN
easat-10684	480	85	comparison	comparison	NOUN
easat-10684	480	86	between	between	ADP
easat-10684	480	87	the	the	DET
easat-10684	480	88	proposed	propose	VERB
easat-10684	480	89	bilstm-2	bilstm-2	NOUN
easat-10684	480	90	-	-	PUNCT
easat-10684	480	91	ml	ml	ADV
easat-10684	480	92	fake	fake	ADJ
easat-10684	480	93	news	news	NOUN
easat-10684	480	94	detection	detection	NOUN
easat-10684	480	95	model	model	NOUN
easat-10684	480	96	,	,	PUNCT
easat-10684	480	97	benchmark	benchmark	NOUN
easat-10684	480	98	model	model	NOUN
easat-10684	480	99	,	,	PUNCT
easat-10684	480	100	pandey	pandey	PROPN
easat-10684	480	101	et	et	PROPN
easat-10684	480	102	al	al	PROPN
easat-10684	480	103	.	.	PUNCT
easat-10684	481	1	[	[	X
easat-10684	481	2	6	6	NUM
easat-10684	481	3	]	]	PUNCT
easat-10684	481	4	.	.	PUNCT
easat-10684	482	1	based	base	VERB
easat-10684	482	2	on	on	ADP
easat-10684	482	3	the	the	DET
easat-10684	482	4	table	table	NOUN
easat-10684	482	5	and	and	CCONJ
easat-10684	482	6	figure	figure	NOUN
easat-10684	482	7	,	,	PUNCT
easat-10684	482	8	it	it	PRON
easat-10684	482	9	is	be	AUX
easat-10684	482	10	revealed	reveal	VERB
easat-10684	482	11	that	that	SCONJ
easat-10684	482	12	the	the	DET
easat-10684	482	13	proposed	propose	VERB
easat-10684	482	14	bi	bi	NOUN
easat-10684	482	15	-	-	PUNCT
easat-10684	482	16	lstm-2	lstm-2	NOUN
easat-10684	482	17	-	-	PUNCT
easat-10684	482	18	ml	ml	ADV
easat-10684	482	19	fake	fake	ADJ
easat-10684	482	20	news	news	NOUN
easat-10684	482	21	detection	detection	NOUN
easat-10684	482	22	model	model	NOUN
easat-10684	482	23	performs	perform	VERB
easat-10684	482	24	1537	1537	NUM
easat-10684	482	25	edelweiss	edelweiss	PROPN
easat-10684	482	26	applied	apply	VERB
easat-10684	482	27	science	science	NOUN
easat-10684	482	28	and	and	CCONJ
easat-10684	482	29	technology	technology	NOUN
easat-10684	482	30	issn	issn	PROPN
easat-10684	482	31	:	:	PUNCT
easat-10684	482	32	2576	2576	NUM
easat-10684	482	33	-	-	SYM
easat-10684	482	34	8484	8484	NUM
easat-10684	482	35	vol	vol	NOUN
easat-10684	482	36	.	.	PROPN
easat-10684	483	1	9	9	NUM
easat-10684	483	2	,	,	PUNCT
easat-10684	483	3	no	no	INTJ
easat-10684	483	4	.	.	NOUN
easat-10684	483	5	10	10	NUM
easat-10684	483	6	:	:	PUNCT
easat-10684	483	7	1515	1515	NUM
easat-10684	483	8	-	-	SYM
easat-10684	483	9	1540	1540	NUM
easat-10684	483	10	,	,	PUNCT
easat-10684	483	11	2025	2025	NUM
easat-10684	483	12	doi	doi	NOUN
easat-10684	483	13	:	:	PUNCT
easat-10684	483	14	10.55214/2576	10.55214/2576	NUM
easat-10684	483	15	-	-	SYM
easat-10684	483	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	484	1	©	©	PROPN
easat-10684	484	2	2025	2025	NUM
easat-10684	484	3	by	by	ADP
easat-10684	484	4	the	the	DET
easat-10684	484	5	authors	author	NOUN
easat-10684	484	6	;	;	PUNCT
easat-10684	484	7	licensee	licensee	PROPN
easat-10684	484	8	learning	learning	NOUN
easat-10684	484	9	gate	gate	NOUN
easat-10684	484	10	significantly	significantly	ADV
easat-10684	484	11	better	well	ADV
easat-10684	484	12	than	than	ADP
easat-10684	484	13	the	the	DET
easat-10684	484	14	existing	exist	VERB
easat-10684	484	15	models	model	NOUN
easat-10684	484	16	,	,	PUNCT
easat-10684	484	17	and	and	CCONJ
easat-10684	484	18	the	the	DET
easat-10684	484	19	bi	bi	ADJ
easat-10684	484	20	-	-	ADJ
easat-10684	484	21	lstm	lstm	ADJ
easat-10684	484	22	pre	pre	ADJ
easat-10684	484	23	-	-	ADJ
easat-10684	484	24	trained	trained	ADJ
easat-10684	484	25	weights	weight	NOUN
easat-10684	484	26	have	have	AUX
easat-10684	484	27	enhanced	enhance	VERB
easat-10684	484	28	the	the	DET
easat-10684	484	29	performance	performance	NOUN
easat-10684	484	30	of	of	ADP
easat-10684	484	31	all	all	DET
easat-10684	484	32	the	the	DET
easat-10684	484	33	machine	machine	NOUN
easat-10684	484	34	learning	learning	NOUN
easat-10684	484	35	models	model	NOUN
easat-10684	484	36	with	with	ADP
easat-10684	484	37	better	well	ADJ
easat-10684	484	38	accuracy	accuracy	NOUN
easat-10684	484	39	than	than	ADP
easat-10684	484	40	the	the	DET
easat-10684	484	41	existing	exist	VERB
easat-10684	484	42	models	model	NOUN
easat-10684	484	43	.	.	PUNCT
easat-10684	485	1	figure	figure	NOUN
easat-10684	485	2	13	13	NUM
easat-10684	485	3	.	.	PUNCT
easat-10684	486	1	proposed	propose	VERB
easat-10684	486	2	bi	bi	PROPN
easat-10684	486	3	-	-	PUNCT
easat-10684	486	4	lstm-2	lstm-2	NOUN
easat-10684	486	5	-	-	PUNCT
easat-10684	486	6	ml	ml	NOUN
easat-10684	486	7	vs	vs	ADP
easat-10684	486	8	benchmark	benchmark	NOUN
easat-10684	486	9	models	model	NOUN
easat-10684	486	10	performance	performance	NOUN
easat-10684	486	11	comparison	comparison	NOUN
easat-10684	486	12	.	.	PUNCT
easat-10684	487	1	6	6	NUM
easat-10684	487	2	.	.	X
easat-10684	487	3	conclusion	conclusion	NOUN
easat-10684	487	4	and	and	CCONJ
easat-10684	487	5	recommendation	recommendation	VERB
easat-10684	487	6	the	the	DET
easat-10684	487	7	research	research	NOUN
easat-10684	487	8	study	study	NOUN
easat-10684	487	9	developed	develop	VERB
easat-10684	487	10	a	a	DET
easat-10684	487	11	fake	fake	ADJ
easat-10684	487	12	news	news	NOUN
easat-10684	487	13	detection	detection	NOUN
easat-10684	487	14	model	model	NOUN
easat-10684	487	15	using	use	VERB
easat-10684	487	16	five	five	NUM
easat-10684	487	17	machine	machine	NOUN
easat-10684	487	18	learning	learning	NOUN
easat-10684	487	19	models	model	NOUN
easat-10684	487	20	,	,	PUNCT
easat-10684	487	21	including	include	VERB
easat-10684	487	22	logistic	logistic	ADJ
easat-10684	487	23	regression	regression	NOUN
easat-10684	487	24	(	(	PUNCT
easat-10684	487	25	lr	lr	NOUN
easat-10684	487	26	)	)	PUNCT
easat-10684	487	27	,	,	PUNCT
easat-10684	487	28	k	k	X
easat-10684	487	29	-	-	PUNCT
easat-10684	487	30	nearest	near	ADJ
easat-10684	487	31	neighbour	neighbour	NOUN
easat-10684	487	32	(	(	PUNCT
easat-10684	487	33	knn	knn	PROPN
easat-10684	487	34	)	)	PUNCT
easat-10684	487	35	,	,	PUNCT
easat-10684	487	36	naïve	naïve	ADJ
easat-10684	487	37	bayes	bayes	PROPN
easat-10684	487	38	(	(	PUNCT
easat-10684	487	39	nb	nb	PROPN
easat-10684	487	40	)	)	PUNCT
easat-10684	487	41	,	,	PUNCT
easat-10684	487	42	support	support	NOUN
easat-10684	487	43	vector	vector	NOUN
easat-10684	487	44	machine	machine	NOUN
easat-10684	487	45	(	(	PUNCT
easat-10684	487	46	svm	svm	PROPN
easat-10684	487	47	)	)	PUNCT
easat-10684	487	48	,	,	PUNCT
easat-10684	487	49	and	and	CCONJ
easat-10684	487	50	decision	decision	NOUN
easat-10684	487	51	tree	tree	NOUN
easat-10684	487	52	(	(	PUNCT
easat-10684	487	53	dt	dt	NOUN
easat-10684	487	54	)	)	PUNCT
easat-10684	487	55	.	.	PUNCT
easat-10684	488	1	the	the	DET
easat-10684	488	2	research	research	NOUN
easat-10684	488	3	utilises	utilise	VERB
easat-10684	488	4	the	the	DET
easat-10684	488	5	bi	bi	ADJ
easat-10684	488	6	-	-	ADJ
easat-10684	488	7	lstm	lstm	ADJ
easat-10684	488	8	deep	deep	ADJ
easat-10684	488	9	learning	learning	NOUN
easat-10684	488	10	model	model	NOUN
easat-10684	488	11	to	to	PART
easat-10684	488	12	enhance	enhance	VERB
easat-10684	488	13	the	the	DET
easat-10684	488	14	performance	performance	NOUN
easat-10684	488	15	of	of	ADP
easat-10684	488	16	the	the	DET
easat-10684	488	17	prior	prior	ADJ
easat-10684	488	18	machine	machine	NOUN
easat-10684	488	19	learning	learning	NOUN
easat-10684	488	20	models	model	NOUN
easat-10684	488	21	mentioned	mention	VERB
easat-10684	488	22	,	,	PUNCT
easat-10684	488	23	using	use	VERB
easat-10684	488	24	the	the	DET
easat-10684	488	25	transfer	transfer	NOUN
easat-10684	488	26	learning	learning	NOUN
easat-10684	488	27	approach	approach	NOUN
easat-10684	488	28	.	.	PUNCT
easat-10684	489	1	hence	hence	ADV
easat-10684	489	2	,	,	PUNCT
easat-10684	489	3	the	the	DET
easat-10684	489	4	implemented	implement	VERB
easat-10684	489	5	bi	bi	NOUN
easat-10684	489	6	-	-	ADJ
easat-10684	489	7	lstm-2	lstm-2	NOUN
easat-10684	489	8	-	-	PUNCT
easat-10684	489	9	ml	ml	ADV
easat-10684	489	10	fake	fake	ADJ
easat-10684	489	11	news	news	NOUN
easat-10684	489	12	detection	detection	NOUN
easat-10684	489	13	model	model	NOUN
easat-10684	489	14	has	have	AUX
easat-10684	489	15	significantly	significantly	ADV
easat-10684	489	16	demonstrated	demonstrate	VERB
easat-10684	489	17	great	great	ADJ
easat-10684	489	18	performance	performance	NOUN
easat-10684	489	19	with	with	ADP
easat-10684	489	20	an	an	DET
easat-10684	489	21	average	average	ADJ
easat-10684	489	22	increase	increase	NOUN
easat-10684	489	23	of	of	ADP
easat-10684	489	24	8.54	8.54	NUM
easat-10684	489	25	%	%	NOUN
easat-10684	489	26	in	in	ADP
easat-10684	489	27	accuracy	accuracy	NOUN
easat-10684	489	28	compared	compare	VERB
easat-10684	489	29	to	to	ADP
easat-10684	489	30	the	the	DET
easat-10684	489	31	benchmark	benchmark	NOUN
easat-10684	489	32	approach	approach	NOUN
easat-10684	489	33	.	.	PUNCT
easat-10684	490	1	pandey	pandey	PROPN
easat-10684	490	2	et	et	PROPN
easat-10684	490	3	al	al	PROPN
easat-10684	490	4	.	.	PUNCT
easat-10684	491	1	[	[	X
easat-10684	491	2	6	6	NUM
easat-10684	491	3	]	]	PUNCT
easat-10684	491	4	proposed	propose	VERB
easat-10684	491	5	a	a	DET
easat-10684	491	6	similar	similar	ADJ
easat-10684	491	7	machine	machine	NOUN
easat-10684	491	8	learning	learning	NOUN
easat-10684	491	9	model	model	NOUN
easat-10684	491	10	for	for	ADP
easat-10684	491	11	fake	fake	ADJ
easat-10684	491	12	news	news	NOUN
easat-10684	491	13	detection	detection	NOUN
easat-10684	491	14	,	,	PUNCT
easat-10684	491	15	but	but	CCONJ
easat-10684	491	16	the	the	DET
easat-10684	491	17	model	model	NOUN
easat-10684	491	18	's	's	PART
easat-10684	491	19	performance	performance	NOUN
easat-10684	491	20	was	be	AUX
easat-10684	491	21	not	not	PART
easat-10684	491	22	satisfactory	satisfactory	ADJ
easat-10684	491	23	,	,	PUNCT
easat-10684	491	24	while	while	SCONJ
easat-10684	491	25	jaybhaye	jaybhaye	NOUN
easat-10684	491	26	[	[	PUNCT
easat-10684	491	27	13	13	NUM
easat-10684	491	28	]	]	PUNCT
easat-10684	491	29	suffered	suffer	VERB
easat-10684	491	30	from	from	ADP
easat-10684	491	31	model	model	NOUN
easat-10684	491	32	overfitting	overfitting	NOUN
easat-10684	491	33	during	during	ADP
easat-10684	491	34	training	training	NOUN
easat-10684	491	35	and	and	CCONJ
easat-10684	491	36	poor	poor	ADJ
easat-10684	491	37	performance	performance	NOUN
easat-10684	491	38	on	on	ADP
easat-10684	491	39	testing	testing	NOUN
easat-10684	491	40	samples	sample	NOUN
easat-10684	491	41	.	.	PUNCT
easat-10684	492	1	this	this	DET
easat-10684	492	2	issue	issue	NOUN
easat-10684	492	3	is	be	AUX
easat-10684	492	4	addressed	address	VERB
easat-10684	492	5	by	by	ADP
easat-10684	492	6	enhancing	enhance	VERB
easat-10684	492	7	the	the	DET
easat-10684	492	8	performance	performance	NOUN
easat-10684	492	9	of	of	ADP
easat-10684	492	10	the	the	DET
easat-10684	492	11	machine	machine	NOUN
easat-10684	492	12	learning	learn	VERB
easat-10684	492	13	model	model	NOUN
easat-10684	492	14	using	use	VERB
easat-10684	492	15	bi	bi	NOUN
easat-10684	492	16	-	-	ADJ
easat-10684	492	17	lstm	lstm	ADJ
easat-10684	492	18	contextualized	contextualize	VERB
easat-10684	492	19	weights	weight	NOUN
easat-10684	492	20	in	in	ADP
easat-10684	492	21	training	train	VERB
easat-10684	492	22	the	the	DET
easat-10684	492	23	machine	machine	NOUN
easat-10684	492	24	learning	learn	VERB
easat-10684	492	25	algorithms	algorithm	NOUN
easat-10684	492	26	,	,	PUNCT
easat-10684	492	27	and	and	CCONJ
easat-10684	492	28	also	also	ADV
easat-10684	492	29	addressing	address	VERB
easat-10684	492	30	the	the	DET
easat-10684	492	31	overfitting	overfitte	VERB
easat-10684	492	32	issue	issue	NOUN
easat-10684	492	33	using	use	VERB
easat-10684	492	34	dropout	dropout	NOUN
easat-10684	492	35	techniques	technique	NOUN
easat-10684	492	36	to	to	PART
easat-10684	492	37	penalize	penalize	VERB
easat-10684	492	38	the	the	DET
easat-10684	492	39	bi	bi	ADJ
easat-10684	492	40	-	-	ADJ
easat-10684	492	41	lstm	lstm	ADJ
easat-10684	492	42	training	training	NOUN
easat-10684	492	43	weights	weight	NOUN
easat-10684	492	44	.	.	PUNCT
easat-10684	493	1	the	the	DET
easat-10684	493	2	performance	performance	NOUN
easat-10684	493	3	comparison	comparison	NOUN
easat-10684	493	4	between	between	ADP
easat-10684	493	5	the	the	DET
easat-10684	493	6	proposed	propose	VERB
easat-10684	493	7	bi	bi	NOUN
easat-10684	493	8	-	-	PUNCT
easat-10684	493	9	lstm-2	lstm-2	NOUN
easat-10684	493	10	-	-	PUNCT
easat-10684	493	11	ml	ml	NOUN
easat-10684	493	12	model	model	NOUN
easat-10684	493	13	and	and	CCONJ
easat-10684	493	14	the	the	DET
easat-10684	493	15	benchmark	benchmark	NOUN
easat-10684	493	16	ml	ml	NOUN
easat-10684	493	17	models	model	NOUN
easat-10684	493	18	shows	show	VERB
easat-10684	493	19	that	that	SCONJ
easat-10684	493	20	the	the	DET
easat-10684	493	21	bi	bi	NOUN
easat-10684	493	22	-	-	ADJ
easat-10684	493	23	lstm	lstm	ADJ
easat-10684	493	24	weights	weight	NOUN
easat-10684	493	25	significantly	significantly	ADV
easat-10684	493	26	enhance	enhance	VERB
easat-10684	493	27	the	the	DET
easat-10684	493	28	performance	performance	NOUN
easat-10684	493	29	accuracy	accuracy	NOUN
easat-10684	493	30	of	of	ADP
easat-10684	493	31	the	the	DET
easat-10684	493	32	ml	ml	PROPN
easat-10684	493	33	model	model	NOUN
easat-10684	493	34	used	use	VERB
easat-10684	493	35	for	for	ADP
easat-10684	493	36	building	build	VERB
easat-10684	493	37	bi	bi	NOUN
easat-10684	493	38	-	-	ADJ
easat-10684	493	39	lstm-2	lstm-2	NOUN
easat-10684	493	40	-	-	PUNCT
easat-10684	493	41	ml	ml	NOUN
easat-10684	493	42	models	model	NOUN
easat-10684	493	43	.	.	PUNCT
easat-10684	494	1	furthermore	furthermore	ADV
easat-10684	494	2	,	,	PUNCT
easat-10684	494	3	an	an	DET
easat-10684	494	4	extensive	extensive	ADJ
easat-10684	494	5	study	study	NOUN
easat-10684	494	6	can	can	AUX
easat-10684	494	7	be	be	AUX
easat-10684	494	8	conducted	conduct	VERB
easat-10684	494	9	on	on	ADP
easat-10684	494	10	other	other	ADJ
easat-10684	494	11	deep	deep	ADJ
easat-10684	494	12	learning	learning	NOUN
easat-10684	494	13	architectures	architecture	NOUN
easat-10684	494	14	or	or	CCONJ
easat-10684	494	15	transformer	transformer	NOUN
easat-10684	494	16	-	-	PUNCT
easat-10684	494	17	based	base	VERB
easat-10684	494	18	architectures	architecture	NOUN
easat-10684	494	19	to	to	PART
easat-10684	494	20	improve	improve	VERB
easat-10684	494	21	machine	machine	NOUN
easat-10684	494	22	learning	learning	NOUN
easat-10684	494	23	using	use	VERB
easat-10684	494	24	a	a	DET
easat-10684	494	25	transfer	transfer	NOUN
easat-10684	494	26	learning	learning	NOUN
easat-10684	494	27	approach	approach	NOUN
easat-10684	494	28	.	.	PUNCT
easat-10684	495	1	architectures	architecture	NOUN
easat-10684	495	2	such	such	ADJ
easat-10684	495	3	as	as	ADP
easat-10684	495	4	lstm	lstm	PROPN
easat-10684	495	5	,	,	PUNCT
easat-10684	495	6	gpt	gpt	NOUN
easat-10684	495	7	,	,	PUNCT
easat-10684	495	8	bert	bert	PROPN
easat-10684	495	9	,	,	PUNCT
easat-10684	495	10	albert	albert	PROPN
easat-10684	495	11	,	,	PUNCT
easat-10684	495	12	gru	gru	PROPN
easat-10684	495	13	,	,	PUNCT
easat-10684	495	14	and	and	CCONJ
easat-10684	495	15	similar	similar	ADJ
easat-10684	495	16	models	model	NOUN
easat-10684	495	17	can	can	AUX
easat-10684	495	18	be	be	AUX
easat-10684	495	19	adopted	adopt	VERB
easat-10684	495	20	.	.	PUNCT
easat-10684	496	1	this	this	PRON
easat-10684	496	2	will	will	AUX
easat-10684	496	3	reveal	reveal	VERB
easat-10684	496	4	further	further	ADJ
easat-10684	496	5	insights	insight	NOUN
easat-10684	496	6	and	and	CCONJ
easat-10684	496	7	confirm	confirm	VERB
easat-10684	496	8	the	the	DET
easat-10684	496	9	impact	impact	NOUN
easat-10684	496	10	of	of	ADP
easat-10684	496	11	deep	deep	ADJ
easat-10684	496	12	learning	learning	NOUN
easat-10684	496	13	models	model	NOUN
easat-10684	496	14	on	on	ADP
easat-10684	496	15	machine	machine	NOUN
easat-10684	496	16	learning	learn	VERB
easat-10684	496	17	performance	performance	NOUN
easat-10684	496	18	through	through	ADP
easat-10684	496	19	a	a	DET
easat-10684	496	20	transfer	transfer	NOUN
easat-10684	496	21	learning	learning	NOUN
easat-10684	496	22	approach	approach	NOUN
easat-10684	496	23	.	.	PUNCT
easat-10684	497	1	data	datum	NOUN
easat-10684	497	2	availability	availability	NOUN
easat-10684	497	3	statement	statement	NOUN
easat-10684	497	4	:	:	PUNCT
easat-10684	497	5	the	the	DET
easat-10684	497	6	data	datum	NOUN
easat-10684	497	7	that	that	PRON
easat-10684	497	8	support	support	VERB
easat-10684	497	9	the	the	DET
easat-10684	497	10	findings	finding	NOUN
easat-10684	497	11	of	of	ADP
easat-10684	497	12	this	this	DET
easat-10684	497	13	study	study	NOUN
easat-10684	497	14	are	be	AUX
easat-10684	497	15	openly	openly	ADV
easat-10684	497	16	available	available	ADJ
easat-10684	497	17	in	in	ADP
easat-10684	497	18	kaggle	kaggle	NOUN
easat-10684	497	19	at	at	ADP
easat-10684	497	20	https://github.com/georgemcintire/fake_real_news_dataset	https://github.com/georgemcintire/fake_real_news_dataset	PROPN
easat-10684	497	21	.	.	PUNCT
easat-10684	498	1	0.00	0.00	NUM
easat-10684	498	2	%	%	NOUN
easat-10684	498	3	10.00	10.00	NUM
easat-10684	498	4	%	%	NOUN
easat-10684	498	5	20.00	20.00	NUM
easat-10684	498	6	%	%	NOUN
easat-10684	498	7	30.00	30.00	NUM
easat-10684	498	8	%	%	NOUN
easat-10684	498	9	40.00	40.00	NUM
easat-10684	498	10	%	%	NOUN
easat-10684	498	11	50.00	50.00	NUM
easat-10684	498	12	%	%	NOUN
easat-10684	498	13	60.00	60.00	NUM
easat-10684	498	14	%	%	NOUN
easat-10684	498	15	70.00	70.00	NUM
easat-10684	498	16	%	%	NOUN
easat-10684	498	17	80.00	80.00	NUM
easat-10684	498	18	%	%	NOUN
easat-10684	498	19	90.00	90.00	NUM
easat-10684	498	20	%	%	NOUN
easat-10684	498	21	100.00	100.00	NUM
easat-10684	498	22	%	%	NOUN
easat-10684	498	23	logistic	logistic	ADJ
easat-10684	498	24	regression	regression	NOUN
easat-10684	498	25	k	k	X
easat-10684	498	26	-	-	PUNCT
easat-10684	498	27	nearest	near	ADJ
easat-10684	498	28	neighbor	neighbor	NOUN
easat-10684	498	29	naive	naive	ADJ
easat-10684	498	30	bayes	bayes	NOUN
easat-10684	498	31	svm	svm	VERB
easat-10684	498	32	decision	decision	NOUN
easat-10684	498	33	tree	tree	NOUN
easat-10684	498	34	proposed	propose	VERB
easat-10684	498	35	bi	bi	NOUN
easat-10684	498	36	-	-	ADJ
easat-10684	498	37	lstm-2	lstm-2	NOUN
easat-10684	498	38	-	-	PUNCT
easat-10684	498	39	ml	ml	NOUN
easat-10684	498	40	vs	vs	ADP
easat-10684	498	41	benchmark	benchmark	NOUN
easat-10684	498	42	models	model	NOUN
easat-10684	498	43	machine	machine	NOUN
easat-10684	498	44	learning	learning	NOUN
easat-10684	498	45	(	(	PUNCT
easat-10684	498	46	benchmark	benchmark	PROPN
easat-10684	498	47	)	)	PUNCT
easat-10684	498	48	pandey	pandey	PROPN
easat-10684	498	49	et	et	PROPN
easat-10684	498	50	al	al	PROPN
easat-10684	498	51	.	.	PROPN
easat-10684	498	52	,	,	PUNCT
easat-10684	498	53	(	(	PUNCT
easat-10684	498	54	2022	2022	NUM
easat-10684	498	55	)	)	PUNCT
easat-10684	498	56	(	(	PUNCT
easat-10684	498	57	jaybhaye	jaybhaye	NOUN
easat-10684	498	58	et	et	PROPN
easat-10684	498	59	al	al	PROPN
easat-10684	498	60	.	.	PROPN
easat-10684	498	61	,	,	PUNCT
easat-10684	498	62	2023	2023	NUM
easat-10684	498	63	)	)	PUNCT
easat-10684	498	64	proposed	propose	VERB
easat-10684	498	65	bi	bi	ADJ
easat-10684	498	66	-	-	ADJ
easat-10684	498	67	lstm	lstm	ADJ
easat-10684	498	68	transfer	transfer	NOUN
easat-10684	498	69	learning-2	learning-2	NOUN
easat-10684	498	70	-	-	PUNCT
easat-10684	498	71	ml	ml	NOUN
easat-10684	498	72	https://github.com/georgemcintire/fake_real_news_dataset	https://github.com/georgemcintire/fake_real_news_dataset	X
easat-10684	498	73	1538	1538	NUM
easat-10684	498	74	edelweiss	edelweiss	PROPN
easat-10684	498	75	applied	apply	VERB
easat-10684	498	76	science	science	NOUN
easat-10684	498	77	and	and	CCONJ
easat-10684	498	78	technology	technology	NOUN
easat-10684	498	79	issn	issn	PROPN
easat-10684	498	80	:	:	PUNCT
easat-10684	498	81	2576	2576	NUM
easat-10684	498	82	-	-	SYM
easat-10684	498	83	8484	8484	NUM
easat-10684	498	84	vol	vol	NOUN
easat-10684	498	85	.	.	PROPN
easat-10684	499	1	9	9	NUM
easat-10684	499	2	,	,	PUNCT
easat-10684	499	3	no	no	INTJ
easat-10684	499	4	.	.	NOUN
easat-10684	499	5	10	10	NUM
easat-10684	499	6	:	:	PUNCT
easat-10684	499	7	1515	1515	NUM
easat-10684	499	8	-	-	SYM
easat-10684	499	9	1540	1540	NUM
easat-10684	499	10	,	,	PUNCT
easat-10684	499	11	2025	2025	NUM
easat-10684	499	12	doi	doi	NOUN
easat-10684	499	13	:	:	PUNCT
easat-10684	499	14	10.55214/2576	10.55214/2576	NUM
easat-10684	499	15	-	-	SYM
easat-10684	499	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	500	1	©	©	PROPN
easat-10684	500	2	2025	2025	NUM
easat-10684	500	3	by	by	ADP
easat-10684	500	4	the	the	DET
easat-10684	500	5	authors	author	NOUN
easat-10684	500	6	;	;	PUNCT
easat-10684	500	7	licensee	licensee	PROPN
easat-10684	500	8	learning	learn	VERB
easat-10684	500	9	gate	gate	PROPN
easat-10684	500	10	declaration	declaration	NOUN
easat-10684	500	11	of	of	ADP
easat-10684	500	12	conflict	conflict	NOUN
easat-10684	500	13	of	of	ADP
easat-10684	500	14	interest	interest	NOUN
easat-10684	500	15	:	:	PUNCT
easat-10684	500	16	all	all	DET
easat-10684	500	17	the	the	DET
easat-10684	500	18	authors	author	NOUN
easat-10684	500	19	wish	wish	VERB
easat-10684	500	20	to	to	PART
easat-10684	500	21	declare	declare	VERB
easat-10684	500	22	that	that	SCONJ
easat-10684	500	23	no	no	DET
easat-10684	500	24	conflict	conflict	NOUN
easat-10684	500	25	of	of	ADP
easat-10684	500	26	interest	interest	NOUN
easat-10684	500	27	exists	exist	VERB
easat-10684	500	28	in	in	ADP
easat-10684	500	29	writing	writing	NOUN
easat-10684	500	30	and	and	CCONJ
easat-10684	500	31	submitting	submit	VERB
easat-10684	500	32	this	this	DET
easat-10684	500	33	manuscript	manuscript	NOUN
easat-10684	500	34	.	.	PUNCT
easat-10684	501	1	all	all	DET
easat-10684	501	2	authors	author	NOUN
easat-10684	501	3	agree	agree	VERB
easat-10684	501	4	to	to	PART
easat-10684	501	5	submit	submit	VERB
easat-10684	501	6	this	this	DET
easat-10684	501	7	manuscript	manuscript	NOUN
easat-10684	501	8	to	to	ADP
easat-10684	501	9	a	a	DET
easat-10684	501	10	reputable	reputable	ADJ
easat-10684	501	11	journal	journal	NOUN
easat-10684	501	12	outlet	outlet	PROPN
easat-10684	501	13	.	.	PUNCT
easat-10684	502	1	declaration	declaration	NOUN
easat-10684	502	2	statement	statement	NOUN
easat-10684	502	3	:	:	PUNCT
easat-10684	502	4	i.	i.	NOUN
easat-10684	502	5	all	all	DET
easat-10684	502	6	authors	author	NOUN
easat-10684	502	7	agree	agree	VERB
easat-10684	502	8	to	to	PART
easat-10684	502	9	be	be	AUX
easat-10684	502	10	accountable	accountable	ADJ
easat-10684	502	11	for	for	ADP
easat-10684	502	12	the	the	DET
easat-10684	502	13	content	content	NOUN
easat-10684	502	14	and	and	CCONJ
easat-10684	502	15	conclusions	conclusion	NOUN
easat-10684	502	16	of	of	ADP
easat-10684	502	17	the	the	DET
easat-10684	502	18	article	article	NOUN
easat-10684	502	19	.	.	PUNCT
easat-10684	503	1	ii	ii	PROPN
easat-10684	503	2	.	.	PUNCT
easat-10684	504	1	no	no	DET
easat-10684	504	2	third	third	ADJ
easat-10684	504	3	-	-	PUNCT
easat-10684	504	4	party	party	NOUN
easat-10684	504	5	services	service	NOUN
easat-10684	504	6	were	be	AUX
easat-10684	504	7	involved	involve	VERB
easat-10684	504	8	in	in	ADP
easat-10684	504	9	the	the	DET
easat-10684	504	10	research	research	NOUN
easat-10684	504	11	or	or	CCONJ
easat-10684	504	12	manuscript	manuscript	NOUN
easat-10684	504	13	preparation	preparation	NOUN
easat-10684	504	14	.	.	PUNCT
easat-10684	505	1	iii	iii	X
easat-10684	505	2	.	.	PUNCT
easat-10684	506	1	all	all	PRON
easat-10684	506	2	contributing	contribute	VERB
easat-10684	506	3	authors	author	NOUN
easat-10684	506	4	of	of	ADP
easat-10684	506	5	this	this	DET
easat-10684	506	6	manuscript	manuscript	NOUN
easat-10684	506	7	are	be	AUX
easat-10684	506	8	properly	properly	ADV
easat-10684	506	9	listed	list	VERB
easat-10684	506	10	as	as	ADP
easat-10684	506	11	authors	author	NOUN
easat-10684	506	12	and	and	CCONJ
easat-10684	506	13	have	have	AUX
easat-10684	506	14	been	be	AUX
easat-10684	506	15	acknowledged	acknowledge	VERB
easat-10684	506	16	.	.	PUNCT
easat-10684	507	1	iv	iv	X
easat-10684	507	2	.	.	PUNCT
easat-10684	508	1	the	the	DET
easat-10684	508	2	authors	author	NOUN
easat-10684	508	3	declared	declare	VERB
easat-10684	508	4	that	that	SCONJ
easat-10684	508	5	no	no	DET
easat-10684	508	6	ai	ai	NOUN
easat-10684	508	7	software	software	NOUN
easat-10684	508	8	has	have	AUX
easat-10684	508	9	been	be	AUX
easat-10684	508	10	used	use	VERB
easat-10684	508	11	to	to	PART
easat-10684	508	12	prepare	prepare	VERB
easat-10684	508	13	the	the	DET
easat-10684	508	14	manuscript	manuscript	NOUN
easat-10684	508	15	.	.	PUNCT
easat-10684	509	1	v.	v.	ADP
easat-10684	509	2	the	the	DET
easat-10684	509	3	authors	author	NOUN
easat-10684	509	4	also	also	ADV
easat-10684	509	5	declared	declare	VERB
easat-10684	509	6	that	that	SCONJ
easat-10684	509	7	all	all	DET
easat-10684	509	8	the	the	DET
easat-10684	509	9	images	image	NOUN
easat-10684	509	10	used	use	VERB
easat-10684	509	11	are	be	AUX
easat-10684	509	12	original	original	ADJ
easat-10684	509	13	,	,	PUNCT
easat-10684	509	14	and	and	CCONJ
easat-10684	509	15	all	all	DET
easat-10684	509	16	the	the	DET
easat-10684	509	17	images	image	NOUN
easat-10684	509	18	were	be	AUX
easat-10684	509	19	obtained	obtain	VERB
easat-10684	509	20	from	from	ADP
easat-10684	509	21	our	our	PRON
easat-10684	509	22	implementation	implementation	NOUN
easat-10684	509	23	.	.	PUNCT
easat-10684	510	1	transparency	transparency	NOUN
easat-10684	510	2	:	:	PUNCT
easat-10684	510	3	the	the	DET
easat-10684	510	4	authors	author	NOUN
easat-10684	510	5	confirm	confirm	VERB
easat-10684	510	6	that	that	SCONJ
easat-10684	510	7	the	the	DET
easat-10684	510	8	manuscript	manuscript	NOUN
easat-10684	510	9	is	be	AUX
easat-10684	510	10	an	an	DET
easat-10684	510	11	honest	honest	ADJ
easat-10684	510	12	,	,	PUNCT
easat-10684	510	13	accurate	accurate	ADJ
easat-10684	510	14	,	,	PUNCT
easat-10684	510	15	and	and	CCONJ
easat-10684	510	16	transparent	transparent	ADJ
easat-10684	510	17	account	account	NOUN
easat-10684	510	18	of	of	ADP
easat-10684	510	19	the	the	DET
easat-10684	510	20	study	study	NOUN
easat-10684	510	21	;	;	PUNCT
easat-10684	510	22	that	that	SCONJ
easat-10684	510	23	no	no	DET
easat-10684	510	24	vital	vital	ADJ
easat-10684	510	25	features	feature	NOUN
easat-10684	510	26	of	of	ADP
easat-10684	510	27	the	the	DET
easat-10684	510	28	study	study	NOUN
easat-10684	510	29	have	have	AUX
easat-10684	510	30	been	be	AUX
easat-10684	510	31	omitted	omit	VERB
easat-10684	510	32	;	;	PUNCT
easat-10684	510	33	and	and	CCONJ
easat-10684	510	34	that	that	SCONJ
easat-10684	510	35	any	any	DET
easat-10684	510	36	discrepancies	discrepancy	NOUN
easat-10684	510	37	from	from	ADP
easat-10684	510	38	the	the	DET
easat-10684	510	39	study	study	NOUN
easat-10684	510	40	as	as	SCONJ
easat-10684	510	41	planned	plan	VERB
easat-10684	510	42	have	have	AUX
easat-10684	510	43	been	be	AUX
easat-10684	510	44	explained	explain	VERB
easat-10684	510	45	.	.	PUNCT
easat-10684	511	1	this	this	DET
easat-10684	511	2	study	study	NOUN
easat-10684	511	3	followed	follow	VERB
easat-10684	511	4	all	all	DET
easat-10684	511	5	ethical	ethical	ADJ
easat-10684	511	6	practices	practice	NOUN
easat-10684	511	7	during	during	ADP
easat-10684	511	8	writing	writing	NOUN
easat-10684	511	9	.	.	PUNCT
easat-10684	512	1	acknowledgements	acknowledgement	NOUN
easat-10684	512	2	:	:	PUNCT
easat-10684	512	3	all	all	DET
easat-10684	512	4	the	the	DET
easat-10684	512	5	authors	author	NOUN
easat-10684	512	6	acknowledge	acknowledge	VERB
easat-10684	512	7	that	that	SCONJ
easat-10684	512	8	the	the	DET
easat-10684	512	9	data	datum	NOUN
easat-10684	512	10	used	use	VERB
easat-10684	512	11	in	in	ADP
easat-10684	512	12	this	this	DET
easat-10684	512	13	study	study	NOUN
easat-10684	512	14	were	be	AUX
easat-10684	512	15	obtained	obtain	VERB
easat-10684	512	16	from	from	ADP
easat-10684	512	17	an	an	DET
easat-10684	512	18	online	online	ADJ
easat-10684	512	19	repository	repository	NOUN
easat-10684	512	20	as	as	SCONJ
easat-10684	512	21	provided	provide	VERB
easat-10684	512	22	in	in	ADP
easat-10684	512	23	the	the	DET
easat-10684	512	24	data	datum	NOUN
easat-10684	512	25	availability	availability	NOUN
easat-10684	512	26	statement	statement	NOUN
easat-10684	512	27	.	.	PUNCT
easat-10684	513	1	copyright	copyright	NOUN
easat-10684	513	2	:	:	PUNCT
easat-10684	513	3	©	©	PROPN
easat-10684	513	4	2025	2025	NUM
easat-10684	513	5	by	by	ADP
easat-10684	513	6	the	the	DET
easat-10684	513	7	authors	author	NOUN
easat-10684	513	8	.	.	PUNCT
easat-10684	514	1	this	this	DET
easat-10684	514	2	article	article	NOUN
easat-10684	514	3	is	be	AUX
easat-10684	514	4	an	an	DET
easat-10684	514	5	open	open	ADJ
easat-10684	514	6	access	access	NOUN
easat-10684	514	7	article	article	NOUN
easat-10684	514	8	distributed	distribute	VERB
easat-10684	514	9	under	under	ADP
easat-10684	514	10	the	the	DET
easat-10684	514	11	terms	term	NOUN
easat-10684	514	12	and	and	CCONJ
easat-10684	514	13	conditions	condition	NOUN
easat-10684	514	14	of	of	ADP
easat-10684	514	15	the	the	DET
easat-10684	514	16	creative	creative	ADJ
easat-10684	514	17	commons	common	NOUN
easat-10684	514	18	attribution	attribution	NOUN
easat-10684	514	19	(	(	PUNCT
easat-10684	514	20	cc	cc	NOUN
easat-10684	514	21	by	by	ADP
easat-10684	514	22	)	)	PUNCT
easat-10684	514	23	license	license	NOUN
easat-10684	514	24	(	(	PUNCT
easat-10684	514	25	https://creativecommons.org/licenses/by/4.0/	https://creativecommons.org/licenses/by/4.0/	PROPN
easat-10684	514	26	)	)	PUNCT
easat-10684	514	27	.	.	PUNCT
easat-10684	515	1	references	reference	NOUN
easat-10684	515	2	[	[	X
easat-10684	515	3	1	1	NUM
easat-10684	515	4	]	]	PUNCT
easat-10684	515	5	c.	c.	NOUN
easat-10684	515	6	song	song	NOUN
easat-10684	515	7	,	,	PUNCT
easat-10684	515	8	n.	n.	NOUN
easat-10684	515	9	ning	ning	PROPN
easat-10684	515	10	,	,	PUNCT
easat-10684	515	11	y.	y.	PROPN
easat-10684	515	12	zhang	zhang	PROPN
easat-10684	515	13	,	,	PUNCT
easat-10684	515	14	and	and	CCONJ
easat-10684	515	15	b.	b.	PROPN
easat-10684	515	16	wu	wu	PROPN
easat-10684	515	17	,	,	PUNCT
easat-10684	515	18	"	"	PUNCT
easat-10684	515	19	a	a	DET
easat-10684	515	20	multimodal	multimodal	ADJ
easat-10684	515	21	fake	fake	ADJ
easat-10684	515	22	news	news	NOUN
easat-10684	515	23	detection	detection	NOUN
easat-10684	515	24	model	model	NOUN
easat-10684	515	25	based	base	VERB
easat-10684	515	26	on	on	ADP
easat-10684	515	27	crossmodal	crossmodal	ADJ
easat-10684	515	28	attention	attention	NOUN
easat-10684	515	29	residual	residual	ADJ
easat-10684	515	30	and	and	CCONJ
easat-10684	515	31	multichannel	multichannel	ADJ
easat-10684	515	32	convolutional	convolutional	ADJ
easat-10684	515	33	neural	neural	ADJ
easat-10684	515	34	networks	network	NOUN
easat-10684	515	35	,	,	PUNCT
easat-10684	515	36	"	"	PUNCT
easat-10684	515	37	information	information	NOUN
easat-10684	515	38	processing	processing	NOUN
easat-10684	515	39	&	&	CCONJ
easat-10684	515	40	management	management	NOUN
easat-10684	515	41	,	,	PUNCT
easat-10684	515	42	vol	vol	NOUN
easat-10684	515	43	.	.	PROPN
easat-10684	516	1	58	58	NUM
easat-10684	516	2	,	,	PUNCT
easat-10684	516	3	no	no	INTJ
easat-10684	516	4	.	.	NOUN
easat-10684	516	5	1	1	NUM
easat-10684	516	6	,	,	PUNCT
easat-10684	516	7	p.	p.	NOUN
easat-10684	516	8	102437	102437	NUM
easat-10684	516	9	,	,	PUNCT
easat-10684	516	10	2021	2021	NUM
easat-10684	516	11	.	.	PUNCT
easat-10684	517	1	https://doi.org/10.1016/j.ipm.2020.102437	https://doi.org/10.1016/j.ipm.2020.102437	PUNCT
easat-10684	518	1	[	[	X
easat-10684	518	2	2	2	NUM
easat-10684	518	3	]	]	PUNCT
easat-10684	518	4	g.	g.	PROPN
easat-10684	518	5	e.	e.	PROPN
easat-10684	518	6	r.	r.	PROPN
easat-10684	518	7	agudelo	agudelo	ADV
easat-10684	518	8	,	,	PUNCT
easat-10684	518	9	o.	o.	PROPN
easat-10684	518	10	j.	j.	PROPN
easat-10684	518	11	s.	s.	PROPN
easat-10684	518	12	parra	parra	PROPN
easat-10684	518	13	,	,	PUNCT
easat-10684	518	14	and	and	CCONJ
easat-10684	518	15	j.	j.	PROPN
easat-10684	518	16	b.	b.	PROPN
easat-10684	518	17	velandia	velandia	PROPN
easat-10684	518	18	,	,	PUNCT
easat-10684	518	19	"	"	PUNCT
easat-10684	518	20	raising	raise	VERB
easat-10684	518	21	a	a	DET
easat-10684	518	22	model	model	NOUN
easat-10684	518	23	for	for	ADP
easat-10684	518	24	fake	fake	ADJ
easat-10684	518	25	news	news	NOUN
easat-10684	518	26	detection	detection	NOUN
easat-10684	518	27	using	use	VERB
easat-10684	518	28	machine	machine	NOUN
easat-10684	518	29	learning	learning	NOUN
easat-10684	518	30	in	in	ADP
easat-10684	518	31	python	python	NOUN
easat-10684	518	32	,	,	PUNCT
easat-10684	518	33	"	"	PUNCT
easat-10684	518	34	in	in	ADP
easat-10684	518	35	conference	conference	NOUN
easat-10684	518	36	on	on	ADP
easat-10684	518	37	e	e	NOUN
easat-10684	518	38	-	-	NOUN
easat-10684	518	39	business	business	NOUN
easat-10684	518	40	,	,	PUNCT
easat-10684	518	41	e	e	NOUN
easat-10684	518	42	-	-	NOUN
easat-10684	518	43	services	service	NOUN
easat-10684	518	44	and	and	CCONJ
easat-10684	518	45	e	e	NOUN
easat-10684	518	46	-	-	NOUN
easat-10684	518	47	society	society	NOUN
easat-10684	518	48	(	(	PUNCT
easat-10684	518	49	pp	pp	ADJ
easat-10684	518	50	.	.	PUNCT
easat-10684	519	1	596	596	NUM
easat-10684	519	2	-	-	NUM
easat-10684	519	3	604	604	NUM
easat-10684	519	4	)	)	PUNCT
easat-10684	519	5	.	.	PUNCT
easat-10684	520	1	cham	cham	PROPN
easat-10684	520	2	:	:	PUNCT
easat-10684	520	3	springer	springer	NOUN
easat-10684	520	4	international	international	ADJ
easat-10684	520	5	publishing	publishing	NOUN
easat-10684	520	6	,	,	PUNCT
easat-10684	520	7	2018	2018	NUM
easat-10684	520	8	.	.	PUNCT
easat-10684	521	1	[	[	X
easat-10684	521	2	3	3	NUM
easat-10684	521	3	]	]	PUNCT
easat-10684	521	4	a.	a.	NOUN
easat-10684	521	5	jain	jain	PROPN
easat-10684	521	6	,	,	PUNCT
easat-10684	521	7	a.	a.	PROPN
easat-10684	521	8	shakya	shakya	PROPN
easat-10684	521	9	,	,	PUNCT
easat-10684	521	10	h.	h.	PROPN
easat-10684	521	11	khatter	khatter	PROPN
easat-10684	521	12	,	,	PUNCT
easat-10684	521	13	and	and	CCONJ
easat-10684	521	14	a.	a.	PROPN
easat-10684	521	15	k.	k.	PROPN
easat-10684	521	16	gupta	gupta	PROPN
easat-10684	521	17	,	,	PUNCT
easat-10684	521	18	"	"	PUNCT
easat-10684	521	19	a	a	DET
easat-10684	521	20	smart	smart	ADJ
easat-10684	521	21	system	system	NOUN
easat-10684	521	22	for	for	ADP
easat-10684	521	23	fake	fake	ADJ
easat-10684	521	24	news	news	NOUN
easat-10684	521	25	detection	detection	NOUN
easat-10684	521	26	using	use	VERB
easat-10684	521	27	machine	machine	NOUN
easat-10684	521	28	learning	learning	NOUN
easat-10684	521	29	,	,	PUNCT
easat-10684	521	30	"	"	PUNCT
easat-10684	521	31	in	in	ADP
easat-10684	521	32	2019	2019	NUM
easat-10684	521	33	international	international	ADJ
easat-10684	521	34	conference	conference	NOUN
easat-10684	521	35	on	on	ADP
easat-10684	521	36	issues	issue	NOUN
easat-10684	521	37	and	and	CCONJ
easat-10684	521	38	challenges	challenge	NOUN
easat-10684	521	39	in	in	ADP
easat-10684	521	40	intelligent	intelligent	ADJ
easat-10684	521	41	computing	computing	NOUN
easat-10684	521	42	techniques	technique	NOUN
easat-10684	521	43	(	(	PUNCT
easat-10684	521	44	icict	icict	NOUN
easat-10684	521	45	)	)	PUNCT
easat-10684	521	46	(	(	PUNCT
easat-10684	521	47	vol	vol	NOUN
easat-10684	521	48	.	.	NOUN
easat-10684	521	49	1	1	NUM
easat-10684	521	50	,	,	PUNCT
easat-10684	521	51	pp	pp	ADJ
easat-10684	521	52	.	.	PUNCT
easat-10684	522	1	1	1	NUM
easat-10684	522	2	-	-	SYM
easat-10684	522	3	4	4	NUM
easat-10684	522	4	)	)	PUNCT
easat-10684	522	5	.	.	PUNCT
easat-10684	523	1	ieee	ieee	PROPN
easat-10684	523	2	,	,	PUNCT
easat-10684	523	3	2019	2019	NUM
easat-10684	523	4	.	.	PUNCT
easat-10684	524	1	[	[	X
easat-10684	524	2	4	4	NUM
easat-10684	524	3	]	]	PUNCT
easat-10684	524	4	a.	a.	NOUN
easat-10684	524	5	m.	m.	NOUN
easat-10684	524	6	braşoveanu	braşoveanu	PROPN
easat-10684	524	7	and	and	CCONJ
easat-10684	524	8	r.	r.	PROPN
easat-10684	524	9	andonie	andonie	PROPN
easat-10684	524	10	,	,	PUNCT
easat-10684	524	11	"	"	PUNCT
easat-10684	524	12	semantic	semantic	ADJ
easat-10684	524	13	fake	fake	ADJ
easat-10684	524	14	news	news	NOUN
easat-10684	524	15	detection	detection	NOUN
easat-10684	524	16	:	:	PUNCT
easat-10684	524	17	a	a	DET
easat-10684	524	18	machine	machine	NOUN
easat-10684	524	19	learning	learn	VERB
easat-10684	524	20	perspective	perspective	NOUN
easat-10684	524	21	,	,	PUNCT
easat-10684	524	22	"	"	PUNCT
easat-10684	524	23	in	in	ADP
easat-10684	524	24	international	international	ADJ
easat-10684	524	25	work	work	NOUN
easat-10684	524	26	-	-	PUNCT
easat-10684	524	27	conference	conference	NOUN
easat-10684	524	28	on	on	ADP
easat-10684	524	29	artificial	artificial	ADJ
easat-10684	524	30	neural	neural	ADJ
easat-10684	524	31	networks	network	NOUN
easat-10684	524	32	(	(	PUNCT
easat-10684	524	33	pp	pp	ADV
easat-10684	524	34	.	.	PUNCT
easat-10684	524	35	656	656	NUM
easat-10684	524	36	-	-	SYM
easat-10684	524	37	667	667	NUM
easat-10684	524	38	)	)	PUNCT
easat-10684	524	39	.	.	PUNCT
easat-10684	525	1	cham	cham	PROPN
easat-10684	525	2	:	:	PUNCT
easat-10684	525	3	springer	springer	NOUN
easat-10684	525	4	international	international	ADJ
easat-10684	525	5	publishing	publishing	NOUN
easat-10684	525	6	,	,	PUNCT
easat-10684	525	7	2019	2019	NUM
easat-10684	525	8	.	.	PUNCT
easat-10684	526	1	[	[	X
easat-10684	526	2	5	5	X
easat-10684	526	3	]	]	PUNCT
easat-10684	526	4	e.	e.	PROPN
easat-10684	526	5	hossain	hossain	PROPN
easat-10684	526	6	,	,	PUNCT
easat-10684	526	7	m.	m.	PROPN
easat-10684	526	8	nadim	nadim	PROPN
easat-10684	526	9	kaysar	kaysar	PROPN
easat-10684	526	10	,	,	PUNCT
easat-10684	526	11	a.	a.	PROPN
easat-10684	526	12	z.	z.	PROPN
easat-10684	526	13	m.	m.	PROPN
easat-10684	526	14	jalal	jalal	PROPN
easat-10684	526	15	uddin	uddin	PROPN
easat-10684	526	16	joy	joy	PROPN
easat-10684	526	17	,	,	PUNCT
easat-10684	526	18	m.	m.	NOUN
easat-10684	526	19	mizanur	mizanur	PROPN
easat-10684	526	20	rahman	rahman	PROPN
easat-10684	526	21	,	,	PUNCT
easat-10684	526	22	and	and	CCONJ
easat-10684	526	23	w.	w.	PROPN
easat-10684	526	24	rahman	rahman	PROPN
easat-10684	526	25	,	,	PUNCT
easat-10684	526	26	"	"	PUNCT
easat-10684	526	27	a	a	DET
easat-10684	526	28	study	study	NOUN
easat-10684	526	29	towards	towards	ADP
easat-10684	526	30	bangla	bangla	PROPN
easat-10684	526	31	fake	fake	ADJ
easat-10684	526	32	news	news	NOUN
easat-10684	526	33	detection	detection	NOUN
easat-10684	526	34	using	use	VERB
easat-10684	526	35	machine	machine	NOUN
easat-10684	526	36	learning	learning	NOUN
easat-10684	526	37	and	and	CCONJ
easat-10684	526	38	deep	deep	ADJ
easat-10684	526	39	learning	learning	NOUN
easat-10684	526	40	,	,	PUNCT
easat-10684	526	41	"	"	PUNCT
easat-10684	526	42	in	in	ADP
easat-10684	526	43	sentimental	sentimental	ADJ
easat-10684	526	44	analysis	analysis	NOUN
easat-10684	526	45	and	and	CCONJ
easat-10684	526	46	deep	deep	ADJ
easat-10684	526	47	learning	learning	NOUN
easat-10684	526	48	:	:	PUNCT
easat-10684	526	49	proceedings	proceeding	NOUN
easat-10684	526	50	of	of	ADP
easat-10684	526	51	icsadl	icsadl	NOUN
easat-10684	526	52	2021	2021	NUM
easat-10684	526	53	(	(	PUNCT
easat-10684	526	54	pp	pp	ADJ
easat-10684	526	55	.	.	PUNCT
easat-10684	527	1	79	79	NUM
easat-10684	527	2	-	-	SYM
easat-10684	527	3	95	95	NUM
easat-10684	527	4	)	)	PUNCT
easat-10684	527	5	.	.	PUNCT
easat-10684	528	1	singapore	singapore	PROPN
easat-10684	528	2	:	:	PUNCT
easat-10684	528	3	springer	springer	PROPN
easat-10684	528	4	singapore	singapore	PROPN
easat-10684	528	5	,	,	PUNCT
easat-10684	528	6	2021	2021	NUM
easat-10684	528	7	.	.	PUNCT
easat-10684	529	1	[	[	X
easat-10684	529	2	6	6	NUM
easat-10684	529	3	]	]	PUNCT
easat-10684	529	4	s.	s.	PROPN
easat-10684	529	5	pandey	pandey	PROPN
easat-10684	529	6	,	,	PUNCT
easat-10684	529	7	s.	s.	PROPN
easat-10684	529	8	prabhakaran	prabhakaran	PROPN
easat-10684	529	9	,	,	PUNCT
easat-10684	529	10	n.	n.	PROPN
easat-10684	529	11	s.	s.	PROPN
easat-10684	529	12	reddy	reddy	PROPN
easat-10684	529	13	,	,	PUNCT
easat-10684	529	14	and	and	CCONJ
easat-10684	529	15	d.	d.	PROPN
easat-10684	529	16	acharya	acharya	PROPN
easat-10684	529	17	,	,	PUNCT
easat-10684	529	18	"	"	PUNCT
easat-10684	529	19	fake	fake	ADJ
easat-10684	529	20	news	news	NOUN
easat-10684	529	21	detection	detection	NOUN
easat-10684	529	22	from	from	ADP
easat-10684	529	23	online	online	ADJ
easat-10684	529	24	media	medium	NOUN
easat-10684	529	25	using	use	VERB
easat-10684	529	26	machine	machine	NOUN
easat-10684	529	27	learning	learn	VERB
easat-10684	529	28	classifiers	classifier	NOUN
easat-10684	529	29	,	,	PUNCT
easat-10684	529	30	"	"	PUNCT
easat-10684	529	31	in	in	ADP
easat-10684	529	32	journal	journal	NOUN
easat-10684	529	33	of	of	ADP
easat-10684	529	34	physics	physics	PROPN
easat-10684	529	35	:	:	PUNCT
easat-10684	529	36	conference	conference	NOUN
easat-10684	529	37	series	series	NOUN
easat-10684	529	38	(	(	PUNCT
easat-10684	529	39	vol	vol	NOUN
easat-10684	529	40	.	.	PUNCT
easat-10684	529	41	2161	2161	NUM
easat-10684	529	42	,	,	PUNCT
easat-10684	529	43	no	no	INTJ
easat-10684	529	44	.	.	NOUN
easat-10684	529	45	1	1	NUM
easat-10684	529	46	,	,	PUNCT
easat-10684	529	47	p.	p.	NOUN
easat-10684	529	48	012027	012027	NUM
easat-10684	529	49	)	)	PUNCT
easat-10684	529	50	.	.	PUNCT
easat-10684	530	1	iop	iop	PROPN
easat-10684	530	2	publishing	publishing	NOUN
easat-10684	530	3	,	,	PUNCT
easat-10684	530	4	2022	2022	NUM
easat-10684	530	5	.	.	PUNCT
easat-10684	531	1	[	[	X
easat-10684	531	2	7	7	X
easat-10684	531	3	]	]	X
easat-10684	531	4	c.	c.	PROPN
easat-10684	531	5	zhang	zhang	PROPN
easat-10684	531	6	and	and	CCONJ
easat-10684	531	7	y.	y.	PROPN
easat-10684	531	8	lu	lu	PROPN
easat-10684	531	9	,	,	PUNCT
easat-10684	531	10	"	"	PUNCT
easat-10684	531	11	study	study	NOUN
easat-10684	531	12	on	on	ADP
easat-10684	531	13	artificial	artificial	ADJ
easat-10684	531	14	intelligence	intelligence	NOUN
easat-10684	531	15	:	:	PUNCT
easat-10684	531	16	the	the	DET
easat-10684	531	17	state	state	NOUN
easat-10684	531	18	of	of	ADP
easat-10684	531	19	the	the	DET
easat-10684	531	20	art	art	NOUN
easat-10684	531	21	and	and	CCONJ
easat-10684	531	22	future	future	ADJ
easat-10684	531	23	prospects	prospect	NOUN
easat-10684	531	24	,	,	PUNCT
easat-10684	531	25	"	"	PUNCT
easat-10684	531	26	journal	journal	NOUN
easat-10684	531	27	of	of	ADP
easat-10684	531	28	industrial	industrial	ADJ
easat-10684	531	29	information	information	NOUN
easat-10684	531	30	integration	integration	NOUN
easat-10684	531	31	,	,	PUNCT
easat-10684	531	32	vol	vol	NOUN
easat-10684	531	33	.	.	PROPN
easat-10684	531	34	23	23	NUM
easat-10684	531	35	,	,	PUNCT
easat-10684	531	36	p.	p.	NOUN
easat-10684	531	37	100224	100224	NUM
easat-10684	531	38	,	,	PUNCT
easat-10684	531	39	2021	2021	NUM
easat-10684	531	40	.	.	PUNCT
easat-10684	532	1	[	[	X
easat-10684	532	2	8	8	NUM
easat-10684	532	3	]	]	X
easat-10684	532	4	p.	p.	NOUN
easat-10684	532	5	goyal	goyal	PROPN
easat-10684	532	6	,	,	PUNCT
easat-10684	532	7	s.	s.	PROPN
easat-10684	532	8	pandey	pandey	PROPN
easat-10684	532	9	,	,	PUNCT
easat-10684	532	10	and	and	CCONJ
easat-10684	532	11	k.	k.	PROPN
easat-10684	532	12	jain	jain	PROPN
easat-10684	532	13	,	,	PUNCT
easat-10684	532	14	deep	deep	ADJ
easat-10684	532	15	learning	learning	NOUN
easat-10684	532	16	for	for	ADP
easat-10684	532	17	natural	natural	ADJ
easat-10684	532	18	language	language	NOUN
easat-10684	532	19	processing	processing	NOUN
easat-10684	532	20	.	.	PUNCT
easat-10684	533	1	berkeley	berkeley	PROPN
easat-10684	533	2	,	,	PUNCT
easat-10684	533	3	ca	ca	NOUN
easat-10684	533	4	:	:	PUNCT
easat-10684	533	5	apress	apress	NOUN
easat-10684	533	6	,	,	PUNCT
easat-10684	533	7	2018	2018	NUM
easat-10684	533	8	.	.	PUNCT
easat-10684	534	1	[	[	X
easat-10684	534	2	9	9	NUM
easat-10684	534	3	]	]	PUNCT
easat-10684	534	4	d.	d.	PROPN
easat-10684	534	5	khurana	khurana	PROPN
easat-10684	534	6	,	,	PUNCT
easat-10684	534	7	a.	a.	PROPN
easat-10684	534	8	koli	koli	PROPN
easat-10684	534	9	,	,	PUNCT
easat-10684	534	10	k.	k.	PROPN
easat-10684	534	11	khatter	khatter	PROPN
easat-10684	534	12	,	,	PUNCT
easat-10684	534	13	and	and	CCONJ
easat-10684	534	14	s.	s.	PROPN
easat-10684	534	15	singh	singh	PROPN
easat-10684	534	16	,	,	PUNCT
easat-10684	534	17	"	"	PUNCT
easat-10684	534	18	natural	natural	ADJ
easat-10684	534	19	language	language	NOUN
easat-10684	534	20	processing	processing	NOUN
easat-10684	534	21	:	:	PUNCT
easat-10684	534	22	state	state	NOUN
easat-10684	534	23	of	of	ADP
easat-10684	534	24	the	the	DET
easat-10684	534	25	art	art	NOUN
easat-10684	534	26	,	,	PUNCT
easat-10684	534	27	current	current	ADJ
easat-10684	534	28	trends	trend	NOUN
easat-10684	534	29	and	and	CCONJ
easat-10684	534	30	challenges	challenge	NOUN
easat-10684	534	31	,	,	PUNCT
easat-10684	534	32	"	"	PUNCT
easat-10684	534	33	multimedia	multimedia	NOUN
easat-10684	534	34	tools	tool	NOUN
easat-10684	534	35	and	and	CCONJ
easat-10684	534	36	applications	application	NOUN
easat-10684	534	37	,	,	PUNCT
easat-10684	534	38	vol	vol	NOUN
easat-10684	534	39	.	.	PROPN
easat-10684	534	40	82	82	NUM
easat-10684	534	41	,	,	PUNCT
easat-10684	534	42	pp	pp	ADJ
easat-10684	534	43	.	.	PUNCT
easat-10684	535	1	3713	3713	NUM
easat-10684	535	2	-	-	SYM
easat-10684	535	3	3744	3744	NUM
easat-10684	535	4	,	,	PUNCT
easat-10684	535	5	2023	2023	NUM
easat-10684	535	6	.	.	PUNCT
easat-10684	536	1	https://doi.org/10.1007/s11042-02213428-4	https://doi.org/10.1007/s11042-02213428-4	NOUN
easat-10684	536	2	[	[	X
easat-10684	536	3	10	10	NUM
easat-10684	536	4	]	]	X
easat-10684	536	5	r.	r.	PROPN
easat-10684	536	6	patel	patel	PROPN
easat-10684	536	7	and	and	CCONJ
easat-10684	536	8	s.	s.	PROPN
easat-10684	536	9	patel	patel	PROPN
easat-10684	536	10	,	,	PUNCT
easat-10684	536	11	"	"	PUNCT
easat-10684	536	12	deep	deep	ADJ
easat-10684	536	13	learning	learning	NOUN
easat-10684	536	14	for	for	ADP
easat-10684	536	15	natural	natural	ADJ
easat-10684	536	16	language	language	NOUN
easat-10684	536	17	processing	processing	NOUN
easat-10684	536	18	,	,	PUNCT
easat-10684	536	19	"	"	PUNCT
easat-10684	536	20	presented	present	VERB
easat-10684	536	21	at	at	ADP
easat-10684	536	22	the	the	DET
easat-10684	536	23	information	information	NOUN
easat-10684	536	24	and	and	CCONJ
easat-10684	536	25	communication	communication	NOUN
easat-10684	536	26	technology	technology	NOUN
easat-10684	536	27	for	for	ADP
easat-10684	536	28	competitive	competitive	ADJ
easat-10684	536	29	strategies	strategy	NOUN
easat-10684	536	30	(	(	PUNCT
easat-10684	536	31	ictcs	ictcs	NOUN
easat-10684	536	32	2020	2020	NUM
easat-10684	536	33	)	)	PUNCT
easat-10684	536	34	intelligent	intelligent	ADJ
easat-10684	536	35	strategies	strategy	NOUN
easat-10684	536	36	for	for	ADP
easat-10684	536	37	ict	ict	PROPN
easat-10684	536	38	(	(	PUNCT
easat-10684	536	39	pp	pp	ADJ
easat-10684	536	40	.	.	PUNCT
easat-10684	537	1	523	523	NUM
easat-10684	537	2	-	-	SYM
easat-10684	537	3	533	533	NUM
easat-10684	537	4	)	)	PUNCT
easat-10684	537	5	.	.	PUNCT
easat-10684	538	1	singapore	singapore	PROPN
easat-10684	538	2	:	:	PUNCT
easat-10684	538	3	springer	springer	NOUN
easat-10684	538	4	nature	nature	PROPN
easat-10684	538	5	singapore	singapore	PROPN
easat-10684	538	6	,	,	PUNCT
easat-10684	538	7	2021	2021	NUM
easat-10684	538	8	.	.	PUNCT
easat-10684	539	1	https://creativecommons.org/licenses/by/4.0/	https://creativecommons.org/licenses/by/4.0/	PROPN
easat-10684	539	2	https://doi.org/10.1016/j.ipm.2020.102437	https://doi.org/10.1016/j.ipm.2020.102437	PROPN
easat-10684	539	3	https://doi.org/10.1007/s11042-022-13428-4	https://doi.org/10.1007/s11042-022-13428-4	NUM
easat-10684	539	4	https://doi.org/10.1007/s11042-022-13428-4	https://doi.org/10.1007/s11042-022-13428-4	NUM
easat-10684	539	5	1539	1539	NUM
easat-10684	539	6	edelweiss	edelweiss	PROPN
easat-10684	539	7	applied	apply	VERB
easat-10684	539	8	science	science	NOUN
easat-10684	539	9	and	and	CCONJ
easat-10684	539	10	technology	technology	NOUN
easat-10684	539	11	issn	issn	PROPN
easat-10684	539	12	:	:	PUNCT
easat-10684	539	13	2576	2576	NUM
easat-10684	539	14	-	-	SYM
easat-10684	539	15	8484	8484	NUM
easat-10684	539	16	vol	vol	NOUN
easat-10684	539	17	.	.	PROPN
easat-10684	540	1	9	9	NUM
easat-10684	540	2	,	,	PUNCT
easat-10684	540	3	no	no	INTJ
easat-10684	540	4	.	.	NOUN
easat-10684	540	5	10	10	NUM
easat-10684	540	6	:	:	PUNCT
easat-10684	540	7	1515	1515	NUM
easat-10684	540	8	-	-	SYM
easat-10684	540	9	1540	1540	NUM
easat-10684	540	10	,	,	PUNCT
easat-10684	540	11	2025	2025	NUM
easat-10684	540	12	doi	doi	NOUN
easat-10684	540	13	:	:	PUNCT
easat-10684	540	14	10.55214/2576	10.55214/2576	NUM
easat-10684	540	15	-	-	SYM
easat-10684	540	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	541	1	©	©	PROPN
easat-10684	541	2	2025	2025	NUM
easat-10684	541	3	by	by	ADP
easat-10684	541	4	the	the	DET
easat-10684	541	5	authors	author	NOUN
easat-10684	541	6	;	;	PUNCT
easat-10684	541	7	licensee	licensee	PROPN
easat-10684	541	8	learning	learning	NOUN
easat-10684	541	9	gate	gate	NOUN
easat-10684	542	1	[	[	X
easat-10684	542	2	11	11	NUM
easat-10684	542	3	]	]	PUNCT
easat-10684	542	4	s.	s.	PROPN
easat-10684	542	5	kumar	kumar	PROPN
easat-10684	542	6	,	,	PUNCT
easat-10684	542	7	r.	r.	PROPN
easat-10684	542	8	asthana	asthana	PROPN
easat-10684	542	9	,	,	PUNCT
easat-10684	542	10	s.	s.	PROPN
easat-10684	542	11	upadhyay	upadhyay	PROPN
easat-10684	542	12	,	,	PUNCT
easat-10684	542	13	n.	n.	PROPN
easat-10684	542	14	upreti	upreti	PROPN
easat-10684	542	15	,	,	PUNCT
easat-10684	542	16	and	and	CCONJ
easat-10684	542	17	m.	m.	NOUN
easat-10684	542	18	akbar	akbar	NOUN
easat-10684	542	19	,	,	PUNCT
easat-10684	542	20	"	"	PUNCT
easat-10684	542	21	fake	fake	ADJ
easat-10684	542	22	news	news	NOUN
easat-10684	542	23	detection	detection	NOUN
easat-10684	542	24	using	use	VERB
easat-10684	542	25	deep	deep	ADJ
easat-10684	542	26	learning	learning	NOUN
easat-10684	542	27	models	model	NOUN
easat-10684	542	28	:	:	PUNCT
easat-10684	542	29	a	a	DET
easat-10684	542	30	novel	novel	ADJ
easat-10684	542	31	approach	approach	NOUN
easat-10684	542	32	,	,	PUNCT
easat-10684	542	33	"	"	PUNCT
easat-10684	542	34	transactions	transaction	NOUN
easat-10684	542	35	on	on	ADP
easat-10684	542	36	emerging	emerge	VERB
easat-10684	542	37	telecommunications	telecommunication	NOUN
easat-10684	542	38	technologies	technology	NOUN
easat-10684	542	39	,	,	PUNCT
easat-10684	542	40	vol	vol	NOUN
easat-10684	542	41	.	.	PROPN
easat-10684	542	42	31	31	NUM
easat-10684	542	43	,	,	PUNCT
easat-10684	542	44	no	no	INTJ
easat-10684	542	45	.	.	NOUN
easat-10684	542	46	2	2	NUM
easat-10684	542	47	,	,	PUNCT
easat-10684	542	48	p.	p.	NOUN
easat-10684	542	49	e3767	e3767	PROPN
easat-10684	542	50	,	,	PUNCT
easat-10684	542	51	2020	2020	NUM
easat-10684	542	52	.	.	PUNCT
easat-10684	543	1	https://doi.org/10.1002/ett.3767	https://doi.org/10.1002/ett.3767	PROPN
easat-10684	544	1	[	[	X
easat-10684	544	2	12	12	NUM
easat-10684	544	3	]	]	PUNCT
easat-10684	544	4	k.	k.	PROPN
easat-10684	544	5	r.	r.	PROPN
easat-10684	544	6	asish	asish	PROPN
easat-10684	544	7	,	,	PUNCT
easat-10684	544	8	a.	a.	PROPN
easat-10684	544	9	gupta	gupta	PROPN
easat-10684	544	10	,	,	PUNCT
easat-10684	544	11	a.	a.	PROPN
easat-10684	544	12	kumar	kumar	PROPN
easat-10684	544	13	,	,	PUNCT
easat-10684	544	14	a.	a.	PROPN
easat-10684	544	15	mason	mason	PROPN
easat-10684	544	16	,	,	PUNCT
easat-10684	544	17	m.	m.	PROPN
easat-10684	544	18	k.	k.	PROPN
easat-10684	544	19	enduri	enduri	PROPN
easat-10684	544	20	,	,	PUNCT
easat-10684	544	21	and	and	CCONJ
easat-10684	544	22	s.	s.	PROPN
easat-10684	544	23	anamalamudi	anamalamudi	PROPN
easat-10684	544	24	,	,	PUNCT
easat-10684	544	25	"	"	PUNCT
easat-10684	544	26	a	a	DET
easat-10684	544	27	tool	tool	NOUN
easat-10684	544	28	for	for	ADP
easat-10684	544	29	fake	fake	ADJ
easat-10684	544	30	news	news	NOUN
easat-10684	544	31	detection	detection	NOUN
easat-10684	544	32	using	use	VERB
easat-10684	544	33	machine	machine	NOUN
easat-10684	544	34	learning	learning	NOUN
easat-10684	544	35	techniques	technique	NOUN
easat-10684	544	36	,	,	PUNCT
easat-10684	544	37	"	"	PUNCT
easat-10684	544	38	in	in	ADP
easat-10684	544	39	2022	2022	NUM
easat-10684	544	40	2nd	2nd	ADJ
easat-10684	544	41	international	international	ADJ
easat-10684	544	42	conference	conference	NOUN
easat-10684	544	43	on	on	ADP
easat-10684	544	44	intelligent	intelligent	ADJ
easat-10684	544	45	technologies	technology	NOUN
easat-10684	544	46	(	(	PUNCT
easat-10684	544	47	conit	conit	NOUN
easat-10684	544	48	)	)	PUNCT
easat-10684	544	49	(	(	PUNCT
easat-10684	544	50	pp	pp	X
easat-10684	544	51	.	.	PUNCT
easat-10684	545	1	1	1	NUM
easat-10684	545	2	-	-	SYM
easat-10684	545	3	6	6	NUM
easat-10684	545	4	)	)	PUNCT
easat-10684	545	5	.	.	PUNCT
easat-10684	546	1	ieee	ieee	NOUN
easat-10684	546	2	,	,	PUNCT
easat-10684	546	3	2022	2022	NUM
easat-10684	546	4	.	.	PUNCT
easat-10684	547	1	[	[	X
easat-10684	547	2	13	13	NUM
easat-10684	547	3	]	]	PUNCT
easat-10684	547	4	m.	m.	NOUN
easat-10684	547	5	jaybhaye	jaybhaye	NOUN
easat-10684	547	6	,	,	PUNCT
easat-10684	547	7	"	"	PUNCT
easat-10684	547	8	fake	fake	ADJ
easat-10684	547	9	news	news	NOUN
easat-10684	547	10	detection	detection	NOUN
easat-10684	547	11	using	use	VERB
easat-10684	547	12	an	an	DET
easat-10684	547	13	lstm	lstm	NOUN
easat-10684	547	14	-	-	PUNCT
easat-10684	547	15	based	base	VERB
easat-10684	547	16	deep	deep	ADJ
easat-10684	547	17	learning	learning	NOUN
easat-10684	547	18	approach	approach	NOUN
easat-10684	547	19	,	,	PUNCT
easat-10684	547	20	"	"	PUNCT
easat-10684	547	21	presented	present	VERB
easat-10684	547	22	at	at	ADP
easat-10684	547	23	the	the	DET
easat-10684	547	24	itm	itm	PROPN
easat-10684	547	25	web	web	NOUN
easat-10684	547	26	of	of	ADP
easat-10684	547	27	conferences	conference	NOUN
easat-10684	547	28	,	,	PUNCT
easat-10684	547	29	2023	2023	NUM
easat-10684	547	30	.	.	PUNCT
easat-10684	548	1	[	[	X
easat-10684	548	2	14	14	NUM
easat-10684	548	3	]	]	PUNCT
easat-10684	548	4	r.	r.	PROPN
easat-10684	548	5	k.	k.	PROPN
easat-10684	548	6	kaliyar	kaliyar	PROPN
easat-10684	548	7	,	,	PUNCT
easat-10684	548	8	a.	a.	NOUN
easat-10684	548	9	goswami	goswami	NOUN
easat-10684	548	10	,	,	PUNCT
easat-10684	548	11	and	and	CCONJ
easat-10684	548	12	p.	p.	PROPN
easat-10684	548	13	narang	narang	PROPN
easat-10684	548	14	,	,	PUNCT
easat-10684	548	15	"	"	PUNCT
easat-10684	548	16	fakebert	fakebert	NOUN
easat-10684	548	17	:	:	PUNCT
easat-10684	548	18	fake	fake	ADJ
easat-10684	548	19	news	news	NOUN
easat-10684	548	20	detection	detection	NOUN
easat-10684	548	21	in	in	ADP
easat-10684	548	22	social	social	ADJ
easat-10684	548	23	media	medium	NOUN
easat-10684	548	24	with	with	ADP
easat-10684	548	25	a	a	DET
easat-10684	548	26	bert	bert	NOUN
easat-10684	548	27	-	-	PUNCT
easat-10684	548	28	based	base	VERB
easat-10684	548	29	deep	deep	ADJ
easat-10684	548	30	learning	learning	NOUN
easat-10684	548	31	approach	approach	NOUN
easat-10684	548	32	,	,	PUNCT
easat-10684	548	33	"	"	PUNCT
easat-10684	548	34	multimedia	multimedia	NOUN
easat-10684	548	35	tools	tool	NOUN
easat-10684	548	36	and	and	CCONJ
easat-10684	548	37	applications	application	NOUN
easat-10684	548	38	,	,	PUNCT
easat-10684	548	39	vol	vol	NOUN
easat-10684	548	40	.	.	PROPN
easat-10684	548	41	80	80	NUM
easat-10684	548	42	,	,	PUNCT
easat-10684	548	43	no	no	INTJ
easat-10684	548	44	.	.	NOUN
easat-10684	548	45	8	8	NUM
easat-10684	548	46	,	,	PUNCT
easat-10684	548	47	pp	pp	ADJ
easat-10684	548	48	.	.	PUNCT
easat-10684	549	1	11765	11765	NUM
easat-10684	549	2	-	-	SYM
easat-10684	549	3	11788	11788	NUM
easat-10684	549	4	,	,	PUNCT
easat-10684	549	5	2021	2021	NUM
easat-10684	549	6	.	.	PUNCT
easat-10684	550	1	[	[	X
easat-10684	550	2	15	15	NUM
easat-10684	550	3	]	]	PUNCT
easat-10684	550	4	k.	k.	PROPN
easat-10684	550	5	sharma	sharma	PROPN
easat-10684	550	6	,	,	PUNCT
easat-10684	550	7	f.	f.	PROPN
easat-10684	550	8	qian	qian	PROPN
easat-10684	550	9	,	,	PUNCT
easat-10684	550	10	h.	h.	PROPN
easat-10684	550	11	jiang	jiang	PROPN
easat-10684	550	12	,	,	PUNCT
easat-10684	550	13	n.	n.	PROPN
easat-10684	550	14	ruchansky	ruchansky	PROPN
easat-10684	550	15	,	,	PUNCT
easat-10684	550	16	m.	m.	PROPN
easat-10684	550	17	zhang	zhang	PROPN
easat-10684	550	18	,	,	PUNCT
easat-10684	550	19	and	and	CCONJ
easat-10684	550	20	y.	y.	PROPN
easat-10684	550	21	liu	liu	PROPN
easat-10684	550	22	,	,	PUNCT
easat-10684	550	23	"	"	PUNCT
easat-10684	550	24	combating	combat	VERB
easat-10684	550	25	fake	fake	ADJ
easat-10684	550	26	news	news	NOUN
easat-10684	550	27	:	:	PUNCT
easat-10684	550	28	a	a	DET
easat-10684	550	29	survey	survey	NOUN
easat-10684	550	30	on	on	ADP
easat-10684	550	31	identification	identification	NOUN
easat-10684	550	32	and	and	CCONJ
easat-10684	550	33	mitigation	mitigation	NOUN
easat-10684	550	34	techniques	technique	NOUN
easat-10684	550	35	,	,	PUNCT
easat-10684	550	36	"	"	PUNCT
easat-10684	550	37	acm	acm	NOUN
easat-10684	550	38	transactions	transaction	NOUN
easat-10684	550	39	on	on	ADP
easat-10684	550	40	intelligent	intelligent	ADJ
easat-10684	550	41	systems	system	NOUN
easat-10684	550	42	and	and	CCONJ
easat-10684	550	43	technology	technology	NOUN
easat-10684	550	44	(	(	PUNCT
easat-10684	550	45	tist	tist	NOUN
easat-10684	550	46	)	)	PUNCT
easat-10684	550	47	,	,	PUNCT
easat-10684	550	48	vol	vol	NOUN
easat-10684	550	49	.	.	PROPN
easat-10684	550	50	10	10	NUM
easat-10684	550	51	,	,	PUNCT
easat-10684	550	52	no	no	INTJ
easat-10684	550	53	.	.	NOUN
easat-10684	550	54	3	3	NUM
easat-10684	550	55	,	,	PUNCT
easat-10684	550	56	pp	pp	ADJ
easat-10684	550	57	.	.	PUNCT
easat-10684	551	1	1	1	NUM
easat-10684	551	2	-	-	SYM
easat-10684	551	3	42	42	NUM
easat-10684	551	4	,	,	PUNCT
easat-10684	551	5	2019	2019	NUM
easat-10684	551	6	.	.	PUNCT
easat-10684	552	1	https://doi.org/10.1145/3305260	https://doi.org/10.1145/3305260	X
easat-10684	553	1	[	[	X
easat-10684	553	2	16	16	NUM
easat-10684	553	3	]	]	X
easat-10684	553	4	h.	h.	PROPN
easat-10684	553	5	thakar	thakar	PROPN
easat-10684	553	6	and	and	CCONJ
easat-10684	553	7	b.	b.	PROPN
easat-10684	553	8	bhatt	bhatt	PROPN
easat-10684	553	9	,	,	PUNCT
easat-10684	553	10	"	"	PUNCT
easat-10684	553	11	fake	fake	ADJ
easat-10684	553	12	news	news	NOUN
easat-10684	553	13	detection	detection	NOUN
easat-10684	553	14	:	:	PUNCT
easat-10684	553	15	recent	recent	ADJ
easat-10684	553	16	trends	trend	NOUN
easat-10684	553	17	and	and	CCONJ
easat-10684	553	18	challenges	challenge	NOUN
easat-10684	553	19	,	,	PUNCT
easat-10684	553	20	"	"	PUNCT
easat-10684	553	21	social	social	ADJ
easat-10684	553	22	network	network	NOUN
easat-10684	553	23	analysis	analysis	NOUN
easat-10684	553	24	and	and	CCONJ
easat-10684	553	25	mining	mining	NOUN
easat-10684	553	26	,	,	PUNCT
easat-10684	553	27	vol	vol	NOUN
easat-10684	553	28	.	.	PROPN
easat-10684	553	29	14	14	NUM
easat-10684	553	30	,	,	PUNCT
easat-10684	553	31	no	no	INTJ
easat-10684	553	32	.	.	NOUN
easat-10684	553	33	1	1	NUM
easat-10684	553	34	,	,	PUNCT
easat-10684	553	35	p.	p.	NOUN
easat-10684	553	36	176	176	NUM
easat-10684	553	37	,	,	PUNCT
easat-10684	553	38	2024	2024	NUM
easat-10684	553	39	.	.	PUNCT
easat-10684	554	1	[	[	X
easat-10684	554	2	17	17	NUM
easat-10684	554	3	]	]	X
easat-10684	554	4	k.	k.	PROPN
easat-10684	554	5	i.	i.	PROPN
easat-10684	554	6	roumeliotis	roumeliotis	PROPN
easat-10684	554	7	,	,	PUNCT
easat-10684	554	8	n.	n.	PROPN
easat-10684	554	9	d.	d.	PROPN
easat-10684	554	10	tselikas	tselikas	PROPN
easat-10684	554	11	,	,	PUNCT
easat-10684	554	12	and	and	CCONJ
easat-10684	554	13	d.	d.	PROPN
easat-10684	554	14	k.	k.	PROPN
easat-10684	554	15	nasiopoulos	nasiopoulos	PROPN
easat-10684	554	16	,	,	PUNCT
easat-10684	554	17	"	"	PUNCT
easat-10684	554	18	unmasking	unmask	VERB
easat-10684	554	19	misinformation	misinformation	NOUN
easat-10684	554	20	:	:	PUNCT
easat-10684	554	21	leveraging	leverage	VERB
easat-10684	554	22	cnn	cnn	PROPN
easat-10684	554	23	,	,	PUNCT
easat-10684	554	24	bert	bert	PROPN
easat-10684	554	25	,	,	PUNCT
easat-10684	554	26	and	and	CCONJ
easat-10684	554	27	gpt	gpt	NOUN
easat-10684	554	28	models	model	NOUN
easat-10684	554	29	for	for	ADP
easat-10684	554	30	robust	robust	ADJ
easat-10684	554	31	fake	fake	ADJ
easat-10684	554	32	news	news	NOUN
easat-10684	554	33	classification	classification	NOUN
easat-10684	554	34	,	,	PUNCT
easat-10684	554	35	"	"	PUNCT
easat-10684	554	36	2024	2024	NUM
easat-10684	554	37	.	.	PUNCT
easat-10684	555	1	[	[	X
easat-10684	555	2	18	18	NUM
easat-10684	555	3	]	]	X
easat-10684	555	4	e.	e.	PROPN
easat-10684	555	5	hashmi	hashmi	PROPN
easat-10684	555	6	,	,	PUNCT
easat-10684	555	7	s.	s.	PROPN
easat-10684	555	8	y.	y.	PROPN
easat-10684	555	9	yayilgan	yayilgan	PROPN
easat-10684	555	10	,	,	PUNCT
easat-10684	555	11	m.	m.	NOUN
easat-10684	555	12	m.	m.	PROPN
easat-10684	555	13	yamin	yamin	PROPN
easat-10684	555	14	,	,	PUNCT
easat-10684	555	15	s.	s.	PROPN
easat-10684	555	16	ali	ali	PROPN
easat-10684	555	17	,	,	PUNCT
easat-10684	555	18	and	and	CCONJ
easat-10684	555	19	m.	m.	NOUN
easat-10684	555	20	abomhara	abomhara	PROPN
easat-10684	555	21	,	,	PUNCT
easat-10684	555	22	"	"	PUNCT
easat-10684	555	23	advancing	advance	VERB
easat-10684	555	24	fake	fake	ADJ
easat-10684	555	25	news	news	NOUN
easat-10684	555	26	detection	detection	NOUN
easat-10684	555	27	:	:	PUNCT
easat-10684	555	28	hybrid	hybrid	ADJ
easat-10684	555	29	deep	deep	ADJ
easat-10684	555	30	learning	learning	NOUN
easat-10684	555	31	with	with	ADP
easat-10684	555	32	fasttext	fasttext	NOUN
easat-10684	555	33	and	and	CCONJ
easat-10684	555	34	explainable	explainable	ADJ
easat-10684	555	35	ai	ai	NOUN
easat-10684	555	36	,	,	PUNCT
easat-10684	555	37	"	"	PUNCT
easat-10684	555	38	ieee	ieee	NOUN
easat-10684	555	39	access	access	NOUN
easat-10684	555	40	,	,	PUNCT
easat-10684	555	41	vol	vol	NOUN
easat-10684	555	42	.	.	PROPN
easat-10684	555	43	12	12	NUM
easat-10684	555	44	,	,	PUNCT
easat-10684	555	45	pp	pp	ADJ
easat-10684	555	46	.	.	PUNCT
easat-10684	555	47	44462	44462	NUM
easat-10684	555	48	-	-	SYM
easat-10684	555	49	44480	44480	NUM
easat-10684	555	50	,	,	PUNCT
easat-10684	555	51	2024	2024	NUM
easat-10684	555	52	.	.	PUNCT
easat-10684	556	1	[	[	X
easat-10684	556	2	19	19	NUM
easat-10684	556	3	]	]	X
easat-10684	556	4	e.	e.	PROPN
easat-10684	556	5	m.	m.	PROPN
easat-10684	556	6	almandouh	almandouh	PROPN
easat-10684	556	7	,	,	PUNCT
easat-10684	556	8	m.	m.	PROPN
easat-10684	556	9	f.	f.	PROPN
easat-10684	556	10	alrahmawy	alrahmawy	PROPN
easat-10684	556	11	,	,	PUNCT
easat-10684	556	12	m.	m.	NOUN
easat-10684	556	13	eisa	eisa	PROPN
easat-10684	556	14	,	,	PUNCT
easat-10684	556	15	m.	m.	NOUN
easat-10684	556	16	elhoseny	elhoseny	PROPN
easat-10684	556	17	,	,	PUNCT
easat-10684	556	18	and	and	CCONJ
easat-10684	556	19	a.	a.	NOUN
easat-10684	556	20	tolba	tolba	PROPN
easat-10684	556	21	,	,	PUNCT
easat-10684	556	22	"	"	PUNCT
easat-10684	556	23	ensemble	ensemble	ADJ
easat-10684	556	24	based	base	VERB
easat-10684	556	25	high	high	ADJ
easat-10684	556	26	performance	performance	NOUN
easat-10684	556	27	deep	deep	ADJ
easat-10684	556	28	learning	learning	NOUN
easat-10684	556	29	models	model	NOUN
easat-10684	556	30	for	for	ADP
easat-10684	556	31	fake	fake	ADJ
easat-10684	556	32	news	news	NOUN
easat-10684	556	33	detection	detection	NOUN
easat-10684	556	34	,	,	PUNCT
easat-10684	556	35	"	"	PUNCT
easat-10684	556	36	scientific	scientific	ADJ
easat-10684	556	37	reports	report	NOUN
easat-10684	556	38	,	,	PUNCT
easat-10684	556	39	vol	vol	NOUN
easat-10684	556	40	.	.	PROPN
easat-10684	557	1	14	14	NUM
easat-10684	557	2	,	,	PUNCT
easat-10684	557	3	no	no	INTJ
easat-10684	557	4	.	.	NOUN
easat-10684	557	5	1	1	NUM
easat-10684	557	6	,	,	PUNCT
easat-10684	557	7	p.	p.	NOUN
easat-10684	557	8	26591	26591	NUM
easat-10684	557	9	,	,	PUNCT
easat-10684	557	10	2024	2024	NUM
easat-10684	557	11	.	.	PUNCT
easat-10684	558	1	https://doi.org/10.1038/s41598-024-76286-0	https://doi.org/10.1038/s41598-024-76286-0	NUM
easat-10684	559	1	[	[	X
easat-10684	559	2	20	20	NUM
easat-10684	559	3	]	]	PUNCT
easat-10684	559	4	m.	m.	NOUN
easat-10684	559	5	al	al	PROPN
easat-10684	559	6	-	-	PUNCT
easat-10684	559	7	alshaqi	alshaqi	PROPN
easat-10684	559	8	,	,	PUNCT
easat-10684	559	9	d.	d.	PROPN
easat-10684	559	10	b.	b.	PROPN
easat-10684	559	11	rawat	rawat	PROPN
easat-10684	559	12	,	,	PUNCT
easat-10684	559	13	and	and	CCONJ
easat-10684	559	14	c.	c.	PROPN
easat-10684	559	15	liu	liu	PROPN
easat-10684	559	16	,	,	PUNCT
easat-10684	559	17	"	"	PUNCT
easat-10684	559	18	ensemble	ensemble	ADJ
easat-10684	559	19	techniques	technique	NOUN
easat-10684	559	20	for	for	ADP
easat-10684	559	21	robust	robust	ADJ
easat-10684	559	22	fake	fake	ADJ
easat-10684	559	23	news	news	NOUN
easat-10684	559	24	detection	detection	NOUN
easat-10684	559	25	:	:	PUNCT
easat-10684	559	26	integrating	integrate	VERB
easat-10684	559	27	transformers	transformer	NOUN
easat-10684	559	28	,	,	PUNCT
easat-10684	559	29	natural	natural	ADJ
easat-10684	559	30	language	language	NOUN
easat-10684	559	31	processing	processing	NOUN
easat-10684	559	32	,	,	PUNCT
easat-10684	559	33	and	and	CCONJ
easat-10684	559	34	machine	machine	NOUN
easat-10684	559	35	learning	learning	NOUN
easat-10684	559	36	,	,	PUNCT
easat-10684	559	37	"	"	PUNCT
easat-10684	559	38	sensors	sensor	NOUN
easat-10684	559	39	,	,	PUNCT
easat-10684	559	40	vol	vol	NOUN
easat-10684	559	41	.	.	PROPN
easat-10684	559	42	24	24	NUM
easat-10684	559	43	,	,	PUNCT
easat-10684	559	44	no	no	INTJ
easat-10684	559	45	.	.	NOUN
easat-10684	559	46	18	18	NUM
easat-10684	559	47	,	,	PUNCT
easat-10684	559	48	p.	p.	NOUN
easat-10684	559	49	6062	6062	NUM
easat-10684	559	50	,	,	PUNCT
easat-10684	559	51	2024	2024	NUM
easat-10684	559	52	.	.	PUNCT
easat-10684	560	1	https://doi.org/10.3390/s24186062	https://doi.org/10.3390/s24186062	PRON
easat-10684	560	2	[	[	X
easat-10684	560	3	21	21	NUM
easat-10684	560	4	]	]	PUNCT
easat-10684	560	5	m.	m.	NOUN
easat-10684	560	6	a.	a.	PROPN
easat-10684	560	7	al	al	PROPN
easat-10684	560	8	-	-	PUNCT
easat-10684	560	9	tarawneh	tarawneh	PROPN
easat-10684	560	10	,	,	PUNCT
easat-10684	560	11	o.	o.	PROPN
easat-10684	560	12	al	al	PROPN
easat-10684	560	13	-	-	PUNCT
easat-10684	560	14	irr	irr	PROPN
easat-10684	560	15	,	,	PUNCT
easat-10684	560	16	k.	k.	PROPN
easat-10684	560	17	s.	s.	PROPN
easat-10684	560	18	al	al	PROPN
easat-10684	560	19	-	-	PUNCT
easat-10684	560	20	maaitah	maaitah	PROPN
easat-10684	560	21	,	,	PUNCT
easat-10684	560	22	h.	h.	PROPN
easat-10684	560	23	kanj	kanj	PROPN
easat-10684	560	24	,	,	PUNCT
easat-10684	560	25	and	and	CCONJ
easat-10684	560	26	w.	w.	PROPN
easat-10684	560	27	h.	h.	PROPN
easat-10684	560	28	f.	f.	PROPN
easat-10684	560	29	aly	aly	PROPN
easat-10684	560	30	,	,	PUNCT
easat-10684	560	31	"	"	PUNCT
easat-10684	560	32	enhancing	enhance	VERB
easat-10684	560	33	fake	fake	ADJ
easat-10684	560	34	news	news	NOUN
easat-10684	560	35	detection	detection	NOUN
easat-10684	560	36	with	with	ADP
easat-10684	560	37	word	word	NOUN
easat-10684	560	38	embedding	embed	VERB
easat-10684	560	39	:	:	PUNCT
easat-10684	560	40	a	a	DET
easat-10684	560	41	machine	machine	NOUN
easat-10684	560	42	learning	learning	NOUN
easat-10684	560	43	and	and	CCONJ
easat-10684	560	44	deep	deep	ADJ
easat-10684	560	45	learning	learning	NOUN
easat-10684	560	46	approach	approach	NOUN
easat-10684	560	47	,	,	PUNCT
easat-10684	560	48	"	"	PUNCT
easat-10684	560	49	computers	computer	NOUN
easat-10684	560	50	,	,	PUNCT
easat-10684	560	51	vol	vol	NOUN
easat-10684	560	52	.	.	PROPN
easat-10684	560	53	13	13	NUM
easat-10684	560	54	,	,	PUNCT
easat-10684	560	55	no	no	INTJ
easat-10684	560	56	.	.	NOUN
easat-10684	560	57	9	9	NUM
easat-10684	560	58	,	,	PUNCT
easat-10684	560	59	p.	p.	NOUN
easat-10684	560	60	239	239	NUM
easat-10684	560	61	,	,	PUNCT
easat-10684	560	62	2024	2024	NUM
easat-10684	560	63	.	.	PUNCT
easat-10684	561	1	https://doi.org/10.3390/computers13090239	https://doi.org/10.3390/computers13090239	NOUN
easat-10684	562	1	[	[	X
easat-10684	562	2	22	22	NUM
easat-10684	562	3	]	]	PUNCT
easat-10684	562	4	e.-s	e.-	NOUN
easat-10684	562	5	.	.	PUNCT
easat-10684	563	1	apostol	apostol	NOUN
easat-10684	563	2	and	and	CCONJ
easat-10684	563	3	c.-o	c.-o	NOUN
easat-10684	563	4	.	.	PUNCT
easat-10684	564	1	truică	truică	NOUN
easat-10684	564	2	,	,	PUNCT
easat-10684	564	3	"	"	PUNCT
easat-10684	564	4	efficient	efficient	ADJ
easat-10684	564	5	machine	machine	NOUN
easat-10684	564	6	learning	learn	VERB
easat-10684	564	7	ensemble	ensemble	ADJ
easat-10684	564	8	methods	method	NOUN
easat-10684	564	9	for	for	ADP
easat-10684	564	10	detecting	detect	VERB
easat-10684	564	11	gravitational	gravitational	ADJ
easat-10684	564	12	wave	wave	NOUN
easat-10684	564	13	glitches	glitch	NOUN
easat-10684	564	14	in	in	ADP
easat-10684	564	15	ligo	ligo	ADJ
easat-10684	564	16	time	time	NOUN
easat-10684	564	17	series	series	PROPN
easat-10684	564	18	,	,	PUNCT
easat-10684	564	19	"	"	PUNCT
easat-10684	564	20	in	in	ADP
easat-10684	564	21	2023	2023	NUM
easat-10684	564	22	ieee	ieee	NOUN
easat-10684	564	23	19th	19th	ADJ
easat-10684	564	24	international	international	ADJ
easat-10684	564	25	conference	conference	NOUN
easat-10684	564	26	on	on	ADP
easat-10684	564	27	intelligent	intelligent	ADJ
easat-10684	564	28	computer	computer	NOUN
easat-10684	564	29	communication	communication	NOUN
easat-10684	564	30	and	and	CCONJ
easat-10684	564	31	processing	processing	NOUN
easat-10684	564	32	(	(	PUNCT
easat-10684	564	33	iccp	iccp	ADV
easat-10684	564	34	)	)	PUNCT
easat-10684	564	35	(	(	PUNCT
easat-10684	564	36	pp	pp	X
easat-10684	564	37	.	.	PUNCT
easat-10684	565	1	79	79	NUM
easat-10684	565	2	-	-	SYM
easat-10684	565	3	86	86	NUM
easat-10684	565	4	)	)	PUNCT
easat-10684	565	5	.	.	PUNCT
easat-10684	566	1	ieee	ieee	NOUN
easat-10684	566	2	,	,	PUNCT
easat-10684	566	3	2023	2023	NUM
easat-10684	566	4	.	.	PUNCT
easat-10684	567	1	[	[	X
easat-10684	567	2	23	23	NUM
easat-10684	567	3	]	]	X
easat-10684	567	4	n.	n.	PROPN
easat-10684	567	5	nasir	nasir	PROPN
easat-10684	567	6	,	,	PUNCT
easat-10684	567	7	f.	f.	PROPN
easat-10684	567	8	rehman	rehman	PROPN
easat-10684	567	9	,	,	PUNCT
easat-10684	567	10	and	and	CCONJ
easat-10684	567	11	s.	s.	PROPN
easat-10684	567	12	f.	f.	PROPN
easat-10684	567	13	omair	omair	PROPN
easat-10684	567	14	,	,	PUNCT
easat-10684	567	15	"	"	PUNCT
easat-10684	567	16	risk	risk	NOUN
easat-10684	567	17	factors	factor	NOUN
easat-10684	567	18	for	for	ADP
easat-10684	567	19	bacterial	bacterial	ADJ
easat-10684	567	20	infections	infection	NOUN
easat-10684	567	21	in	in	ADP
easat-10684	567	22	patients	patient	NOUN
easat-10684	567	23	with	with	ADP
easat-10684	567	24	moderate	moderate	ADJ
easat-10684	567	25	to	to	ADP
easat-10684	567	26	severe	severe	ADJ
easat-10684	567	27	covid‐19	covid‐19	PROPN
easat-10684	567	28	:	:	PUNCT
easat-10684	567	29	a	a	DET
easat-10684	567	30	case‐control	case‐control	NOUN
easat-10684	567	31	study	study	NOUN
easat-10684	567	32	,	,	PUNCT
easat-10684	567	33	"	"	PUNCT
easat-10684	567	34	journal	journal	NOUN
easat-10684	567	35	of	of	ADP
easat-10684	567	36	medical	medical	ADJ
easat-10684	567	37	virology	virology	NOUN
easat-10684	567	38	,	,	PUNCT
easat-10684	567	39	vol	vol	NOUN
easat-10684	567	40	.	.	PROPN
easat-10684	567	41	93	93	NUM
easat-10684	567	42	,	,	PUNCT
easat-10684	567	43	no	no	INTJ
easat-10684	567	44	.	.	NOUN
easat-10684	567	45	7	7	NUM
easat-10684	567	46	,	,	PUNCT
easat-10684	567	47	pp	pp	ADJ
easat-10684	567	48	.	.	PUNCT
easat-10684	567	49	4564	4564	NUM
easat-10684	567	50	-	-	SYM
easat-10684	567	51	4569	4569	NUM
easat-10684	567	52	,	,	PUNCT
easat-10684	567	53	2021	2021	NUM
easat-10684	567	54	.	.	PUNCT
easat-10684	568	1	[	[	X
easat-10684	568	2	24	24	NUM
easat-10684	568	3	]	]	PUNCT
easat-10684	568	4	s.	s.	PROPN
easat-10684	568	5	k.	k.	PROPN
easat-10684	568	6	thota	thota	PROPN
easat-10684	568	7	,	,	PUNCT
easat-10684	568	8	m.	m.	NOUN
easat-10684	568	9	gujjalapudi	gujjalapudi	PROPN
easat-10684	568	10	,	,	PUNCT
easat-10684	568	11	c.	c.	PROPN
easat-10684	568	12	anam	anam	PROPN
easat-10684	568	13	,	,	PUNCT
easat-10684	568	14	r.	r.	PROPN
easat-10684	568	15	chiluka	chiluka	PROPN
easat-10684	568	16	,	,	PUNCT
easat-10684	568	17	p.	p.	NOUN
easat-10684	568	18	mamidi	mamidi	NOUN
easat-10684	568	19	,	,	PUNCT
easat-10684	568	20	and	and	CCONJ
easat-10684	568	21	s.	s.	PROPN
easat-10684	568	22	m.	m.	PROPN
easat-10684	568	23	magar	magar	PROPN
easat-10684	568	24	,	,	PUNCT
easat-10684	568	25	"	"	PUNCT
easat-10684	568	26	comparative	comparative	ADJ
easat-10684	568	27	evaluation	evaluation	NOUN
easat-10684	568	28	of	of	ADP
easat-10684	568	29	the	the	DET
easat-10684	568	30	retention	retention	NOUN
easat-10684	568	31	of	of	ADP
easat-10684	568	32	maxillary	maxillary	NOUN
easat-10684	568	33	complete	complete	ADJ
easat-10684	568	34	denture	denture	NOUN
easat-10684	568	35	bases	basis	NOUN
easat-10684	568	36	made	make	VERB
easat-10684	568	37	by	by	ADP
easat-10684	568	38	heat	heat	NOUN
easat-10684	568	39	cure	cure	NOUN
easat-10684	568	40	acrylic	acrylic	NOUN
easat-10684	568	41	resin	resin	NOUN
easat-10684	568	42	processed	process	VERB
easat-10684	568	43	on	on	ADP
easat-10684	568	44	high	high	ADJ
easat-10684	568	45	expansion	expansion	NOUN
easat-10684	568	46	stone	stone	NOUN
easat-10684	568	47	and	and	CCONJ
easat-10684	568	48	type	type	NOUN
easat-10684	568	49	iii	iii	NUM
easat-10684	568	50	dental	dental	ADJ
easat-10684	568	51	stone	stone	NOUN
easat-10684	568	52	:	:	PUNCT
easat-10684	568	53	an	an	DET
easat-10684	568	54	in	in	ADP
easat-10684	568	55	-	-	PUNCT
easat-10684	568	56	vivo	vivo	NOUN
easat-10684	568	57	study	study	NOUN
easat-10684	568	58	,	,	PUNCT
easat-10684	568	59	"	"	PUNCT
easat-10684	568	60	int	int	NOUN
easat-10684	568	61	j	j	PROPN
easat-10684	568	62	cur	cur	PROPN
easat-10684	568	63	res	res	PROPN
easat-10684	568	64	rev	rev	PROPN
easat-10684	568	65	,	,	PUNCT
easat-10684	568	66	vol	vol	NOUN
easat-10684	568	67	.	.	PROPN
easat-10684	568	68	13	13	NUM
easat-10684	568	69	,	,	PUNCT
easat-10684	568	70	no	no	INTJ
easat-10684	568	71	.	.	NOUN
easat-10684	568	72	15	15	NUM
easat-10684	568	73	,	,	PUNCT
easat-10684	568	74	p.	p.	NOUN
easat-10684	568	75	14	14	NUM
easat-10684	568	76	,	,	PUNCT
easat-10684	568	77	2021	2021	NUM
easat-10684	568	78	.	.	PUNCT
easat-10684	569	1	[	[	X
easat-10684	569	2	25	25	NUM
easat-10684	569	3	]	]	PUNCT
easat-10684	569	4	a.	a.	NOUN
easat-10684	569	5	choudhary	choudhary	PROPN
easat-10684	569	6	and	and	CCONJ
easat-10684	569	7	a.	a.	PROPN
easat-10684	569	8	arora	arora	PROPN
easat-10684	569	9	,	,	PUNCT
easat-10684	569	10	"	"	PUNCT
easat-10684	569	11	linguistic	linguistic	ADJ
easat-10684	569	12	feature	feature	NOUN
easat-10684	569	13	based	base	VERB
easat-10684	569	14	learning	learn	VERB
easat-10684	569	15	model	model	NOUN
easat-10684	569	16	for	for	ADP
easat-10684	569	17	fake	fake	ADJ
easat-10684	569	18	news	news	NOUN
easat-10684	569	19	detection	detection	NOUN
easat-10684	569	20	and	and	CCONJ
easat-10684	569	21	classification	classification	NOUN
easat-10684	569	22	,	,	PUNCT
easat-10684	569	23	"	"	PUNCT
easat-10684	569	24	expert	expert	NOUN
easat-10684	569	25	systems	system	NOUN
easat-10684	569	26	with	with	ADP
easat-10684	569	27	applications	application	NOUN
easat-10684	569	28	,	,	PUNCT
easat-10684	569	29	vol	vol	NOUN
easat-10684	569	30	.	.	PROPN
easat-10684	569	31	169	169	NUM
easat-10684	569	32	,	,	PUNCT
easat-10684	569	33	p.	p.	NOUN
easat-10684	569	34	114171	114171	NUM
easat-10684	569	35	,	,	PUNCT
easat-10684	569	36	2021	2021	NUM
easat-10684	569	37	.	.	PUNCT
easat-10684	570	1	[	[	X
easat-10684	570	2	26	26	NUM
easat-10684	570	3	]	]	X
easat-10684	570	4	r.	r.	PROPN
easat-10684	570	5	vijjali	vijjali	PROPN
easat-10684	570	6	,	,	PUNCT
easat-10684	570	7	p.	p.	PROPN
easat-10684	570	8	potluri	potluri	PROPN
easat-10684	570	9	,	,	PUNCT
easat-10684	570	10	s.	s.	PROPN
easat-10684	570	11	kumar	kumar	PROPN
easat-10684	570	12	,	,	PUNCT
easat-10684	570	13	and	and	CCONJ
easat-10684	570	14	s.	s.	PROPN
easat-10684	570	15	teki	teki	PROPN
easat-10684	570	16	,	,	PUNCT
easat-10684	570	17	"	"	PUNCT
easat-10684	570	18	two	two	NUM
easat-10684	570	19	stage	stage	NOUN
easat-10684	570	20	transformer	transformer	NOUN
easat-10684	570	21	model	model	NOUN
easat-10684	570	22	for	for	ADP
easat-10684	570	23	covid-19	covid-19	PROPN
easat-10684	570	24	fake	fake	ADJ
easat-10684	570	25	news	news	NOUN
easat-10684	570	26	detection	detection	NOUN
easat-10684	570	27	and	and	CCONJ
easat-10684	570	28	fact	fact	NOUN
easat-10684	570	29	checking	check	VERB
easat-10684	570	30	,	,	PUNCT
easat-10684	570	31	"	"	PUNCT
easat-10684	570	32	arxiv	arxiv	PROPN
easat-10684	570	33	preprint	preprint	NOUN
easat-10684	570	34	arxiv:2011.13253	arxiv:2011.13253	NOUN
easat-10684	570	35	,	,	PUNCT
easat-10684	570	36	2020	2020	NUM
easat-10684	570	37	.	.	PUNCT
easat-10684	571	1	[	[	X
easat-10684	571	2	27	27	NUM
easat-10684	571	3	]	]	X
easat-10684	571	4	s.	s.	PROPN
easat-10684	571	5	singhal	singhal	PROPN
easat-10684	571	6	,	,	PUNCT
easat-10684	571	7	s.	s.	PROPN
easat-10684	571	8	choudhary	choudhary	PROPN
easat-10684	571	9	,	,	PUNCT
easat-10684	571	10	and	and	CCONJ
easat-10684	571	11	p.	p.	PROPN
easat-10684	571	12	c.	c.	PROPN
easat-10684	571	13	biswal	biswal	PROPN
easat-10684	571	14	,	,	PUNCT
easat-10684	571	15	"	"	PUNCT
easat-10684	571	16	return	return	NOUN
easat-10684	571	17	and	and	CCONJ
easat-10684	571	18	volatility	volatility	NOUN
easat-10684	571	19	linkages	linkage	NOUN
easat-10684	571	20	among	among	ADP
easat-10684	571	21	international	international	ADJ
easat-10684	571	22	crude	crude	ADJ
easat-10684	571	23	oil	oil	NOUN
easat-10684	571	24	price	price	NOUN
easat-10684	571	25	,	,	PUNCT
easat-10684	571	26	gold	gold	NOUN
easat-10684	571	27	price	price	NOUN
easat-10684	571	28	,	,	PUNCT
easat-10684	571	29	exchange	exchange	NOUN
easat-10684	571	30	rate	rate	NOUN
easat-10684	571	31	and	and	CCONJ
easat-10684	571	32	stock	stock	NOUN
easat-10684	571	33	markets	market	NOUN
easat-10684	571	34	:	:	PUNCT
easat-10684	571	35	evidence	evidence	NOUN
easat-10684	571	36	from	from	ADP
easat-10684	571	37	mexico	mexico	PROPN
easat-10684	571	38	,	,	PUNCT
easat-10684	571	39	"	"	PUNCT
easat-10684	571	40	resources	resource	NOUN
easat-10684	571	41	policy	policy	NOUN
easat-10684	571	42	,	,	PUNCT
easat-10684	571	43	vol	vol	NOUN
easat-10684	571	44	.	.	PROPN
easat-10684	571	45	60	60	NUM
easat-10684	571	46	,	,	PUNCT
easat-10684	571	47	pp	pp	ADJ
easat-10684	571	48	.	.	PUNCT
easat-10684	572	1	255	255	NUM
easat-10684	572	2	-	-	SYM
easat-10684	572	3	261	261	NUM
easat-10684	572	4	,	,	PUNCT
easat-10684	572	5	2019	2019	NUM
easat-10684	572	6	.	.	PUNCT
easat-10684	573	1	[	[	X
easat-10684	573	2	28	28	NUM
easat-10684	573	3	]	]	X
easat-10684	573	4	s.	s.	PROPN
easat-10684	573	5	kumari	kumari	PROPN
easat-10684	573	6	and	and	CCONJ
easat-10684	573	7	m.	m.	NOUN
easat-10684	573	8	p.	p.	PROPN
easat-10684	573	9	singh	singh	PROPN
easat-10684	573	10	,	,	PUNCT
easat-10684	573	11	"	"	PUNCT
easat-10684	573	12	a	a	DET
easat-10684	573	13	deep	deep	ADJ
easat-10684	573	14	learning	learn	VERB
easat-10684	573	15	multimodal	multimodal	NOUN
easat-10684	573	16	framework	framework	NOUN
easat-10684	573	17	for	for	ADP
easat-10684	573	18	fake	fake	ADJ
easat-10684	573	19	news	news	NOUN
easat-10684	573	20	detection	detection	NOUN
easat-10684	573	21	,	,	PUNCT
easat-10684	573	22	"	"	PUNCT
easat-10684	573	23	engineering	engineering	NOUN
easat-10684	573	24	,	,	PUNCT
easat-10684	573	25	technology	technology	NOUN
easat-10684	573	26	&	&	CCONJ
easat-10684	573	27	applied	apply	VERB
easat-10684	573	28	science	science	NOUN
easat-10684	573	29	research	research	NOUN
easat-10684	573	30	,	,	PUNCT
easat-10684	573	31	vol	vol	NOUN
easat-10684	573	32	.	.	PROPN
easat-10684	573	33	14	14	NUM
easat-10684	573	34	,	,	PUNCT
easat-10684	573	35	no	no	INTJ
easat-10684	573	36	.	.	NOUN
easat-10684	573	37	5	5	NUM
easat-10684	573	38	,	,	PUNCT
easat-10684	573	39	pp	pp	ADJ
easat-10684	573	40	.	.	PUNCT
easat-10684	574	1	16527	16527	NUM
easat-10684	574	2	-	-	SYM
easat-10684	574	3	16533	16533	NUM
easat-10684	574	4	,	,	PUNCT
easat-10684	574	5	2024	2024	NUM
easat-10684	574	6	.	.	PUNCT
easat-10684	575	1	[	[	X
easat-10684	575	2	29	29	NUM
easat-10684	575	3	]	]	X
easat-10684	575	4	h.	h.	PROPN
easat-10684	575	5	gul	gul	PROPN
easat-10684	575	6	,	,	PUNCT
easat-10684	575	7	f.	f.	PROPN
easat-10684	575	8	al	al	PROPN
easat-10684	575	9	-	-	PUNCT
easat-10684	575	10	obeidat	obeidat	PROPN
easat-10684	575	11	,	,	PUNCT
easat-10684	575	12	m.	m.	NOUN
easat-10684	575	13	wasim	wasim	PROPN
easat-10684	575	14	,	,	PUNCT
easat-10684	575	15	a.	a.	NOUN
easat-10684	575	16	amin	amin	PROPN
easat-10684	575	17	,	,	PUNCT
easat-10684	575	18	and	and	CCONJ
easat-10684	575	19	f.	f.	PROPN
easat-10684	575	20	moreira	moreira	PROPN
easat-10684	575	21	,	,	PUNCT
easat-10684	575	22	"	"	PUNCT
easat-10684	575	23	advancing	advance	VERB
easat-10684	575	24	fake	fake	ADJ
easat-10684	575	25	news	news	NOUN
easat-10684	575	26	detection	detection	NOUN
easat-10684	575	27	with	with	ADP
easat-10684	575	28	graph	graph	NOUN
easat-10684	575	29	neural	neural	ADJ
easat-10684	575	30	network	network	NOUN
easat-10684	575	31	and	and	CCONJ
easat-10684	575	32	deep	deep	ADJ
easat-10684	575	33	learning	learning	NOUN
easat-10684	575	34	,	,	PUNCT
easat-10684	575	35	"	"	PUNCT
easat-10684	575	36	journal	journal	NOUN
easat-10684	575	37	of	of	ADP
easat-10684	575	38	physics	physics	PROPN
easat-10684	575	39	:	:	PUNCT
easat-10684	575	40	complexity	complexity	NOUN
easat-10684	575	41	,	,	PUNCT
easat-10684	575	42	vol	vol	NOUN
easat-10684	575	43	.	.	PROPN
easat-10684	575	44	6	6	NUM
easat-10684	575	45	,	,	PUNCT
easat-10684	575	46	no	no	INTJ
easat-10684	575	47	.	.	NOUN
easat-10684	575	48	2	2	NUM
easat-10684	575	49	,	,	PUNCT
easat-10684	575	50	p.	p.	NOUN
easat-10684	575	51	025001	025001	NUM
easat-10684	575	52	,	,	PUNCT
easat-10684	575	53	2025	2025	NUM
easat-10684	575	54	.	.	PUNCT
easat-10684	576	1	[	[	X
easat-10684	576	2	30	30	NUM
easat-10684	576	3	]	]	X
easat-10684	576	4	w.	w.	PROPN
easat-10684	576	5	k.	k.	PROPN
easat-10684	576	6	sari	sari	PROPN
easat-10684	576	7	,	,	PUNCT
easat-10684	576	8	i.	i.	PROPN
easat-10684	576	9	s.	s.	PROPN
easat-10684	576	10	b.	b.	PROPN
easat-10684	576	11	azhar	azhar	PROPN
easat-10684	576	12	,	,	PUNCT
easat-10684	576	13	z.	z.	PROPN
easat-10684	576	14	yamani	yamani	PROPN
easat-10684	576	15	,	,	PUNCT
easat-10684	576	16	and	and	CCONJ
easat-10684	576	17	y.	y.	PROPN
easat-10684	576	18	florensia	florensia	PROPN
easat-10684	576	19	,	,	PUNCT
easat-10684	576	20	"	"	PUNCT
easat-10684	576	21	fake	fake	ADJ
easat-10684	576	22	news	news	NOUN
easat-10684	576	23	detection	detection	NOUN
easat-10684	576	24	using	use	VERB
easat-10684	576	25	optimized	optimize	VERB
easat-10684	576	26	convolutional	convolutional	ADJ
easat-10684	576	27	neural	neural	ADJ
easat-10684	576	28	network	network	NOUN
easat-10684	576	29	and	and	CCONJ
easat-10684	576	30	bidirectional	bidirectional	ADJ
easat-10684	576	31	long	long	ADJ
easat-10684	576	32	short	short	ADJ
easat-10684	576	33	-	-	PUNCT
easat-10684	576	34	term	term	NOUN
easat-10684	576	35	memory	memory	NOUN
easat-10684	576	36	,	,	PUNCT
easat-10684	576	37	"	"	PUNCT
easat-10684	576	38	computer	computer	NOUN
easat-10684	576	39	engineering	engineering	NOUN
easat-10684	576	40	&	&	CCONJ
easat-10684	576	41	applications	application	NOUN
easat-10684	576	42	journal	journal	PROPN
easat-10684	576	43	,	,	PUNCT
easat-10684	576	44	vol	vol	NOUN
easat-10684	576	45	.	.	PROPN
easat-10684	576	46	13	13	NUM
easat-10684	576	47	,	,	PUNCT
easat-10684	576	48	no	no	INTJ
easat-10684	576	49	.	.	NOUN
easat-10684	576	50	3	3	NUM
easat-10684	576	51	,	,	PUNCT
easat-10684	576	52	pp	pp	ADJ
easat-10684	576	53	.	.	PUNCT
easat-10684	576	54	25	25	NUM
easat-10684	576	55	-	-	SYM
easat-10684	576	56	33	33	NUM
easat-10684	576	57	,	,	PUNCT
easat-10684	576	58	2024	2024	NUM
easat-10684	576	59	.	.	PUNCT
easat-10684	577	1	[	[	X
easat-10684	577	2	31	31	NUM
easat-10684	577	3	]	]	PUNCT
easat-10684	577	4	b.	b.	PROPN
easat-10684	577	5	brinda	brinda	PROPN
easat-10684	577	6	,	,	PUNCT
easat-10684	577	7	c.	c.	PROPN
easat-10684	577	8	rajan	rajan	PROPN
easat-10684	577	9	,	,	PUNCT
easat-10684	577	10	and	and	CCONJ
easat-10684	577	11	k.	k.	PROPN
easat-10684	577	12	geetha	geetha	PROPN
easat-10684	577	13	,	,	PUNCT
easat-10684	577	14	"	"	PUNCT
easat-10684	577	15	detecting	detect	VERB
easat-10684	577	16	evolving	evolve	VERB
easat-10684	577	17	fake	fake	ADJ
easat-10684	577	18	news	news	NOUN
easat-10684	577	19	in	in	ADP
easat-10684	577	20	social	social	ADJ
easat-10684	577	21	media	medium	NOUN
easat-10684	577	22	by	by	ADP
easat-10684	577	23	leveraging	leverage	VERB
easat-10684	577	24	heterogeneous	heterogeneous	ADJ
easat-10684	577	25	deep	deep	ADJ
easat-10684	577	26	learning	learning	NOUN
easat-10684	577	27	model	model	NOUN
easat-10684	577	28	,	,	PUNCT
easat-10684	577	29	"	"	PUNCT
easat-10684	577	30	in	in	ADP
easat-10684	577	31	2024	2024	NUM
easat-10684	577	32	second	second	ADJ
easat-10684	577	33	international	international	ADJ
easat-10684	577	34	conference	conference	NOUN
easat-10684	577	35	on	on	ADP
easat-10684	577	36	advances	advance	NOUN
easat-10684	577	37	in	in	ADP
easat-10684	577	38	information	information	NOUN
easat-10684	577	39	technology	technology	NOUN
easat-10684	577	40	(	(	PUNCT
easat-10684	577	41	icait	icait	PROPN
easat-10684	577	42	)	)	PUNCT
easat-10684	577	43	,	,	PUNCT
easat-10684	577	44	2024	2024	NUM
easat-10684	577	45	,	,	PUNCT
easat-10684	577	46	vol	vol	NOUN
easat-10684	577	47	.	.	NOUN
easat-10684	577	48	1	1	NUM
easat-10684	577	49	,	,	PUNCT
easat-10684	577	50	pp	pp	ADJ
easat-10684	577	51	.	.	PUNCT
easat-10684	578	1	1	1	NUM
easat-10684	578	2	-	-	SYM
easat-10684	578	3	5	5	NUM
easat-10684	578	4	.	.	PUNCT
easat-10684	579	1	[	[	X
easat-10684	579	2	32	32	NUM
easat-10684	579	3	]	]	PUNCT
easat-10684	579	4	k.	k.	PROPN
easat-10684	579	5	deepthi	deepthi	PROPN
easat-10684	579	6	and	and	CCONJ
easat-10684	579	7	a.	a.	PROPN
easat-10684	579	8	k.	k.	PROPN
easat-10684	579	9	shastry	shastry	PROPN
easat-10684	579	10	,	,	PUNCT
easat-10684	579	11	"	"	PUNCT
easat-10684	579	12	a	a	DET
easat-10684	579	13	deep	deep	ADJ
easat-10684	579	14	learning	learning	NOUN
easat-10684	579	15	methodology	methodology	NOUN
easat-10684	579	16	-	-	PUNCT
easat-10684	579	17	based	base	VERB
easat-10684	579	18	unsupervised	unsupervised	ADJ
easat-10684	579	19	misinformation	misinformation	NOUN
easat-10684	579	20	detection	detection	NOUN
easat-10684	579	21	(	(	PUNCT
easat-10684	579	22	umd	umd	NOUN
easat-10684	579	23	)	)	PUNCT
easat-10684	579	24	technique	technique	NOUN
easat-10684	579	25	for	for	ADP
easat-10684	579	26	identifying	identify	VERB
easat-10684	579	27	fabricated	fabricate	VERB
easat-10684	579	28	narratives	narrative	NOUN
easat-10684	579	29	,	,	PUNCT
easat-10684	579	30	"	"	PUNCT
easat-10684	579	31	in	in	ADP
easat-10684	579	32	2024	2024	NUM
easat-10684	579	33	second	second	ADJ
easat-10684	579	34	international	international	ADJ
easat-10684	579	35	conference	conference	NOUN
easat-10684	579	36	on	on	ADP
easat-10684	579	37	networks	network	NOUN
easat-10684	579	38	,	,	PUNCT
easat-10684	579	39	multimedia	multimedia	NOUN
easat-10684	579	40	and	and	CCONJ
easat-10684	579	41	information	information	NOUN
easat-10684	579	42	technology	technology	NOUN
easat-10684	579	43	(	(	PUNCT
easat-10684	579	44	nmitcon	nmitcon	PROPN
easat-10684	579	45	)	)	PUNCT
easat-10684	579	46	,	,	PUNCT
easat-10684	579	47	2024	2024	NUM
easat-10684	579	48	:	:	PUNCT
easat-10684	579	49	ieee	ieee	NOUN
easat-10684	579	50	,	,	PUNCT
easat-10684	579	51	pp	pp	ADJ
easat-10684	579	52	.	.	PUNCT
easat-10684	580	1	1	1	NUM
easat-10684	580	2	-	-	SYM
easat-10684	580	3	6	6	NUM
easat-10684	580	4	.	.	PUNCT
easat-10684	581	1	[	[	X
easat-10684	581	2	33	33	NUM
easat-10684	581	3	]	]	PUNCT
easat-10684	581	4	h.	h.	PROPN
easat-10684	581	5	zhu	zhu	PROPN
easat-10684	581	6	,	,	PUNCT
easat-10684	581	7	"	"	PUNCT
easat-10684	581	8	application	application	NOUN
easat-10684	581	9	of	of	ADP
easat-10684	581	10	deep	deep	ADJ
easat-10684	581	11	learning	learning	NOUN
easat-10684	581	12	models	model	NOUN
easat-10684	581	13	in	in	ADP
easat-10684	581	14	the	the	DET
easat-10684	581	15	identification	identification	NOUN
easat-10684	581	16	and	and	CCONJ
easat-10684	581	17	screening	screening	NOUN
easat-10684	581	18	of	of	ADP
easat-10684	581	19	fake	fake	ADJ
easat-10684	581	20	news	news	NOUN
easat-10684	581	21	,	,	PUNCT
easat-10684	581	22	"	"	PUNCT
easat-10684	581	23	applied	apply	VERB
easat-10684	581	24	and	and	CCONJ
easat-10684	581	25	computational	computational	ADJ
easat-10684	581	26	engineering	engineering	NOUN
easat-10684	581	27	,	,	PUNCT
easat-10684	581	28	vol	vol	NOUN
easat-10684	581	29	.	.	PROPN
easat-10684	582	1	87	87	NUM
easat-10684	582	2	,	,	PUNCT
easat-10684	582	3	pp	pp	ADJ
easat-10684	582	4	.	.	PUNCT
easat-10684	583	1	162	162	NUM
easat-10684	583	2	-	-	SYM
easat-10684	583	3	166	166	NUM
easat-10684	583	4	,	,	PUNCT
easat-10684	583	5	2024	2024	NUM
easat-10684	583	6	.	.	PUNCT
easat-10684	584	1	[	[	X
easat-10684	584	2	34	34	NUM
easat-10684	584	3	]	]	X
easat-10684	584	4	r.	r.	PROPN
easat-10684	584	5	ladouceur	ladouceur	PROPN
easat-10684	584	6	,	,	PUNCT
easat-10684	584	7	a.	a.	PROPN
easat-10684	584	8	hassan	hassan	PROPN
easat-10684	584	9	,	,	PUNCT
easat-10684	584	10	m.	m.	NOUN
easat-10684	584	11	mejri	mejri	NOUN
easat-10684	584	12	,	,	PUNCT
easat-10684	584	13	and	and	CCONJ
easat-10684	584	14	f.	f.	PROPN
easat-10684	584	15	jaafar	jaafar	PROPN
easat-10684	584	16	,	,	PUNCT
easat-10684	584	17	"	"	PUNCT
easat-10684	584	18	can	can	AUX
easat-10684	584	19	deep	deep	ADV
easat-10684	584	20	learning	learn	VERB
easat-10684	584	21	detect	detect	VERB
easat-10684	584	22	fake	fake	ADJ
easat-10684	584	23	news	news	NOUN
easat-10684	584	24	better	well	ADV
easat-10684	584	25	when	when	SCONJ
easat-10684	584	26	adding	add	VERB
easat-10684	584	27	context	context	NOUN
easat-10684	584	28	features	feature	NOUN
easat-10684	584	29	?	?	PUNCT
easat-10684	584	30	,	,	PUNCT
easat-10684	584	31	"	"	PUNCT
easat-10684	584	32	presented	present	VERB
easat-10684	584	33	at	at	ADP
easat-10684	584	34	the	the	DET
easat-10684	584	35	2024	2024	NUM
easat-10684	584	36	ieee	ieee	NOUN
easat-10684	584	37	international	international	ADJ
easat-10684	584	38	conference	conference	NOUN
easat-10684	584	39	on	on	ADP
easat-10684	584	40	cyber	cyber	ADJ
easat-10684	584	41	security	security	NOUN
easat-10684	584	42	and	and	CCONJ
easat-10684	584	43	resilience	resilience	NOUN
easat-10684	584	44	(	(	PUNCT
easat-10684	584	45	csr	csr	NOUN
easat-10684	584	46	)	)	PUNCT
easat-10684	584	47	,	,	PUNCT
easat-10684	584	48	2024	2024	NUM
easat-10684	584	49	.	.	PUNCT
easat-10684	585	1	https://doi.org/10.1002/ett.3767	https://doi.org/10.1002/ett.3767	PROPN
easat-10684	585	2	https://doi.org/10.1145/3305260	https://doi.org/10.1145/3305260	NOUN
easat-10684	585	3	https://doi.org/10.1038/s41598-024-76286-0	https://doi.org/10.1038/s41598-024-76286-0	CCONJ
easat-10684	585	4	https://doi.org/10.3390/s24186062	https://doi.org/10.3390/s24186062	NUM
easat-10684	585	5	https://doi.org/10.3390/computers13090239	https://doi.org/10.3390/computers13090239	PROPN
easat-10684	585	6	1540	1540	NUM
easat-10684	585	7	edelweiss	edelweiss	PROPN
easat-10684	585	8	applied	apply	VERB
easat-10684	585	9	science	science	NOUN
easat-10684	585	10	and	and	CCONJ
easat-10684	585	11	technology	technology	NOUN
easat-10684	585	12	issn	issn	PROPN
easat-10684	585	13	:	:	PUNCT
easat-10684	585	14	2576	2576	NUM
easat-10684	585	15	-	-	SYM
easat-10684	585	16	8484	8484	NUM
easat-10684	585	17	vol	vol	NOUN
easat-10684	585	18	.	.	PROPN
easat-10684	586	1	9	9	NUM
easat-10684	586	2	,	,	PUNCT
easat-10684	586	3	no	no	INTJ
easat-10684	586	4	.	.	NOUN
easat-10684	586	5	10	10	NUM
easat-10684	586	6	:	:	PUNCT
easat-10684	586	7	1515	1515	NUM
easat-10684	586	8	-	-	SYM
easat-10684	586	9	1540	1540	NUM
easat-10684	586	10	,	,	PUNCT
easat-10684	586	11	2025	2025	NUM
easat-10684	586	12	doi	doi	NOUN
easat-10684	586	13	:	:	PUNCT
easat-10684	586	14	10.55214/2576	10.55214/2576	NUM
easat-10684	586	15	-	-	SYM
easat-10684	586	16	8484.v9i10.10684	8484.v9i10.10684	NOUN
easat-10684	587	1	©	©	PROPN
easat-10684	587	2	2025	2025	NUM
easat-10684	587	3	by	by	ADP
easat-10684	587	4	the	the	DET
easat-10684	587	5	authors	author	NOUN
easat-10684	587	6	;	;	PUNCT
easat-10684	587	7	licensee	licensee	PROPN
easat-10684	587	8	learning	learning	NOUN
easat-10684	587	9	gate	gate	NOUN
easat-10684	588	1	[	[	X
easat-10684	588	2	35	35	NUM
easat-10684	588	3	]	]	PUNCT
easat-10684	588	4	a.	a.	NOUN
easat-10684	588	5	porya	porya	PROPN
easat-10684	588	6	,	,	PUNCT
easat-10684	588	7	v.	v.	ADP
easat-10684	588	8	singh	singh	PROPN
easat-10684	588	9	,	,	PUNCT
easat-10684	588	10	and	and	CCONJ
easat-10684	588	11	b.	b.	PROPN
easat-10684	588	12	b.	b.	PROPN
easat-10684	588	13	sinha	sinha	PROPN
easat-10684	588	14	,	,	PUNCT
easat-10684	588	15	"	"	PUNCT
easat-10684	588	16	towards	towards	ADP
easat-10684	588	17	robust	robust	ADJ
easat-10684	588	18	fake	fake	ADJ
easat-10684	588	19	news	news	NOUN
easat-10684	588	20	detection	detection	NOUN
easat-10684	588	21	:	:	PUNCT
easat-10684	588	22	an	an	DET
easat-10684	588	23	emperical	emperical	ADJ
easat-10684	588	24	investigation	investigation	NOUN
easat-10684	588	25	of	of	ADP
easat-10684	588	26	preprocessing	preprocessing	NOUN
easat-10684	588	27	and	and	CCONJ
easat-10684	588	28	modeling	model	VERB
easat-10684	588	29	techniques	technique	NOUN
easat-10684	588	30	,	,	PUNCT
easat-10684	588	31	"	"	PUNCT
easat-10684	588	32	in	in	ADP
easat-10684	588	33	2024	2024	NUM
easat-10684	588	34	international	international	ADJ
easat-10684	588	35	conference	conference	NOUN
easat-10684	588	36	on	on	ADP
easat-10684	588	37	computational	computational	ADJ
easat-10684	588	38	intelligence	intelligence	NOUN
easat-10684	588	39	and	and	CCONJ
easat-10684	588	40	computing	computing	NOUN
easat-10684	588	41	applications	application	NOUN
easat-10684	588	42	(	(	PUNCT
easat-10684	588	43	iccica	iccica	PROPN
easat-10684	588	44	)	)	PUNCT
easat-10684	588	45	,	,	PUNCT
easat-10684	588	46	2024	2024	NUM
easat-10684	588	47	,	,	PUNCT
easat-10684	588	48	vol	vol	NOUN
easat-10684	588	49	.	.	NOUN
easat-10684	589	1	1	1	NUM
easat-10684	589	2	:	:	PUNCT
easat-10684	589	3	ieee	ieee	NOUN
easat-10684	589	4	,	,	PUNCT
easat-10684	589	5	pp	pp	PROPN
easat-10684	589	6	.	.	PUNCT
easat-10684	590	1	80	80	NUM
easat-10684	590	2	-	-	SYM
easat-10684	590	3	85	85	NUM
easat-10684	590	4	.	.	PUNCT
easat-10684	591	1	[	[	X
easat-10684	591	2	36	36	NUM
easat-10684	591	3	]	]	X
easat-10684	591	4	j.	j.	PROPN
easat-10684	591	5	b.	b.	PROPN
easat-10684	591	6	awotunde	awotunde	PROPN
easat-10684	591	7	,	,	PUNCT
easat-10684	591	8	r.	r.	PROPN
easat-10684	591	9	g.	g.	PROPN
easat-10684	591	10	jimoh	jimoh	PROPN
easat-10684	591	11	,	,	PUNCT
easat-10684	591	12	a.	a.	PROPN
easat-10684	591	13	l.	l.	PROPN
easat-10684	591	14	imoize	imoize	PROPN
easat-10684	591	15	,	,	PUNCT
easat-10684	591	16	a.	a.	NOUN
easat-10684	591	17	t.	t.	PROPN
easat-10684	591	18	abdulrazaq	abdulrazaq	PROPN
easat-10684	591	19	,	,	PUNCT
easat-10684	591	20	c.-t	c.-t	PROPN
easat-10684	591	21	.	.	PUNCT
easat-10684	592	1	li	li	PROPN
easat-10684	592	2	,	,	PUNCT
easat-10684	592	3	and	and	CCONJ
easat-10684	592	4	c.-c	c.-c	PROPN
easat-10684	592	5	.	.	PUNCT
easat-10684	593	1	lee	lee	PROPN
easat-10684	593	2	,	,	PUNCT
easat-10684	593	3	"	"	PUNCT
easat-10684	593	4	an	an	DET
easat-10684	593	5	enhanced	enhanced	ADJ
easat-10684	593	6	deep	deep	ADJ
easat-10684	593	7	learningbased	learningbase	VERB
easat-10684	593	8	deepfake	deepfake	NOUN
easat-10684	593	9	video	video	NOUN
easat-10684	593	10	detection	detection	NOUN
easat-10684	593	11	and	and	CCONJ
easat-10684	593	12	classification	classification	NOUN
easat-10684	593	13	system	system	NOUN
easat-10684	593	14	,	,	PUNCT
easat-10684	593	15	"	"	PUNCT
easat-10684	593	16	electronics	electronic	NOUN
easat-10684	593	17	,	,	PUNCT
easat-10684	593	18	vol	vol	NOUN
easat-10684	593	19	.	.	PROPN
easat-10684	593	20	12	12	NUM
easat-10684	593	21	,	,	PUNCT
easat-10684	593	22	no	no	INTJ
easat-10684	593	23	.	.	NOUN
easat-10684	593	24	1	1	NUM
easat-10684	593	25	,	,	PUNCT
easat-10684	593	26	p.	p.	NOUN
easat-10684	593	27	87	87	NUM
easat-10684	593	28	,	,	PUNCT
easat-10684	593	29	2022	2022	NUM
easat-10684	593	30	.	.	PUNCT
easat-10684	594	1	https://doi.org/10.3390/electronics12010087	https://doi.org/10.3390/electronics12010087	PROPN
easat-10684	595	1	[	[	X
easat-10684	595	2	37	37	NUM
easat-10684	595	3	]	]	PUNCT
easat-10684	595	4	j.	j.	PROPN
easat-10684	595	5	alghamdi	alghamdi	PROPN
easat-10684	595	6	,	,	PUNCT
easat-10684	595	7	y.	y.	PROPN
easat-10684	595	8	lin	lin	PROPN
easat-10684	595	9	,	,	PUNCT
easat-10684	595	10	and	and	CCONJ
easat-10684	595	11	s.	s.	PROPN
easat-10684	595	12	luo	luo	PROPN
easat-10684	595	13	,	,	PUNCT
easat-10684	595	14	"	"	PUNCT
easat-10684	595	15	towards	towards	ADP
easat-10684	595	16	covid-19	covid-19	PROPN
easat-10684	595	17	fake	fake	ADJ
easat-10684	595	18	news	news	NOUN
easat-10684	595	19	detection	detection	NOUN
easat-10684	595	20	using	use	VERB
easat-10684	595	21	transformer	transformer	NOUN
easat-10684	595	22	-	-	PUNCT
easat-10684	595	23	based	base	VERB
easat-10684	595	24	models	model	NOUN
easat-10684	595	25	,	,	PUNCT
easat-10684	595	26	"	"	PUNCT
easat-10684	595	27	knowledge	knowledge	NOUN
easat-10684	595	28	-	-	PUNCT
easat-10684	595	29	based	base	VERB
easat-10684	595	30	systems	system	NOUN
easat-10684	595	31	,	,	PUNCT
easat-10684	595	32	vol	vol	NOUN
easat-10684	595	33	.	.	PROPN
easat-10684	595	34	274	274	NUM
easat-10684	595	35	,	,	PUNCT
easat-10684	595	36	p.	p.	NOUN
easat-10684	595	37	110642	110642	NUM
easat-10684	595	38	,	,	PUNCT
easat-10684	595	39	2023	2023	NUM
easat-10684	595	40	.	.	PUNCT
easat-10684	596	1	https://doi.org/10.1016/j.knosys.2023.110642	https://doi.org/10.1016/j.knosys.2023.110642	PROPN
easat-10684	596	2	[	[	X
easat-10684	596	3	38	38	NUM
easat-10684	596	4	]	]	PUNCT
easat-10684	596	5	m.	m.	NOUN
easat-10684	596	6	ghayoomi	ghayoomi	NOUN
easat-10684	596	7	and	and	CCONJ
easat-10684	596	8	m.	m.	NOUN
easat-10684	596	9	mousavian	mousavian	PROPN
easat-10684	596	10	,	,	PUNCT
easat-10684	596	11	"	"	PUNCT
easat-10684	596	12	deep	deep	ADJ
easat-10684	596	13	transfer	transfer	NOUN
easat-10684	596	14	learning	learning	NOUN
easat-10684	596	15	for	for	ADP
easat-10684	596	16	covid‐19	covid‐19	PROPN
easat-10684	596	17	fake	fake	ADJ
easat-10684	596	18	news	news	NOUN
easat-10684	596	19	detection	detection	NOUN
easat-10684	596	20	in	in	ADP
easat-10684	596	21	persian	persian	ADJ
easat-10684	596	22	,	,	PUNCT
easat-10684	596	23	"	"	PUNCT
easat-10684	596	24	expert	expert	NOUN
easat-10684	596	25	systems	system	NOUN
easat-10684	596	26	,	,	PUNCT
easat-10684	596	27	vol	vol	NOUN
easat-10684	596	28	.	.	PROPN
easat-10684	596	29	39	39	NUM
easat-10684	596	30	,	,	PUNCT
easat-10684	596	31	no	no	INTJ
easat-10684	596	32	.	.	NOUN
easat-10684	596	33	8	8	NUM
easat-10684	596	34	,	,	PUNCT
easat-10684	596	35	p.	p.	NOUN
easat-10684	596	36	e13008	e13008	NOUN
easat-10684	596	37	,	,	PUNCT
easat-10684	596	38	2022	2022	NUM
easat-10684	596	39	.	.	PUNCT
easat-10684	597	1	https://doi.org/10.1111/exsy.13008	https://doi.org/10.1111/exsy.13008	NOUN
easat-10684	598	1	[	[	X
easat-10684	598	2	39	39	NUM
easat-10684	598	3	]	]	PUNCT
easat-10684	598	4	m.	m.	PROPN
easat-10684	598	5	f.	f.	PROPN
easat-10684	598	6	mridha	mridha	PROPN
easat-10684	598	7	,	,	PUNCT
easat-10684	598	8	a.	a.	PROPN
easat-10684	598	9	j.	j.	PROPN
easat-10684	598	10	keya	keya	PROPN
easat-10684	598	11	,	,	PUNCT
easat-10684	598	12	m.	m.	NOUN
easat-10684	598	13	a.	a.	PROPN
easat-10684	598	14	hamid	hamid	PROPN
easat-10684	598	15	,	,	PUNCT
easat-10684	598	16	m.	m.	NOUN
easat-10684	598	17	m.	m.	NOUN
easat-10684	598	18	monowar	monowar	PROPN
easat-10684	598	19	,	,	PUNCT
easat-10684	598	20	and	and	CCONJ
easat-10684	598	21	m.	m.	PROPN
easat-10684	598	22	s.	s.	PROPN
easat-10684	598	23	rahman	rahman	PROPN
easat-10684	598	24	,	,	PUNCT
easat-10684	598	25	"	"	PUNCT
easat-10684	598	26	a	a	DET
easat-10684	598	27	comprehensive	comprehensive	ADJ
easat-10684	598	28	review	review	NOUN
easat-10684	598	29	on	on	ADP
easat-10684	598	30	fake	fake	ADJ
easat-10684	598	31	news	news	NOUN
easat-10684	598	32	detection	detection	NOUN
easat-10684	598	33	with	with	ADP
easat-10684	598	34	deep	deep	ADJ
easat-10684	598	35	learning	learning	NOUN
easat-10684	598	36	,	,	PUNCT
easat-10684	598	37	"	"	PUNCT
easat-10684	598	38	ieee	ieee	NOUN
easat-10684	598	39	access	access	NOUN
easat-10684	598	40	,	,	PUNCT
easat-10684	598	41	vol	vol	NOUN
easat-10684	598	42	.	.	NOUN
easat-10684	598	43	9	9	NUM
easat-10684	598	44	,	,	PUNCT
easat-10684	598	45	pp	pp	ADJ
easat-10684	598	46	.	.	PUNCT
easat-10684	599	1	156151	156151	NUM
easat-10684	599	2	-	-	SYM
easat-10684	599	3	156170	156170	NUM
easat-10684	599	4	,	,	PUNCT
easat-10684	599	5	2021	2021	NUM
easat-10684	599	6	.	.	PUNCT
easat-10684	600	1	[	[	X
easat-10684	600	2	40	40	NUM
easat-10684	600	3	]	]	PUNCT
easat-10684	600	4	j.	j.	PROPN
easat-10684	600	5	y.	y.	PROPN
easat-10684	600	6	khan	khan	PROPN
easat-10684	600	7	,	,	PUNCT
easat-10684	600	8	m.	m.	NOUN
easat-10684	600	9	t.	t.	PROPN
easat-10684	600	10	i.	i.	PROPN
easat-10684	600	11	khondaker	khondaker	PROPN
easat-10684	600	12	,	,	PUNCT
easat-10684	600	13	s.	s.	PROPN
easat-10684	600	14	afroz	afroz	PROPN
easat-10684	600	15	,	,	PUNCT
easat-10684	600	16	g.	g.	PROPN
easat-10684	600	17	uddin	uddin	PROPN
easat-10684	600	18	,	,	PUNCT
easat-10684	600	19	and	and	CCONJ
easat-10684	600	20	a.	a.	PROPN
easat-10684	600	21	iqbal	iqbal	PROPN
easat-10684	600	22	,	,	PUNCT
easat-10684	600	23	"	"	PUNCT
easat-10684	600	24	a	a	DET
easat-10684	600	25	benchmark	benchmark	ADJ
easat-10684	600	26	study	study	NOUN
easat-10684	600	27	of	of	ADP
easat-10684	600	28	machine	machine	NOUN
easat-10684	600	29	learning	learning	NOUN
easat-10684	600	30	models	model	NOUN
easat-10684	600	31	for	for	ADP
easat-10684	600	32	online	online	ADJ
easat-10684	600	33	fake	fake	ADJ
easat-10684	600	34	news	news	NOUN
easat-10684	600	35	detection	detection	NOUN
easat-10684	600	36	,	,	PUNCT
easat-10684	600	37	"	"	PUNCT
easat-10684	600	38	machine	machine	NOUN
easat-10684	600	39	learning	learn	VERB
easat-10684	600	40	with	with	ADP
easat-10684	600	41	applications	application	NOUN
easat-10684	600	42	,	,	PUNCT
easat-10684	600	43	vol	vol	NOUN
easat-10684	600	44	.	.	PROPN
easat-10684	600	45	4	4	NUM
easat-10684	600	46	,	,	PUNCT
easat-10684	600	47	p.	p.	NOUN
easat-10684	600	48	100032	100032	NUM
easat-10684	600	49	,	,	PUNCT
easat-10684	600	50	2021	2021	NUM
easat-10684	600	51	.	.	PUNCT
easat-10684	601	1	https://doi.org/10.1016/j.mlwa.2021.100032	https://doi.org/10.1016/j.mlwa.2021.100032	NOUN
easat-10684	602	1	[	[	X
easat-10684	602	2	41	41	NUM
easat-10684	602	3	]	]	PUNCT
easat-10684	602	4	j.	j.	PROPN
easat-10684	602	5	alghamdi	alghamdi	PROPN
easat-10684	602	6	,	,	PUNCT
easat-10684	602	7	y.	y.	PROPN
easat-10684	602	8	lin	lin	PROPN
easat-10684	602	9	,	,	PUNCT
easat-10684	602	10	and	and	CCONJ
easat-10684	602	11	s.	s.	PROPN
easat-10684	602	12	luo	luo	PROPN
easat-10684	602	13	,	,	PUNCT
easat-10684	602	14	"	"	PUNCT
easat-10684	602	15	a	a	DET
easat-10684	602	16	comparative	comparative	ADJ
easat-10684	602	17	study	study	NOUN
easat-10684	602	18	of	of	ADP
easat-10684	602	19	machine	machine	NOUN
easat-10684	602	20	learning	learning	NOUN
easat-10684	602	21	and	and	CCONJ
easat-10684	602	22	deep	deep	ADJ
easat-10684	602	23	learning	learning	NOUN
easat-10684	602	24	techniques	technique	NOUN
easat-10684	602	25	for	for	ADP
easat-10684	602	26	fake	fake	ADJ
easat-10684	602	27	news	news	NOUN
easat-10684	602	28	detection	detection	NOUN
easat-10684	602	29	,	,	PUNCT
easat-10684	602	30	"	"	PUNCT
easat-10684	602	31	information	information	NOUN
easat-10684	602	32	,	,	PUNCT
easat-10684	602	33	vol	vol	NOUN
easat-10684	602	34	.	.	PROPN
easat-10684	602	35	13	13	NUM
easat-10684	602	36	,	,	PUNCT
easat-10684	602	37	no	no	INTJ
easat-10684	602	38	.	.	NOUN
easat-10684	602	39	12	12	NUM
easat-10684	602	40	,	,	PUNCT
easat-10684	602	41	p.	p.	NOUN
easat-10684	602	42	576	576	NUM
easat-10684	602	43	,	,	PUNCT
easat-10684	602	44	2022	2022	NUM
easat-10684	602	45	.	.	PUNCT
easat-10684	603	1	https://doi.org/10.3390/info13120576	https://doi.org/10.3390/info13120576	PROPN
easat-10684	604	1	[	[	X
easat-10684	604	2	42	42	NUM
easat-10684	604	3	]	]	PUNCT
easat-10684	604	4	m.	m.	NOUN
easat-10684	604	5	q.	q.	PROPN
easat-10684	604	6	alnabhan	alnabhan	PROPN
easat-10684	604	7	and	and	CCONJ
easat-10684	604	8	p.	p.	PROPN
easat-10684	604	9	branco	branco	PROPN
easat-10684	604	10	,	,	PUNCT
easat-10684	604	11	"	"	PUNCT
easat-10684	604	12	fake	fake	ADJ
easat-10684	604	13	news	news	NOUN
easat-10684	604	14	detection	detection	NOUN
easat-10684	604	15	using	use	VERB
easat-10684	604	16	deep	deep	ADJ
easat-10684	604	17	learning	learning	NOUN
easat-10684	604	18	:	:	PUNCT
easat-10684	604	19	a	a	DET
easat-10684	604	20	systematic	systematic	ADJ
easat-10684	604	21	literature	literature	NOUN
easat-10684	604	22	review	review	NOUN
easat-10684	604	23	,	,	PUNCT
easat-10684	604	24	"	"	PUNCT
easat-10684	604	25	ieee	ieee	NOUN
easat-10684	604	26	access	access	NOUN
easat-10684	604	27	,	,	PUNCT
easat-10684	604	28	vol	vol	NOUN
easat-10684	604	29	.	.	PROPN
easat-10684	604	30	12	12	NUM
easat-10684	604	31	,	,	PUNCT
easat-10684	604	32	pp	pp	ADJ
easat-10684	604	33	.	.	PUNCT
easat-10684	605	1	114435	114435	NUM
easat-10684	605	2	114459	114459	NUM
easat-10684	605	3	,	,	PUNCT
easat-10684	605	4	2024	2024	NUM
easat-10684	605	5	.	.	PUNCT
easat-10684	606	1	https://doi.org/10.1109/access.2024.3435497	https://doi.org/10.1109/access.2024.3435497	PROPN
easat-10684	606	2	[	[	X
easat-10684	606	3	43	43	NUM
easat-10684	606	4	]	]	PUNCT
easat-10684	606	5	m.	m.	NOUN
easat-10684	606	6	h.	h.	PROPN
easat-10684	606	7	al	al	PROPN
easat-10684	606	8	-	-	PUNCT
easat-10684	606	9	tai	tai	PROPN
easat-10684	606	10	,	,	PUNCT
easat-10684	606	11	b.	b.	PROPN
easat-10684	606	12	m.	m.	PROPN
easat-10684	606	13	nema	nema	PROPN
easat-10684	606	14	,	,	PUNCT
easat-10684	606	15	and	and	CCONJ
easat-10684	606	16	a.	a.	PROPN
easat-10684	606	17	al	al	PROPN
easat-10684	606	18	-	-	PUNCT
easat-10684	606	19	sherbaz	sherbaz	PROPN
easat-10684	606	20	,	,	PUNCT
easat-10684	606	21	"	"	PUNCT
easat-10684	606	22	deep	deep	ADJ
easat-10684	606	23	learning	learning	NOUN
easat-10684	606	24	for	for	ADP
easat-10684	606	25	fake	fake	ADJ
easat-10684	606	26	news	news	NOUN
easat-10684	606	27	detection	detection	NOUN
easat-10684	606	28	:	:	PUNCT
easat-10684	606	29	literature	literature	PROPN
easat-10684	606	30	review	review	PROPN
easat-10684	606	31	,	,	PUNCT
easat-10684	606	32	"	"	PUNCT
easat-10684	606	33	almustansiriyah	almustansiriyah	ADJ
easat-10684	606	34	journal	journal	NOUN
easat-10684	606	35	of	of	ADP
easat-10684	606	36	science	science	NOUN
easat-10684	606	37	,	,	PUNCT
easat-10684	606	38	vol	vol	NOUN
easat-10684	606	39	.	.	PROPN
easat-10684	607	1	34	34	NUM
easat-10684	607	2	,	,	PUNCT
easat-10684	607	3	no	no	INTJ
easat-10684	607	4	.	.	NOUN
easat-10684	607	5	2	2	NUM
easat-10684	607	6	,	,	PUNCT
easat-10684	607	7	pp	pp	ADJ
easat-10684	607	8	.	.	PUNCT
easat-10684	608	1	70	70	NUM
easat-10684	608	2	-	-	SYM
easat-10684	608	3	81	81	NUM
easat-10684	608	4	,	,	PUNCT
easat-10684	608	5	2023	2023	NUM
easat-10684	608	6	.	.	PUNCT
easat-10684	609	1	https://doi.org/10.23851/mjs.v34i2.1292	https://doi.org/10.23851/mjs.v34i2.1292	PROPN
easat-10684	609	2	[	[	X
easat-10684	609	3	44	44	NUM
easat-10684	609	4	]	]	PUNCT
easat-10684	609	5	m.	m.	NOUN
easat-10684	609	6	villagracia	villagracia	PROPN
easat-10684	609	7	octaviano	octaviano	NOUN
easat-10684	609	8	,	,	PUNCT
easat-10684	609	9	"	"	PUNCT
easat-10684	609	10	fake	fake	ADJ
easat-10684	609	11	news	news	NOUN
easat-10684	609	12	detection	detection	NOUN
easat-10684	609	13	using	use	VERB
easat-10684	609	14	machine	machine	NOUN
easat-10684	609	15	learning	learning	NOUN
easat-10684	609	16	,	,	PUNCT
easat-10684	609	17	"	"	PUNCT
easat-10684	609	18	in	in	ADP
easat-10684	609	19	2021	2021	NUM
easat-10684	609	20	5th	5th	ADJ
easat-10684	609	21	international	international	ADJ
easat-10684	609	22	conference	conference	NOUN
easat-10684	609	23	on	on	ADP
easat-10684	609	24	esociety	esociety	NOUN
easat-10684	609	25	,	,	PUNCT
easat-10684	609	26	e	e	NOUN
easat-10684	609	27	-	-	NOUN
easat-10684	609	28	education	education	NOUN
easat-10684	609	29	and	and	CCONJ
easat-10684	609	30	e	e	NOUN
easat-10684	609	31	-	-	NOUN
easat-10684	609	32	technology	technology	NOUN
easat-10684	609	33	,	,	PUNCT
easat-10684	609	34	2021	2021	NUM
easat-10684	609	35	,	,	PUNCT
easat-10684	609	36	pp	pp	ADJ
easat-10684	609	37	.	.	PUNCT
easat-10684	610	1	177	177	NUM
easat-10684	610	2	-	-	SYM
easat-10684	610	3	180	180	NUM
easat-10684	610	4	.	.	PUNCT
easat-10684	611	1	[	[	X
easat-10684	611	2	45	45	NUM
easat-10684	611	3	]	]	X
easat-10684	611	4	e.	e.	PROPN
easat-10684	611	5	m.	m.	PROPN
easat-10684	611	6	sabido	sabido	PROPN
easat-10684	611	7	et	et	PROPN
easat-10684	611	8	al	al	PROPN
easat-10684	611	9	.	.	PROPN
easat-10684	611	10	,	,	PUNCT
easat-10684	611	11	"	"	PUNCT
easat-10684	611	12	insights	insight	NOUN
easat-10684	611	13	into	into	ADP
easat-10684	611	14	the	the	DET
easat-10684	611	15	variation	variation	NOUN
easat-10684	611	16	in	in	ADP
easat-10684	611	17	bioactivities	bioactivitie	NOUN
easat-10684	611	18	of	of	ADP
easat-10684	611	19	closely	closely	ADV
easat-10684	611	20	related	relate	VERB
easat-10684	611	21	streptomyces	streptomyce	NOUN
easat-10684	611	22	strains	strain	NOUN
easat-10684	611	23	from	from	ADP
easat-10684	611	24	marine	marine	ADJ
easat-10684	611	25	sediments	sediment	NOUN
easat-10684	611	26	of	of	ADP
easat-10684	611	27	the	the	DET
easat-10684	611	28	visayan	visayan	ADJ
easat-10684	611	29	sea	sea	NOUN
easat-10684	611	30	against	against	ADP
easat-10684	611	31	eskape	eskape	NOUN
easat-10684	611	32	and	and	CCONJ
easat-10684	611	33	ovarian	ovarian	ADJ
easat-10684	611	34	cancer	cancer	NOUN
easat-10684	611	35	,	,	PUNCT
easat-10684	611	36	"	"	PUNCT
easat-10684	611	37	marine	marine	ADJ
easat-10684	611	38	drugs	drug	NOUN
easat-10684	611	39	,	,	PUNCT
easat-10684	611	40	vol	vol	NOUN
easat-10684	611	41	.	.	PROPN
easat-10684	611	42	19	19	NUM
easat-10684	611	43	,	,	PUNCT
easat-10684	611	44	no	no	INTJ
easat-10684	611	45	.	.	NOUN
easat-10684	611	46	8	8	NUM
easat-10684	611	47	,	,	PUNCT
easat-10684	611	48	p.	p.	NOUN
easat-10684	611	49	441	441	NUM
easat-10684	611	50	,	,	PUNCT
easat-10684	611	51	2021	2021	NUM
easat-10684	611	52	.	.	PUNCT
easat-10684	612	1	[	[	X
easat-10684	612	2	46	46	NUM
easat-10684	612	3	]	]	PUNCT
easat-10684	612	4	z.	z.	PROPN
easat-10684	612	5	khanam	khanam	PROPN
easat-10684	612	6	,	,	PUNCT
easat-10684	612	7	b.	b.	PROPN
easat-10684	612	8	n.	n.	PROPN
easat-10684	612	9	alwasel	alwasel	PROPN
easat-10684	612	10	,	,	PUNCT
easat-10684	612	11	h.	h.	PROPN
easat-10684	612	12	sirafi	sirafi	PROPN
easat-10684	612	13	,	,	PUNCT
easat-10684	612	14	and	and	CCONJ
easat-10684	612	15	m.	m.	NOUN
easat-10684	612	16	rashid	rashid	PROPN
easat-10684	612	17	,	,	PUNCT
easat-10684	612	18	"	"	PUNCT
easat-10684	612	19	fake	fake	ADJ
easat-10684	612	20	news	news	NOUN
easat-10684	612	21	detection	detection	NOUN
easat-10684	612	22	using	use	VERB
easat-10684	612	23	machine	machine	NOUN
easat-10684	612	24	learning	learning	NOUN
easat-10684	612	25	approaches	approach	NOUN
easat-10684	612	26	,	,	PUNCT
easat-10684	612	27	"	"	PUNCT
easat-10684	612	28	iop	iop	PROPN
easat-10684	612	29	conference	conference	NOUN
easat-10684	612	30	series	series	NOUN
easat-10684	612	31	:	:	PUNCT
easat-10684	612	32	materials	material	NOUN
easat-10684	612	33	science	science	NOUN
easat-10684	612	34	and	and	CCONJ
easat-10684	612	35	engineering	engineering	NOUN
easat-10684	612	36	,	,	PUNCT
easat-10684	612	37	vol	vol	NOUN
easat-10684	612	38	.	.	PROPN
easat-10684	612	39	1099	1099	NUM
easat-10684	612	40	,	,	PUNCT
easat-10684	612	41	no	no	INTJ
easat-10684	612	42	.	.	NOUN
easat-10684	612	43	1	1	NUM
easat-10684	612	44	,	,	PUNCT
easat-10684	612	45	p.	p.	NOUN
easat-10684	612	46	012040	012040	NUM
easat-10684	612	47	,	,	PUNCT
easat-10684	612	48	2021	2021	NUM
easat-10684	612	49	.	.	PUNCT
easat-10684	613	1	[	[	X
easat-10684	613	2	47	47	NUM
easat-10684	613	3	]	]	X
easat-10684	613	4	s.	s.	PROPN
easat-10684	613	5	sangamnerkar	sangamnerkar	PROPN
easat-10684	613	6	,	,	PUNCT
easat-10684	613	7	r.	r.	PROPN
easat-10684	613	8	srinivasan	srinivasan	PROPN
easat-10684	613	9	,	,	PUNCT
easat-10684	613	10	m.	m.	NOUN
easat-10684	613	11	christhuraj	christhuraj	NOUN
easat-10684	613	12	,	,	PUNCT
easat-10684	613	13	and	and	CCONJ
easat-10684	613	14	r.	r.	PROPN
easat-10684	613	15	sukumaran	sukumaran	PROPN
easat-10684	613	16	,	,	PUNCT
easat-10684	613	17	"	"	PUNCT
easat-10684	613	18	an	an	DET
easat-10684	613	19	ensemble	ensemble	ADJ
easat-10684	613	20	technique	technique	NOUN
easat-10684	613	21	to	to	PART
easat-10684	613	22	detect	detect	VERB
easat-10684	613	23	fabricated	fabricate	VERB
easat-10684	613	24	news	news	NOUN
easat-10684	613	25	article	article	NOUN
easat-10684	613	26	using	use	VERB
easat-10684	613	27	machine	machine	NOUN
easat-10684	613	28	learning	learning	NOUN
easat-10684	613	29	and	and	CCONJ
easat-10684	613	30	natural	natural	ADJ
easat-10684	613	31	language	language	NOUN
easat-10684	613	32	processing	processing	NOUN
easat-10684	613	33	techniques	technique	NOUN
easat-10684	613	34	,	,	PUNCT
easat-10684	613	35	"	"	PUNCT
easat-10684	613	36	in	in	ADP
easat-10684	613	37	2020	2020	NUM
easat-10684	613	38	international	international	ADJ
easat-10684	613	39	conference	conference	NOUN
easat-10684	613	40	for	for	ADP
easat-10684	613	41	emerging	emerge	VERB
easat-10684	613	42	technology	technology	NOUN
easat-10684	613	43	(	(	PUNCT
easat-10684	613	44	incet	incet	PROPN
easat-10684	613	45	)	)	PUNCT
easat-10684	613	46	,	,	PUNCT
easat-10684	613	47	pp	pp	PROPN
easat-10684	613	48	.	.	PUNCT
easat-10684	614	1	1	1	NUM
easat-10684	614	2	-	-	SYM
easat-10684	614	3	7	7	NUM
easat-10684	614	4	.	.	PUNCT
easat-10684	614	5	ieee	ieee	NOUN
easat-10684	614	6	,	,	PUNCT
easat-10684	614	7	2020	2020	NUM
easat-10684	614	8	.	.	PUNCT
easat-10684	615	1	[	[	X
easat-10684	615	2	48	48	NUM
easat-10684	615	3	]	]	PUNCT
easat-10684	615	4	a.	a.	NOUN
easat-10684	615	5	i.	i.	PROPN
easat-10684	615	6	kadhim	kadhim	PROPN
easat-10684	615	7	,	,	PUNCT
easat-10684	615	8	"	"	PUNCT
easat-10684	615	9	an	an	DET
easat-10684	615	10	evaluation	evaluation	NOUN
easat-10684	615	11	of	of	ADP
easat-10684	615	12	preprocessing	preprocesse	VERB
easat-10684	615	13	techniques	technique	NOUN
easat-10684	615	14	for	for	ADP
easat-10684	615	15	text	text	NOUN
easat-10684	615	16	classification	classification	NOUN
easat-10684	615	17	,	,	PUNCT
easat-10684	615	18	"	"	PUNCT
easat-10684	615	19	international	international	ADJ
easat-10684	615	20	journal	journal	NOUN
easat-10684	615	21	of	of	ADP
easat-10684	615	22	computer	computer	NOUN
easat-10684	615	23	science	science	NOUN
easat-10684	615	24	and	and	CCONJ
easat-10684	615	25	information	information	NOUN
easat-10684	615	26	security	security	NOUN
easat-10684	615	27	,	,	PUNCT
easat-10684	615	28	vol	vol	NOUN
easat-10684	615	29	.	.	PROPN
easat-10684	615	30	16	16	NUM
easat-10684	615	31	,	,	PUNCT
easat-10684	615	32	no	no	INTJ
easat-10684	615	33	.	.	NOUN
easat-10684	615	34	6	6	NUM
easat-10684	615	35	,	,	PUNCT
easat-10684	615	36	pp	pp	ADJ
easat-10684	615	37	.	.	PUNCT
easat-10684	616	1	22	22	NUM
easat-10684	616	2	-	-	SYM
easat-10684	616	3	32	32	NUM
easat-10684	616	4	,	,	PUNCT
easat-10684	616	5	2018	2018	NUM
easat-10684	616	6	.	.	PUNCT
easat-10684	617	1	https://doi.org/10.3390/electronics12010087	https://doi.org/10.3390/electronics12010087	PROPN
easat-10684	617	2	https://doi.org/10.1016/j.knosys.2023.110642	https://doi.org/10.1016/j.knosys.2023.110642	PROPN
easat-10684	617	3	https://doi.org/10.1111/exsy.13008	https://doi.org/10.1111/exsy.13008	NOUN
easat-10684	617	4	https://doi.org/10.1016/j.mlwa.2021.100032	https://doi.org/10.1016/j.mlwa.2021.100032	NOUN
easat-10684	617	5	https://doi.org/10.3390/info13120576	https://doi.org/10.3390/info13120576	NOUN
easat-10684	618	1	https://doi.org/10.1109/access.2024.3435497	https://doi.org/10.1109/access.2024.3435497	PROPN
easat-10684	618	2	https://doi.org/10.23851/mjs.v34i2.1292	https://doi.org/10.23851/mjs.v34i2.1292	PROPN