id	sid	tid	token	lemma	pos
cana-4362	1	1	communications	communication	NOUN
cana-4362	1	2	on	on	ADP
cana-4362	1	3	applied	apply	VERB
cana-4362	1	4	nonlinear	nonlinear	ADJ
cana-4362	1	5	analysis	analysis	NOUN
cana-4362	1	6	issn	issn	NOUN
cana-4362	1	7	:	:	PUNCT
cana-4362	1	8	1074	1074	NUM
cana-4362	1	9	-	-	PUNCT
cana-4362	1	10	133x	133x	NUM
cana-4362	1	11	vol	vol	NOUN
cana-4362	1	12	32	32	NUM
cana-4362	1	13	no	no	NOUN
cana-4362	1	14	.	.	PUNCT
cana-4362	2	1	9s	9s	NUM
cana-4362	2	2	(	(	PUNCT
cana-4362	2	3	2025	2025	NUM
cana-4362	2	4	)	)	PUNCT
cana-4362	2	5	1899	1899	NUM
cana-4362	3	1	https://internationalpubls.com	https://internationalpubls.com	X
cana-4362	3	2	hybrid	hybrid	ADJ
cana-4362	3	3	feature	feature	NOUN
cana-4362	3	4	selection	selection	NOUN
cana-4362	3	5	on	on	ADP
cana-4362	3	6	social	social	ADJ
cana-4362	3	7	media	medium	NOUN
cana-4362	3	8	dataset	dataset	VERB
cana-4362	3	9	for	for	ADP
cana-4362	3	10	sentiment	sentiment	NOUN
cana-4362	3	11	classification	classification	NOUN
cana-4362	3	12	using	use	VERB
cana-4362	3	13	deep	deep	ADJ
cana-4362	3	14	learning	learning	NOUN
cana-4362	3	15	techniques	technique	NOUN
cana-4362	3	16	1sudeep	1sudeep	NUM
cana-4362	3	17	k.	k.	PROPN
cana-4362	3	18	hase	hase	PROPN
cana-4362	3	19	,	,	PUNCT
cana-4362	3	20	2dr	2dr	NOUN
cana-4362	3	21	.	.	PUNCT
cana-4362	4	1	rashmi	rashmi	PROPN
cana-4362	4	2	soni	soni	PROPN
cana-4362	4	3	1research	1research	NUM
cana-4362	4	4	scholar	scholar	NOUN
cana-4362	4	5	,	,	PUNCT
cana-4362	4	6	department	department	NOUN
cana-4362	4	7	of	of	ADP
cana-4362	4	8	cse	cse	PROPN
cana-4362	4	9	,	,	PUNCT
cana-4362	4	10	oriental	oriental	ADJ
cana-4362	4	11	university	university	PROPN
cana-4362	4	12	,	,	PUNCT
cana-4362	4	13	indore	indore	PROPN
cana-4362	4	14	,	,	PUNCT
cana-4362	4	15	india	india	PROPN
cana-4362	4	16	2research	2research	NUM
cana-4362	4	17	supervisor	supervisor	NOUN
cana-4362	4	18	,	,	PUNCT
cana-4362	4	19	department	department	NOUN
cana-4362	4	20	of	of	ADP
cana-4362	4	21	cse	cse	PROPN
cana-4362	4	22	,	,	PUNCT
cana-4362	4	23	oriental	oriental	ADJ
cana-4362	4	24	university	university	PROPN
cana-4362	4	25	,	,	PUNCT
cana-4362	4	26	indore	indore	PROPN
cana-4362	4	27	,	,	PUNCT
cana-4362	4	28	india	india	PROPN
cana-4362	4	29	1hase.sudeep@gmail.com	1hase.sudeep@gmail.com	NUM
cana-4362	4	30	,	,	PUNCT
cana-4362	4	31	2drrashmicseofficial@gmail.com	2drrashmicseofficial@gmail.com	NUM
cana-4362	4	32	article	article	NOUN
cana-4362	4	33	history	history	NOUN
cana-4362	4	34	:	:	PUNCT
cana-4362	4	35	received	receive	VERB
cana-4362	4	36	:	:	PUNCT
cana-4362	4	37	12	12	NUM
cana-4362	4	38	-	-	SYM
cana-4362	4	39	01	01	NUM
cana-4362	4	40	-	-	PUNCT
cana-4362	4	41	2025	2025	NUM
cana-4362	4	42	revised	revise	VERB
cana-4362	4	43	:	:	PUNCT
cana-4362	4	44	15	15	NUM
cana-4362	4	45	-	-	NUM
cana-4362	4	46	02	02	NUM
cana-4362	4	47	-	-	PUNCT
cana-4362	4	48	2025	2025	NUM
cana-4362	4	49	accepted	accept	VERB
cana-4362	4	50	:	:	PUNCT
cana-4362	4	51	01	01	NUM
cana-4362	4	52	-	-	SYM
cana-4362	4	53	03	03	NUM
cana-4362	4	54	-	-	PUNCT
cana-4362	4	55	2025	2025	NUM
cana-4362	4	56	abstract	abstract	NOUN
cana-4362	4	57	:	:	PUNCT
cana-4362	4	58	sentiment	sentiment	NOUN
cana-4362	4	59	classification	classification	NOUN
cana-4362	4	60	involves	involve	VERB
cana-4362	4	61	determining	determine	VERB
cana-4362	4	62	the	the	DET
cana-4362	4	63	sentiment	sentiment	NOUN
cana-4362	4	64	expressed	express	VERB
cana-4362	4	65	in	in	ADP
cana-4362	4	66	text	text	NOUN
cana-4362	4	67	,	,	PUNCT
cana-4362	4	68	such	such	ADJ
cana-4362	4	69	as	as	ADP
cana-4362	4	70	positive	positive	ADJ
cana-4362	4	71	,	,	PUNCT
cana-4362	4	72	negative	negative	ADJ
cana-4362	4	73	,	,	PUNCT
cana-4362	4	74	or	or	CCONJ
cana-4362	4	75	neutral	neutral	ADJ
cana-4362	4	76	,	,	PUNCT
cana-4362	4	77	but	but	CCONJ
cana-4362	4	78	social	social	ADJ
cana-4362	4	79	media	medium	NOUN
cana-4362	4	80	data	datum	NOUN
cana-4362	4	81	presents	present	VERB
cana-4362	4	82	challenges	challenge	NOUN
cana-4362	4	83	due	due	ADP
cana-4362	4	84	to	to	ADP
cana-4362	4	85	its	its	PRON
cana-4362	4	86	high	high	ADJ
cana-4362	4	87	dimensionality	dimensionality	NOUN
cana-4362	4	88	,	,	PUNCT
cana-4362	4	89	noise	noise	NOUN
cana-4362	4	90	,	,	PUNCT
cana-4362	4	91	and	and	CCONJ
cana-4362	4	92	unstructured	unstructured	ADJ
cana-4362	4	93	nature	nature	NOUN
cana-4362	4	94	.	.	PUNCT
cana-4362	5	1	this	this	DET
cana-4362	5	2	study	study	NOUN
cana-4362	5	3	proposes	propose	VERB
cana-4362	5	4	a	a	DET
cana-4362	5	5	novel	novel	ADJ
cana-4362	5	6	sentiment	sentiment	NOUN
cana-4362	5	7	classification	classification	NOUN
cana-4362	5	8	approach	approach	NOUN
cana-4362	5	9	by	by	ADP
cana-4362	5	10	combining	combine	VERB
cana-4362	5	11	hybrid	hybrid	ADJ
cana-4362	5	12	feature	feature	NOUN
cana-4362	5	13	selection	selection	NOUN
cana-4362	5	14	methods	method	NOUN
cana-4362	5	15	with	with	ADP
cana-4362	5	16	deep	deep	ADJ
cana-4362	5	17	learning	learning	NOUN
cana-4362	5	18	techniques	technique	NOUN
cana-4362	5	19	.	.	PUNCT
cana-4362	6	1	social	social	ADJ
cana-4362	6	2	media	medium	NOUN
cana-4362	6	3	platforms	platform	NOUN
cana-4362	6	4	generate	generate	VERB
cana-4362	6	5	vast	vast	ADJ
cana-4362	6	6	amounts	amount	NOUN
cana-4362	6	7	of	of	ADP
cana-4362	6	8	data	datum	NOUN
cana-4362	6	9	daily	daily	ADV
cana-4362	6	10	,	,	PUNCT
cana-4362	6	11	which	which	PRON
cana-4362	6	12	is	be	AUX
cana-4362	6	13	often	often	ADV
cana-4362	6	14	noisy	noisy	ADJ
cana-4362	6	15	,	,	PUNCT
cana-4362	6	16	redundant	redundant	ADJ
cana-4362	6	17	,	,	PUNCT
cana-4362	6	18	and	and	CCONJ
cana-4362	6	19	irrelevant	irrelevant	ADJ
cana-4362	6	20	for	for	ADP
cana-4362	6	21	sentiment	sentiment	NOUN
cana-4362	6	22	analysis	analysis	NOUN
cana-4362	6	23	.	.	PUNCT
cana-4362	7	1	hybrid	hybrid	ADJ
cana-4362	7	2	feature	feature	NOUN
cana-4362	7	3	selection	selection	NOUN
cana-4362	7	4	techniques	technique	NOUN
cana-4362	7	5	,	,	PUNCT
cana-4362	7	6	which	which	PRON
cana-4362	7	7	integrate	integrate	VERB
cana-4362	7	8	filter	filter	NOUN
cana-4362	7	9	and	and	CCONJ
cana-4362	7	10	wrapper	wrapper	NOUN
cana-4362	7	11	-	-	PUNCT
cana-4362	7	12	based	base	VERB
cana-4362	7	13	methods	method	NOUN
cana-4362	7	14	,	,	PUNCT
cana-4362	7	15	assist	assist	VERB
cana-4362	7	16	in	in	ADP
cana-4362	7	17	reducing	reduce	VERB
cana-4362	7	18	the	the	DET
cana-4362	7	19	feature	feature	NOUN
cana-4362	7	20	space	space	NOUN
cana-4362	7	21	while	while	SCONJ
cana-4362	7	22	retaining	retain	VERB
cana-4362	7	23	the	the	DET
cana-4362	7	24	most	most	ADV
cana-4362	7	25	informative	informative	ADJ
cana-4362	7	26	features	feature	NOUN
cana-4362	7	27	.	.	PUNCT
cana-4362	8	1	by	by	ADP
cana-4362	8	2	applying	apply	VERB
cana-4362	8	3	deep	deep	ADJ
cana-4362	8	4	learning	learning	NOUN
cana-4362	8	5	models	model	NOUN
cana-4362	8	6	,	,	PUNCT
cana-4362	8	7	such	such	ADJ
cana-4362	8	8	as	as	ADP
cana-4362	8	9	recurrent	recurrent	ADJ
cana-4362	8	10	neural	neural	ADJ
cana-4362	8	11	networks	network	NOUN
cana-4362	8	12	(	(	PUNCT
cana-4362	8	13	rnns	rnns	PROPN
cana-4362	8	14	)	)	PUNCT
cana-4362	8	15	and	and	CCONJ
cana-4362	8	16	long	long	ADJ
cana-4362	8	17	short	short	ADJ
cana-4362	8	18	-	-	PUNCT
cana-4362	8	19	term	term	NOUN
cana-4362	8	20	memory	memory	NOUN
cana-4362	8	21	(	(	PUNCT
cana-4362	8	22	lstm	lstm	NOUN
cana-4362	8	23	)	)	PUNCT
cana-4362	8	24	networks	network	NOUN
cana-4362	8	25	,	,	PUNCT
cana-4362	8	26	classification	classification	NOUN
cana-4362	8	27	performance	performance	NOUN
cana-4362	8	28	can	can	AUX
cana-4362	8	29	be	be	AUX
cana-4362	8	30	substantially	substantially	ADV
cana-4362	8	31	enhanced	enhance	VERB
cana-4362	8	32	.	.	PUNCT
cana-4362	9	1	the	the	DET
cana-4362	9	2	proposed	propose	VERB
cana-4362	9	3	framework	framework	NOUN
cana-4362	9	4	uses	use	VERB
cana-4362	9	5	hybrid	hybrid	ADJ
cana-4362	9	6	feature	feature	NOUN
cana-4362	9	7	selection	selection	NOUN
cana-4362	9	8	to	to	PART
cana-4362	9	9	eliminate	eliminate	VERB
cana-4362	9	10	noisy	noisy	ADJ
cana-4362	9	11	and	and	CCONJ
cana-4362	9	12	irrelevant	irrelevant	ADJ
cana-4362	9	13	features	feature	NOUN
cana-4362	9	14	,	,	PUNCT
cana-4362	9	15	thereby	thereby	ADV
cana-4362	9	16	improving	improve	VERB
cana-4362	9	17	the	the	DET
cana-4362	9	18	model	model	NOUN
cana-4362	9	19	's	's	PART
cana-4362	9	20	generalization	generalization	NOUN
cana-4362	9	21	capabilities	capability	NOUN
cana-4362	9	22	.	.	PUNCT
cana-4362	10	1	experimental	experimental	ADJ
cana-4362	10	2	results	result	NOUN
cana-4362	10	3	reveal	reveal	VERB
cana-4362	10	4	that	that	SCONJ
cana-4362	10	5	the	the	DET
cana-4362	10	6	combination	combination	NOUN
cana-4362	10	7	of	of	ADP
cana-4362	10	8	hybrid	hybrid	ADJ
cana-4362	10	9	feature	feature	NOUN
cana-4362	10	10	selection	selection	NOUN
cana-4362	10	11	and	and	CCONJ
cana-4362	10	12	deep	deep	ADJ
cana-4362	10	13	learning	learning	NOUN
cana-4362	10	14	techniques	technique	NOUN
cana-4362	10	15	not	not	PART
cana-4362	10	16	only	only	ADV
cana-4362	10	17	boosts	boost	VERB
cana-4362	10	18	sentiment	sentiment	NOUN
cana-4362	10	19	classification	classification	NOUN
cana-4362	10	20	accuracy	accuracy	NOUN
cana-4362	10	21	but	but	CCONJ
cana-4362	10	22	also	also	ADV
cana-4362	10	23	decreases	decrease	VERB
cana-4362	10	24	computational	computational	ADJ
cana-4362	10	25	overhead	overhead	ADV
cana-4362	10	26	.	.	PUNCT
cana-4362	11	1	this	this	DET
cana-4362	11	2	study	study	NOUN
cana-4362	11	3	highlights	highlight	VERB
cana-4362	11	4	the	the	DET
cana-4362	11	5	effectiveness	effectiveness	NOUN
cana-4362	11	6	of	of	ADP
cana-4362	11	7	merging	merge	VERB
cana-4362	11	8	traditional	traditional	ADJ
cana-4362	11	9	feature	feature	NOUN
cana-4362	11	10	selection	selection	NOUN
cana-4362	11	11	methods	method	NOUN
cana-4362	11	12	with	with	ADP
cana-4362	11	13	modern	modern	ADJ
cana-4362	11	14	deep	deep	ADJ
cana-4362	11	15	learning	learning	NOUN
cana-4362	11	16	models	model	NOUN
cana-4362	11	17	to	to	PART
cana-4362	11	18	better	well	ADV
cana-4362	11	19	address	address	VERB
cana-4362	11	20	the	the	DET
cana-4362	11	21	complexities	complexity	NOUN
cana-4362	11	22	of	of	ADP
cana-4362	11	23	social	social	ADJ
cana-4362	11	24	media	medium	NOUN
cana-4362	11	25	datasets	dataset	NOUN
cana-4362	11	26	and	and	CCONJ
cana-4362	11	27	deliver	deliver	VERB
cana-4362	11	28	more	more	ADV
cana-4362	11	29	precise	precise	ADJ
cana-4362	11	30	sentiment	sentiment	NOUN
cana-4362	11	31	analysis	analysis	NOUN
cana-4362	11	32	.	.	PUNCT
cana-4362	12	1	the	the	DET
cana-4362	12	2	results	result	NOUN
cana-4362	12	3	achieved	achieve	VERB
cana-4362	12	4	by	by	ADP
cana-4362	12	5	proposed	propose	VERB
cana-4362	12	6	model	model	NOUN
cana-4362	12	7	is	be	AUX
cana-4362	12	8	98.50	98.50	NUM
cana-4362	12	9	%	%	NOUN
cana-4362	12	10	on	on	ADP
cana-4362	12	11	social	social	ADJ
cana-4362	12	12	media	medium	NOUN
cana-4362	12	13	dataset	dataset	NOUN
cana-4362	12	14	which	which	PRON
cana-4362	12	15	is	be	AUX
cana-4362	12	16	higher	high	ADJ
cana-4362	12	17	than	than	ADP
cana-4362	12	18	conventional	conventional	ADJ
cana-4362	12	19	approaches	approach	NOUN
cana-4362	12	20	.	.	PUNCT
cana-4362	13	1	keywords	keyword	NOUN
cana-4362	13	2	:	:	PUNCT
cana-4362	13	3	sentiment	sentiment	NOUN
cana-4362	13	4	classification	classification	NOUN
cana-4362	13	5	,	,	PUNCT
cana-4362	13	6	feature	feature	NOUN
cana-4362	13	7	extraction	extraction	NOUN
cana-4362	13	8	,	,	PUNCT
cana-4362	13	9	feature	feature	NOUN
cana-4362	13	10	selection	selection	NOUN
cana-4362	13	11	,	,	PUNCT
cana-4362	13	12	machine	machine	NOUN
cana-4362	13	13	learning	learning	NOUN
cana-4362	13	14	,	,	PUNCT
cana-4362	13	15	hybrid	hybrid	ADJ
cana-4362	13	16	features	feature	NOUN
cana-4362	13	17	,	,	PUNCT
cana-4362	13	18	social	social	ADJ
cana-4362	13	19	media	medium	NOUN
cana-4362	13	20	introduction	introduction	NOUN
cana-4362	13	21	the	the	DET
cana-4362	13	22	exponential	exponential	ADJ
cana-4362	13	23	growth	growth	NOUN
cana-4362	13	24	of	of	ADP
cana-4362	13	25	social	social	ADJ
cana-4362	13	26	media	medium	NOUN
cana-4362	13	27	platforms	platform	NOUN
cana-4362	13	28	such	such	ADJ
cana-4362	13	29	as	as	ADP
cana-4362	13	30	twitter	twitter	NOUN
cana-4362	13	31	,	,	PUNCT
cana-4362	13	32	facebook	facebook	NOUN
cana-4362	13	33	,	,	PUNCT
cana-4362	13	34	instagram	instagram	PROPN
cana-4362	13	35	,	,	PUNCT
cana-4362	13	36	and	and	CCONJ
cana-4362	13	37	reddit	reddit	NOUN
cana-4362	13	38	has	have	AUX
cana-4362	13	39	led	lead	VERB
cana-4362	13	40	to	to	ADP
cana-4362	13	41	the	the	DET
cana-4362	13	42	generation	generation	NOUN
cana-4362	13	43	of	of	ADP
cana-4362	13	44	massive	massive	ADJ
cana-4362	13	45	volumes	volume	NOUN
cana-4362	13	46	of	of	ADP
cana-4362	13	47	user	user	NOUN
cana-4362	13	48	-	-	PUNCT
cana-4362	13	49	generated	generate	VERB
cana-4362	13	50	content	content	NOUN
cana-4362	13	51	.	.	PUNCT
cana-4362	14	1	analyzing	analyze	VERB
cana-4362	14	2	and	and	CCONJ
cana-4362	14	3	understanding	understand	VERB
cana-4362	14	4	the	the	DET
cana-4362	14	5	sentiment	sentiment	NOUN
cana-4362	14	6	behind	behind	ADP
cana-4362	14	7	this	this	DET
cana-4362	14	8	content	content	NOUN
cana-4362	14	9	has	have	AUX
cana-4362	14	10	become	become	VERB
cana-4362	14	11	a	a	DET
cana-4362	14	12	crucial	crucial	ADJ
cana-4362	14	13	task	task	NOUN
cana-4362	14	14	in	in	ADP
cana-4362	14	15	many	many	ADJ
cana-4362	14	16	domains	domain	NOUN
cana-4362	14	17	,	,	PUNCT
cana-4362	14	18	including	include	VERB
cana-4362	14	19	marketing	marketing	NOUN
cana-4362	14	20	,	,	PUNCT
cana-4362	14	21	politics	politic	NOUN
cana-4362	14	22	,	,	PUNCT
cana-4362	14	23	and	and	CCONJ
cana-4362	14	24	public	public	ADJ
cana-4362	14	25	opinion	opinion	NOUN
cana-4362	14	26	analysis	analysis	NOUN
cana-4362	14	27	.	.	PUNCT
cana-4362	15	1	sentiment	sentiment	NOUN
cana-4362	15	2	classification	classification	NOUN
cana-4362	15	3	,	,	PUNCT
cana-4362	15	4	a	a	DET
cana-4362	15	5	branch	branch	NOUN
cana-4362	15	6	of	of	ADP
cana-4362	15	7	natural	natural	ADJ
cana-4362	15	8	language	language	NOUN
cana-4362	15	9	processing	processing	NOUN
cana-4362	15	10	(	(	PUNCT
cana-4362	15	11	nlp	nlp	NOUN
cana-4362	15	12	)	)	PUNCT
cana-4362	15	13	,	,	PUNCT
cana-4362	15	14	focuses	focus	VERB
cana-4362	15	15	on	on	ADP
cana-4362	15	16	identifying	identify	VERB
cana-4362	15	17	and	and	CCONJ
cana-4362	15	18	categorizing	categorize	VERB
cana-4362	15	19	opinions	opinion	NOUN
cana-4362	15	20	expressed	express	VERB
cana-4362	15	21	in	in	ADP
cana-4362	15	22	texts	text	NOUN
cana-4362	15	23	into	into	ADP
cana-4362	15	24	positive	positive	ADJ
cana-4362	15	25	,	,	PUNCT
cana-4362	15	26	negative	negative	ADJ
cana-4362	15	27	,	,	PUNCT
cana-4362	15	28	or	or	CCONJ
cana-4362	15	29	neutral	neutral	ADJ
cana-4362	15	30	sentiments	sentiment	NOUN
cana-4362	15	31	.	.	PUNCT
cana-4362	16	1	however	however	ADV
cana-4362	16	2	,	,	PUNCT
cana-4362	16	3	the	the	DET
cana-4362	16	4	vast	vast	ADJ
cana-4362	16	5	amount	amount	NOUN
cana-4362	16	6	of	of	ADP
cana-4362	16	7	unstructured	unstructured	ADJ
cana-4362	16	8	data	datum	NOUN
cana-4362	16	9	,	,	PUNCT
cana-4362	16	10	combined	combine	VERB
cana-4362	16	11	with	with	ADP
cana-4362	16	12	its	its	PRON
cana-4362	16	13	noisy	noisy	ADJ
cana-4362	16	14	and	and	CCONJ
cana-4362	16	15	highdimensional	highdimensional	ADJ
cana-4362	16	16	nature	nature	NOUN
cana-4362	16	17	,	,	PUNCT
cana-4362	16	18	presents	present	VERB
cana-4362	16	19	significant	significant	ADJ
cana-4362	16	20	challenges	challenge	NOUN
cana-4362	16	21	to	to	ADP
cana-4362	16	22	effective	effective	ADJ
cana-4362	16	23	sentiment	sentiment	NOUN
cana-4362	16	24	analysis	analysis	NOUN
cana-4362	16	25	.	.	PUNCT
cana-4362	17	1	to	to	PART
cana-4362	17	2	address	address	VERB
cana-4362	17	3	these	these	DET
cana-4362	17	4	challenges	challenge	NOUN
cana-4362	17	5	,	,	PUNCT
cana-4362	17	6	deep	deep	ADJ
cana-4362	17	7	learning	learning	NOUN
cana-4362	17	8	techniques	technique	NOUN
cana-4362	17	9	have	have	AUX
cana-4362	17	10	been	be	AUX
cana-4362	17	11	extensively	extensively	ADV
cana-4362	17	12	employed	employ	VERB
cana-4362	17	13	due	due	ADP
cana-4362	17	14	to	to	ADP
cana-4362	17	15	their	their	PRON
cana-4362	17	16	ability	ability	NOUN
cana-4362	17	17	to	to	PART
cana-4362	17	18	capture	capture	VERB
cana-4362	17	19	complex	complex	ADJ
cana-4362	17	20	patterns	pattern	NOUN
cana-4362	17	21	in	in	ADP
cana-4362	17	22	large	large	ADJ
cana-4362	17	23	datasets	dataset	NOUN
cana-4362	17	24	.	.	PUNCT
cana-4362	18	1	however	however	ADV
cana-4362	18	2	,	,	PUNCT
cana-4362	18	3	feature	feature	NOUN
cana-4362	18	4	selection	selection	NOUN
cana-4362	18	5	remains	remain	VERB
cana-4362	18	6	an	an	DET
cana-4362	18	7	essential	essential	ADJ
cana-4362	18	8	step	step	NOUN
cana-4362	18	9	to	to	PART
cana-4362	18	10	optimize	optimize	VERB
cana-4362	18	11	mailto:1hase.sudeep@gmail.com	mailto:1hase.sudeep@gmail.com	PROPN
cana-4362	18	12	mailto:2drrashmicseofficial@gmail.com	mailto:2drrashmicseofficial@gmail.com	PROPN
cana-4362	18	13	communications	communication	NOUN
cana-4362	18	14	on	on	ADP
cana-4362	18	15	applied	apply	VERB
cana-4362	18	16	nonlinear	nonlinear	ADJ
cana-4362	18	17	analysis	analysis	NOUN
cana-4362	18	18	issn	issn	NOUN
cana-4362	18	19	:	:	PUNCT
cana-4362	18	20	1074	1074	NUM
cana-4362	18	21	-	-	PUNCT
cana-4362	18	22	133x	133x	NUM
cana-4362	18	23	vol	vol	NOUN
cana-4362	18	24	32	32	NUM
cana-4362	18	25	no	no	NOUN
cana-4362	18	26	.	.	PUNCT
cana-4362	19	1	9s	9s	NUM
cana-4362	19	2	(	(	PUNCT
cana-4362	19	3	2025	2025	NUM
cana-4362	19	4	)	)	PUNCT
cana-4362	19	5	1900	1900	NUM
cana-4362	19	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-4362	19	7	these	these	DET
cana-4362	19	8	models	model	NOUN
cana-4362	19	9	,	,	PUNCT
cana-4362	19	10	reduce	reduce	VERB
cana-4362	19	11	computational	computational	ADJ
cana-4362	19	12	complexity	complexity	NOUN
cana-4362	19	13	,	,	PUNCT
cana-4362	19	14	and	and	CCONJ
cana-4362	19	15	improve	improve	VERB
cana-4362	19	16	accuracy	accuracy	NOUN
cana-4362	19	17	.	.	PUNCT
cana-4362	20	1	hybrid	hybrid	ADJ
cana-4362	20	2	feature	feature	NOUN
cana-4362	20	3	selection	selection	NOUN
cana-4362	20	4	methods	method	NOUN
cana-4362	20	5	,	,	PUNCT
cana-4362	20	6	which	which	PRON
cana-4362	20	7	combine	combine	VERB
cana-4362	20	8	multiple	multiple	ADJ
cana-4362	20	9	feature	feature	NOUN
cana-4362	20	10	selection	selection	NOUN
cana-4362	20	11	techniques	technique	NOUN
cana-4362	20	12	,	,	PUNCT
cana-4362	20	13	have	have	AUX
cana-4362	20	14	emerged	emerge	VERB
cana-4362	20	15	as	as	ADP
cana-4362	20	16	a	a	DET
cana-4362	20	17	robust	robust	ADJ
cana-4362	20	18	solution	solution	NOUN
cana-4362	20	19	for	for	ADP
cana-4362	20	20	enhancing	enhance	VERB
cana-4362	20	21	sentiment	sentiment	NOUN
cana-4362	20	22	classification	classification	NOUN
cana-4362	20	23	performance	performance	NOUN
cana-4362	20	24	in	in	ADP
cana-4362	20	25	social	social	ADJ
cana-4362	20	26	media	medium	NOUN
cana-4362	20	27	datasets	dataset	NOUN
cana-4362	20	28	.	.	PUNCT
cana-4362	21	1	importance	importance	NOUN
cana-4362	21	2	of	of	ADP
cana-4362	21	3	feature	feature	NOUN
cana-4362	21	4	selection	selection	NOUN
cana-4362	21	5	in	in	ADP
cana-4362	21	6	sentiment	sentiment	NOUN
cana-4362	21	7	classification	classification	NOUN
cana-4362	21	8	feature	feature	NOUN
cana-4362	21	9	selection	selection	NOUN
cana-4362	21	10	is	be	AUX
cana-4362	21	11	the	the	DET
cana-4362	21	12	process	process	NOUN
cana-4362	21	13	of	of	ADP
cana-4362	21	14	selecting	select	VERB
cana-4362	21	15	a	a	DET
cana-4362	21	16	subset	subset	NOUN
cana-4362	21	17	of	of	ADP
cana-4362	21	18	relevant	relevant	ADJ
cana-4362	21	19	features	feature	NOUN
cana-4362	21	20	(	(	PUNCT
cana-4362	21	21	variables	variable	NOUN
cana-4362	21	22	or	or	CCONJ
cana-4362	21	23	predictors	predictor	NOUN
cana-4362	21	24	)	)	PUNCT
cana-4362	21	25	for	for	ADP
cana-4362	21	26	model	model	NOUN
cana-4362	21	27	construction	construction	NOUN
cana-4362	21	28	.	.	PUNCT
cana-4362	22	1	in	in	ADP
cana-4362	22	2	the	the	DET
cana-4362	22	3	context	context	NOUN
cana-4362	22	4	of	of	ADP
cana-4362	22	5	sentiment	sentiment	NOUN
cana-4362	22	6	analysis	analysis	NOUN
cana-4362	22	7	on	on	ADP
cana-4362	22	8	social	social	ADJ
cana-4362	22	9	media	medium	NOUN
cana-4362	22	10	datasets	dataset	NOUN
cana-4362	22	11	,	,	PUNCT
cana-4362	22	12	these	these	DET
cana-4362	22	13	features	feature	NOUN
cana-4362	22	14	typically	typically	ADV
cana-4362	22	15	include	include	VERB
cana-4362	22	16	linguistic	linguistic	ADJ
cana-4362	22	17	and	and	CCONJ
cana-4362	22	18	semantic	semantic	ADJ
cana-4362	22	19	attributes	attribute	NOUN
cana-4362	22	20	such	such	ADJ
cana-4362	22	21	as	as	ADP
cana-4362	22	22	words	word	NOUN
cana-4362	22	23	,	,	PUNCT
cana-4362	22	24	phrases	phrase	NOUN
cana-4362	22	25	,	,	PUNCT
cana-4362	22	26	and	and	CCONJ
cana-4362	22	27	syntactic	syntactic	ADJ
cana-4362	22	28	patterns	pattern	NOUN
cana-4362	22	29	.	.	PUNCT
cana-4362	23	1	however	however	ADV
cana-4362	23	2	,	,	PUNCT
cana-4362	23	3	social	social	ADJ
cana-4362	23	4	media	medium	NOUN
cana-4362	23	5	data	datum	NOUN
cana-4362	23	6	is	be	AUX
cana-4362	23	7	often	often	ADV
cana-4362	23	8	characterized	characterize	VERB
cana-4362	23	9	by	by	ADP
cana-4362	23	10	high	high	ADJ
cana-4362	23	11	dimensionality	dimensionality	NOUN
cana-4362	23	12	and	and	CCONJ
cana-4362	23	13	sparsity	sparsity	NOUN
cana-4362	23	14	,	,	PUNCT
cana-4362	23	15	with	with	ADP
cana-4362	23	16	a	a	DET
cana-4362	23	17	large	large	ADJ
cana-4362	23	18	number	number	NOUN
cana-4362	23	19	of	of	ADP
cana-4362	23	20	irrelevant	irrelevant	ADJ
cana-4362	23	21	or	or	CCONJ
cana-4362	23	22	redundant	redundant	ADJ
cana-4362	23	23	features	feature	NOUN
cana-4362	23	24	that	that	PRON
cana-4362	23	25	can	can	AUX
cana-4362	23	26	degrade	degrade	VERB
cana-4362	23	27	the	the	DET
cana-4362	23	28	performance	performance	NOUN
cana-4362	23	29	of	of	ADP
cana-4362	23	30	machine	machine	NOUN
cana-4362	23	31	learning	learning	NOUN
cana-4362	23	32	models	model	NOUN
cana-4362	23	33	.	.	PUNCT
cana-4362	24	1	feature	feature	NOUN
cana-4362	24	2	selection	selection	NOUN
cana-4362	24	3	helps	help	VERB
cana-4362	24	4	in	in	ADP
cana-4362	24	5	reducing	reduce	VERB
cana-4362	24	6	the	the	DET
cana-4362	24	7	dimensionality	dimensionality	NOUN
cana-4362	24	8	of	of	ADP
cana-4362	24	9	the	the	DET
cana-4362	24	10	dataset	dataset	NOUN
cana-4362	24	11	,	,	PUNCT
cana-4362	24	12	enhancing	enhance	VERB
cana-4362	24	13	the	the	DET
cana-4362	24	14	model	model	NOUN
cana-4362	24	15	’s	’s	PART
cana-4362	24	16	generalization	generalization	NOUN
cana-4362	24	17	ability	ability	NOUN
cana-4362	24	18	,	,	PUNCT
cana-4362	24	19	and	and	CCONJ
cana-4362	24	20	speeding	speed	VERB
cana-4362	24	21	up	up	ADP
cana-4362	24	22	the	the	DET
cana-4362	24	23	training	training	NOUN
cana-4362	24	24	process	process	NOUN
cana-4362	24	25	by	by	ADP
cana-4362	24	26	eliminating	eliminate	VERB
cana-4362	24	27	irrelevant	irrelevant	ADJ
cana-4362	24	28	or	or	CCONJ
cana-4362	24	29	redundant	redundant	ADJ
cana-4362	24	30	data	datum	NOUN
cana-4362	24	31	.	.	PUNCT
cana-4362	25	1	traditionally	traditionally	ADV
cana-4362	25	2	,	,	PUNCT
cana-4362	25	3	feature	feature	NOUN
cana-4362	25	4	selection	selection	NOUN
cana-4362	25	5	techniques	technique	NOUN
cana-4362	25	6	are	be	AUX
cana-4362	25	7	categorized	categorize	VERB
cana-4362	25	8	into	into	ADP
cana-4362	25	9	three	three	NUM
cana-4362	25	10	main	main	ADJ
cana-4362	25	11	types	type	NOUN
cana-4362	25	12	:	:	PUNCT
cana-4362	25	13	filter	filter	NOUN
cana-4362	25	14	,	,	PUNCT
cana-4362	25	15	wrapper	wrapper	NOUN
cana-4362	25	16	,	,	PUNCT
cana-4362	25	17	and	and	CCONJ
cana-4362	25	18	embedded	embed	VERB
cana-4362	25	19	methods	method	NOUN
cana-4362	25	20	.	.	PUNCT
cana-4362	26	1	filter	filter	NOUN
cana-4362	26	2	methods	method	NOUN
cana-4362	26	3	,	,	PUNCT
cana-4362	26	4	such	such	ADJ
cana-4362	26	5	as	as	ADP
cana-4362	26	6	chi	chi	ADJ
cana-4362	26	7	-	-	PUNCT
cana-4362	26	8	square	square	ADJ
cana-4362	26	9	and	and	CCONJ
cana-4362	26	10	mutual	mutual	ADJ
cana-4362	26	11	information	information	NOUN
cana-4362	26	12	,	,	PUNCT
cana-4362	26	13	evaluate	evaluate	VERB
cana-4362	26	14	the	the	DET
cana-4362	26	15	relevance	relevance	NOUN
cana-4362	26	16	of	of	ADP
cana-4362	26	17	features	feature	NOUN
cana-4362	26	18	based	base	VERB
cana-4362	26	19	on	on	ADP
cana-4362	26	20	statistical	statistical	ADJ
cana-4362	26	21	measures	measure	NOUN
cana-4362	26	22	and	and	CCONJ
cana-4362	26	23	are	be	AUX
cana-4362	26	24	computationally	computationally	ADV
cana-4362	26	25	efficient	efficient	ADJ
cana-4362	26	26	.	.	PUNCT
cana-4362	27	1	wrapper	wrapper	NOUN
cana-4362	27	2	methods	method	NOUN
cana-4362	27	3	,	,	PUNCT
cana-4362	27	4	such	such	ADJ
cana-4362	27	5	as	as	ADP
cana-4362	27	6	recursive	recursive	ADJ
cana-4362	27	7	feature	feature	NOUN
cana-4362	27	8	elimination	elimination	NOUN
cana-4362	27	9	(	(	PUNCT
cana-4362	27	10	rfe	rfe	NOUN
cana-4362	27	11	)	)	PUNCT
cana-4362	27	12	,	,	PUNCT
cana-4362	27	13	evaluate	evaluate	VERB
cana-4362	27	14	subsets	subset	NOUN
cana-4362	27	15	of	of	ADP
cana-4362	27	16	features	feature	NOUN
cana-4362	27	17	by	by	ADP
cana-4362	27	18	training	train	VERB
cana-4362	27	19	a	a	DET
cana-4362	27	20	model	model	NOUN
cana-4362	27	21	and	and	CCONJ
cana-4362	27	22	assessing	assess	VERB
cana-4362	27	23	its	its	PRON
cana-4362	27	24	performance	performance	NOUN
cana-4362	27	25	,	,	PUNCT
cana-4362	27	26	though	though	SCONJ
cana-4362	27	27	these	these	DET
cana-4362	27	28	methods	method	NOUN
cana-4362	27	29	are	be	AUX
cana-4362	27	30	computationally	computationally	ADV
cana-4362	27	31	expensive	expensive	ADJ
cana-4362	27	32	.	.	PUNCT
cana-4362	28	1	embedded	embed	VERB
cana-4362	28	2	methods	method	NOUN
cana-4362	28	3	,	,	PUNCT
cana-4362	28	4	such	such	ADJ
cana-4362	28	5	as	as	ADP
cana-4362	28	6	decision	decision	NOUN
cana-4362	28	7	tree	tree	NOUN
cana-4362	28	8	-	-	PUNCT
cana-4362	28	9	based	base	VERB
cana-4362	28	10	algorithms	algorithm	NOUN
cana-4362	28	11	,	,	PUNCT
cana-4362	28	12	perform	perform	VERB
cana-4362	28	13	feature	feature	NOUN
cana-4362	28	14	selection	selection	NOUN
cana-4362	28	15	during	during	ADP
cana-4362	28	16	the	the	DET
cana-4362	28	17	model	model	NOUN
cana-4362	28	18	training	training	NOUN
cana-4362	28	19	process	process	NOUN
cana-4362	28	20	.	.	PUNCT
cana-4362	29	1	each	each	DET
cana-4362	29	2	method	method	NOUN
cana-4362	29	3	has	have	VERB
cana-4362	29	4	its	its	PRON
cana-4362	29	5	advantages	advantage	NOUN
cana-4362	29	6	and	and	CCONJ
cana-4362	29	7	limitations	limitation	NOUN
cana-4362	29	8	,	,	PUNCT
cana-4362	29	9	and	and	CCONJ
cana-4362	29	10	no	no	DET
cana-4362	29	11	single	single	ADJ
cana-4362	29	12	technique	technique	NOUN
cana-4362	29	13	is	be	AUX
cana-4362	29	14	universally	universally	ADV
cana-4362	29	15	effective	effective	ADJ
cana-4362	29	16	across	across	ADP
cana-4362	29	17	all	all	DET
cana-4362	29	18	datasets	dataset	NOUN
cana-4362	29	19	and	and	CCONJ
cana-4362	29	20	tasks	task	NOUN
cana-4362	29	21	.	.	PUNCT
cana-4362	30	1	the	the	DET
cana-4362	30	2	hybrid	hybrid	ADJ
cana-4362	30	3	approach	approach	NOUN
cana-4362	30	4	:	:	PUNCT
cana-4362	30	5	combining	combine	VERB
cana-4362	30	6	the	the	DET
cana-4362	30	7	best	good	ADJ
cana-4362	30	8	of	of	ADP
cana-4362	30	9	both	both	DET
cana-4362	30	10	worlds	world	NOUN
cana-4362	30	11	to	to	PART
cana-4362	30	12	overcome	overcome	VERB
cana-4362	30	13	the	the	DET
cana-4362	30	14	limitations	limitation	NOUN
cana-4362	30	15	of	of	ADP
cana-4362	30	16	individual	individual	ADJ
cana-4362	30	17	feature	feature	NOUN
cana-4362	30	18	selection	selection	NOUN
cana-4362	30	19	methods	method	NOUN
cana-4362	30	20	,	,	PUNCT
cana-4362	30	21	hybrid	hybrid	ADJ
cana-4362	30	22	approaches	approach	NOUN
cana-4362	30	23	have	have	AUX
cana-4362	30	24	been	be	AUX
cana-4362	30	25	developed	develop	VERB
cana-4362	30	26	.	.	PUNCT
cana-4362	31	1	hybrid	hybrid	ADJ
cana-4362	31	2	feature	feature	NOUN
cana-4362	31	3	selection	selection	NOUN
cana-4362	31	4	methods	method	NOUN
cana-4362	31	5	combine	combine	VERB
cana-4362	31	6	the	the	DET
cana-4362	31	7	strengths	strength	NOUN
cana-4362	31	8	of	of	ADP
cana-4362	31	9	multiple	multiple	ADJ
cana-4362	31	10	techniques	technique	NOUN
cana-4362	31	11	to	to	PART
cana-4362	31	12	improve	improve	VERB
cana-4362	31	13	the	the	DET
cana-4362	31	14	selection	selection	NOUN
cana-4362	31	15	process	process	NOUN
cana-4362	31	16	,	,	PUNCT
cana-4362	31	17	particularly	particularly	ADV
cana-4362	31	18	in	in	ADP
cana-4362	31	19	complex	complex	ADJ
cana-4362	31	20	and	and	CCONJ
cana-4362	31	21	high	high	ADJ
cana-4362	31	22	-	-	PUNCT
cana-4362	31	23	dimensional	dimensional	ADJ
cana-4362	31	24	environments	environment	NOUN
cana-4362	31	25	like	like	ADP
cana-4362	31	26	social	social	ADJ
cana-4362	31	27	media	medium	NOUN
cana-4362	31	28	sentiment	sentiment	NOUN
cana-4362	31	29	analysis	analysis	NOUN
cana-4362	31	30	.	.	PUNCT
cana-4362	32	1	these	these	DET
cana-4362	32	2	approaches	approach	NOUN
cana-4362	32	3	often	often	ADV
cana-4362	32	4	integrate	integrate	VERB
cana-4362	32	5	filter	filter	NOUN
cana-4362	32	6	and	and	CCONJ
cana-4362	32	7	wrapper	wrapper	NOUN
cana-4362	32	8	methods	method	NOUN
cana-4362	32	9	,	,	PUNCT
cana-4362	32	10	or	or	CCONJ
cana-4362	32	11	combine	combine	VERB
cana-4362	32	12	traditional	traditional	ADJ
cana-4362	32	13	feature	feature	NOUN
cana-4362	32	14	selection	selection	NOUN
cana-4362	32	15	techniques	technique	NOUN
cana-4362	32	16	with	with	ADP
cana-4362	32	17	deep	deep	ADJ
cana-4362	32	18	learning	learning	NOUN
cana-4362	32	19	architectures	architecture	NOUN
cana-4362	32	20	to	to	PART
cana-4362	32	21	optimize	optimize	VERB
cana-4362	32	22	the	the	DET
cana-4362	32	23	selection	selection	NOUN
cana-4362	32	24	of	of	ADP
cana-4362	32	25	relevant	relevant	ADJ
cana-4362	32	26	features	feature	NOUN
cana-4362	32	27	.	.	PUNCT
cana-4362	33	1	for	for	ADP
cana-4362	33	2	example	example	NOUN
cana-4362	33	3	,	,	PUNCT
cana-4362	33	4	a	a	DET
cana-4362	33	5	hybrid	hybrid	ADJ
cana-4362	33	6	model	model	NOUN
cana-4362	33	7	may	may	AUX
cana-4362	33	8	use	use	VERB
cana-4362	33	9	a	a	DET
cana-4362	33	10	filter	filter	NOUN
cana-4362	33	11	method	method	NOUN
cana-4362	33	12	such	such	ADJ
cana-4362	33	13	as	as	ADP
cana-4362	33	14	chi	chi	NOUN
cana-4362	33	15	-	-	PUNCT
cana-4362	33	16	square	square	NOUN
cana-4362	33	17	to	to	PART
cana-4362	33	18	perform	perform	VERB
cana-4362	33	19	an	an	DET
cana-4362	33	20	initial	initial	ADJ
cana-4362	33	21	reduction	reduction	NOUN
cana-4362	33	22	of	of	ADP
cana-4362	33	23	irrelevant	irrelevant	ADJ
cana-4362	33	24	features	feature	NOUN
cana-4362	33	25	,	,	PUNCT
cana-4362	33	26	followed	follow	VERB
cana-4362	33	27	by	by	ADP
cana-4362	33	28	a	a	DET
cana-4362	33	29	wrapper	wrapper	NOUN
cana-4362	33	30	method	method	NOUN
cana-4362	33	31	like	like	ADP
cana-4362	33	32	genetic	genetic	ADJ
cana-4362	33	33	algorithms	algorithm	NOUN
cana-4362	33	34	(	(	PUNCT
cana-4362	33	35	ga	ga	NOUN
cana-4362	33	36	)	)	PUNCT
cana-4362	33	37	or	or	CCONJ
cana-4362	33	38	particle	particle	NOUN
cana-4362	33	39	swarm	swarm	NOUN
cana-4362	33	40	optimization	optimization	NOUN
cana-4362	33	41	(	(	PUNCT
cana-4362	33	42	pso	pso	NOUN
cana-4362	33	43	)	)	PUNCT
cana-4362	33	44	to	to	ADP
cana-4362	33	45	fine	fine	ADJ
cana-4362	33	46	-	-	PUNCT
cana-4362	33	47	tune	tune	NOUN
cana-4362	33	48	the	the	DET
cana-4362	33	49	feature	feature	NOUN
cana-4362	33	50	subset	subset	VERB
cana-4362	33	51	.	.	PUNCT
cana-4362	34	1	in	in	ADP
cana-4362	34	2	some	some	DET
cana-4362	34	3	models	model	NOUN
cana-4362	34	4	,	,	PUNCT
cana-4362	34	5	deep	deep	ADJ
cana-4362	34	6	learning	learning	NOUN
cana-4362	34	7	techniques	technique	NOUN
cana-4362	34	8	like	like	ADP
cana-4362	34	9	convolutional	convolutional	ADJ
cana-4362	34	10	neural	neural	ADJ
cana-4362	34	11	networks	network	NOUN
cana-4362	34	12	(	(	PUNCT
cana-4362	34	13	cnns	cnns	PROPN
cana-4362	34	14	)	)	PUNCT
cana-4362	34	15	or	or	CCONJ
cana-4362	34	16	long	long	ADJ
cana-4362	34	17	short	short	ADJ
cana-4362	34	18	-	-	PUNCT
cana-4362	34	19	term	term	NOUN
cana-4362	34	20	memory	memory	NOUN
cana-4362	34	21	(	(	PUNCT
cana-4362	34	22	lstm	lstm	NOUN
cana-4362	34	23	)	)	PUNCT
cana-4362	34	24	networks	network	NOUN
cana-4362	34	25	are	be	AUX
cana-4362	34	26	incorporated	incorporate	VERB
cana-4362	34	27	to	to	PART
cana-4362	34	28	further	far	ADV
cana-4362	34	29	optimize	optimize	VERB
cana-4362	34	30	feature	feature	NOUN
cana-4362	34	31	selection	selection	NOUN
cana-4362	34	32	by	by	ADP
cana-4362	34	33	identifying	identify	VERB
cana-4362	34	34	intricate	intricate	ADJ
cana-4362	34	35	patterns	pattern	NOUN
cana-4362	34	36	and	and	CCONJ
cana-4362	34	37	relationships	relationship	NOUN
cana-4362	34	38	in	in	ADP
cana-4362	34	39	the	the	DET
cana-4362	34	40	data	datum	NOUN
cana-4362	34	41	.	.	PUNCT
cana-4362	35	1	the	the	DET
cana-4362	35	2	hybrid	hybrid	ADJ
cana-4362	35	3	approach	approach	NOUN
cana-4362	35	4	offers	offer	VERB
cana-4362	35	5	several	several	ADJ
cana-4362	35	6	advantages	advantage	NOUN
cana-4362	35	7	.	.	PUNCT
cana-4362	36	1	first	first	ADV
cana-4362	36	2	,	,	PUNCT
cana-4362	36	3	it	it	PRON
cana-4362	36	4	strikes	strike	VERB
cana-4362	36	5	a	a	DET
cana-4362	36	6	balance	balance	NOUN
cana-4362	36	7	between	between	ADP
cana-4362	36	8	computational	computational	ADJ
cana-4362	36	9	efficiency	efficiency	NOUN
cana-4362	36	10	and	and	CCONJ
cana-4362	36	11	accuracy	accuracy	NOUN
cana-4362	36	12	.	.	PUNCT
cana-4362	37	1	filter	filter	NOUN
cana-4362	37	2	methods	method	NOUN
cana-4362	37	3	can	can	AUX
cana-4362	37	4	quickly	quickly	ADV
cana-4362	37	5	eliminate	eliminate	VERB
cana-4362	37	6	irrelevant	irrelevant	ADJ
cana-4362	37	7	features	feature	NOUN
cana-4362	37	8	,	,	PUNCT
cana-4362	37	9	reducing	reduce	VERB
cana-4362	37	10	the	the	DET
cana-4362	37	11	size	size	NOUN
cana-4362	37	12	of	of	ADP
cana-4362	37	13	the	the	DET
cana-4362	37	14	dataset	dataset	NOUN
cana-4362	37	15	,	,	PUNCT
cana-4362	37	16	while	while	SCONJ
cana-4362	37	17	wrapper	wrapper	NOUN
cana-4362	37	18	methods	method	NOUN
cana-4362	37	19	and	and	CCONJ
cana-4362	37	20	deep	deep	ADJ
cana-4362	37	21	learning	learning	NOUN
cana-4362	37	22	models	model	NOUN
cana-4362	37	23	can	can	AUX
cana-4362	37	24	focus	focus	VERB
cana-4362	37	25	on	on	ADP
cana-4362	37	26	fine	fine	ADV
cana-4362	37	27	-	-	PUNCT
cana-4362	37	28	tuning	tuning	NOUN
cana-4362	37	29	and	and	CCONJ
cana-4362	37	30	identifying	identify	VERB
cana-4362	37	31	the	the	DET
cana-4362	37	32	most	most	ADV
cana-4362	37	33	informative	informative	ADJ
cana-4362	37	34	features	feature	NOUN
cana-4362	37	35	.	.	PUNCT
cana-4362	38	1	second	second	ADJ
cana-4362	38	2	,	,	PUNCT
cana-4362	38	3	hybrid	hybrid	ADJ
cana-4362	38	4	methods	method	NOUN
cana-4362	38	5	reduce	reduce	VERB
cana-4362	38	6	the	the	DET
cana-4362	38	7	risk	risk	NOUN
cana-4362	38	8	of	of	ADP
cana-4362	38	9	overfitting	overfitte	VERB
cana-4362	38	10	by	by	ADP
cana-4362	38	11	preventing	prevent	VERB
cana-4362	38	12	the	the	DET
cana-4362	38	13	model	model	NOUN
cana-4362	38	14	from	from	ADP
cana-4362	38	15	becoming	become	VERB
cana-4362	38	16	too	too	ADV
cana-4362	38	17	complex	complex	ADJ
cana-4362	38	18	or	or	CCONJ
cana-4362	38	19	overly	overly	ADV
cana-4362	38	20	reliant	reliant	ADJ
cana-4362	38	21	on	on	ADP
cana-4362	38	22	a	a	DET
cana-4362	38	23	large	large	ADJ
cana-4362	38	24	number	number	NOUN
cana-4362	38	25	of	of	ADP
cana-4362	38	26	features	feature	NOUN
cana-4362	38	27	.	.	PUNCT
cana-4362	39	1	finally	finally	ADV
cana-4362	39	2	,	,	PUNCT
cana-4362	39	3	these	these	DET
cana-4362	39	4	methods	method	NOUN
cana-4362	39	5	are	be	AUX
cana-4362	39	6	better	well	ADV
cana-4362	39	7	suited	suited	ADJ
cana-4362	39	8	for	for	ADP
cana-4362	39	9	real	real	ADJ
cana-4362	39	10	-	-	PUNCT
cana-4362	39	11	time	time	NOUN
cana-4362	39	12	applications	application	NOUN
cana-4362	39	13	like	like	ADP
cana-4362	39	14	sentiment	sentiment	NOUN
cana-4362	39	15	analysis	analysis	NOUN
cana-4362	39	16	in	in	ADP
cana-4362	39	17	social	social	ADJ
cana-4362	39	18	media	medium	NOUN
cana-4362	39	19	,	,	PUNCT
cana-4362	39	20	where	where	SCONJ
cana-4362	39	21	the	the	DET
cana-4362	39	22	ability	ability	NOUN
cana-4362	39	23	to	to	PART
cana-4362	39	24	process	process	VERB
cana-4362	39	25	and	and	CCONJ
cana-4362	39	26	analyze	analyze	VERB
cana-4362	39	27	large	large	ADJ
cana-4362	39	28	volumes	volume	NOUN
cana-4362	39	29	of	of	ADP
cana-4362	39	30	data	datum	NOUN
cana-4362	39	31	quickly	quickly	ADV
cana-4362	39	32	and	and	CCONJ
cana-4362	39	33	accurately	accurately	ADV
cana-4362	39	34	is	be	AUX
cana-4362	39	35	crucial	crucial	ADJ
cana-4362	39	36	.	.	PUNCT
cana-4362	40	1	communications	communication	NOUN
cana-4362	40	2	on	on	ADP
cana-4362	40	3	applied	apply	VERB
cana-4362	40	4	nonlinear	nonlinear	ADJ
cana-4362	40	5	analysis	analysis	NOUN
cana-4362	40	6	issn	issn	NOUN
cana-4362	40	7	:	:	PUNCT
cana-4362	40	8	1074	1074	NUM
cana-4362	40	9	-	-	PUNCT
cana-4362	40	10	133x	133x	NUM
cana-4362	40	11	vol	vol	NOUN
cana-4362	40	12	32	32	NUM
cana-4362	40	13	no	no	NOUN
cana-4362	40	14	.	.	PUNCT
cana-4362	41	1	9s	9s	NUM
cana-4362	41	2	(	(	PUNCT
cana-4362	41	3	2025	2025	NUM
cana-4362	41	4	)	)	PUNCT
cana-4362	41	5	1901	1901	NUM
cana-4362	41	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-4362	42	1	the	the	DET
cana-4362	42	2	role	role	NOUN
cana-4362	42	3	of	of	ADP
cana-4362	42	4	deep	deep	ADJ
cana-4362	42	5	learning	learning	NOUN
cana-4362	42	6	in	in	ADP
cana-4362	42	7	hybrid	hybrid	ADJ
cana-4362	42	8	feature	feature	NOUN
cana-4362	42	9	selection	selection	NOUN
cana-4362	42	10	deep	deep	ADJ
cana-4362	42	11	learning	learning	NOUN
cana-4362	42	12	models	model	NOUN
cana-4362	42	13	,	,	PUNCT
cana-4362	42	14	especially	especially	ADV
cana-4362	42	15	cnns	cnns	ADJ
cana-4362	42	16	and	and	CCONJ
cana-4362	42	17	lstms	lstms	ADJ
cana-4362	42	18	,	,	PUNCT
cana-4362	42	19	have	have	AUX
cana-4362	42	20	demonstrated	demonstrate	VERB
cana-4362	42	21	remarkable	remarkable	ADJ
cana-4362	42	22	success	success	NOUN
cana-4362	42	23	in	in	ADP
cana-4362	42	24	sentiment	sentiment	NOUN
cana-4362	42	25	classification	classification	NOUN
cana-4362	42	26	tasks	task	NOUN
cana-4362	42	27	due	due	ADP
cana-4362	42	28	to	to	ADP
cana-4362	42	29	their	their	PRON
cana-4362	42	30	ability	ability	NOUN
cana-4362	42	31	to	to	PART
cana-4362	42	32	automatically	automatically	ADV
cana-4362	42	33	learn	learn	VERB
cana-4362	42	34	feature	feature	NOUN
cana-4362	42	35	representations	representation	NOUN
cana-4362	42	36	from	from	ADP
cana-4362	42	37	raw	raw	ADJ
cana-4362	42	38	data	datum	NOUN
cana-4362	42	39	.	.	PUNCT
cana-4362	43	1	these	these	DET
cana-4362	43	2	models	model	NOUN
cana-4362	43	3	are	be	AUX
cana-4362	43	4	adept	adept	ADJ
cana-4362	43	5	at	at	ADP
cana-4362	43	6	capturing	capture	VERB
cana-4362	43	7	the	the	DET
cana-4362	43	8	temporal	temporal	ADJ
cana-4362	43	9	dependencies	dependency	NOUN
cana-4362	43	10	and	and	CCONJ
cana-4362	43	11	spatial	spatial	ADJ
cana-4362	43	12	relationships	relationship	NOUN
cana-4362	43	13	inherent	inherent	ADJ
cana-4362	43	14	in	in	ADP
cana-4362	43	15	social	social	ADJ
cana-4362	43	16	media	medium	NOUN
cana-4362	43	17	texts	text	NOUN
cana-4362	43	18	.	.	PUNCT
cana-4362	44	1	when	when	SCONJ
cana-4362	44	2	combined	combine	VERB
cana-4362	44	3	with	with	ADP
cana-4362	44	4	hybrid	hybrid	ADJ
cana-4362	44	5	feature	feature	NOUN
cana-4362	44	6	selection	selection	NOUN
cana-4362	44	7	methods	method	NOUN
cana-4362	44	8	,	,	PUNCT
cana-4362	44	9	deep	deep	ADJ
cana-4362	44	10	learning	learning	NOUN
cana-4362	44	11	models	model	NOUN
cana-4362	44	12	can	can	AUX
cana-4362	44	13	further	far	ADV
cana-4362	44	14	enhance	enhance	VERB
cana-4362	44	15	the	the	DET
cana-4362	44	16	classification	classification	NOUN
cana-4362	44	17	process	process	NOUN
cana-4362	44	18	by	by	ADP
cana-4362	44	19	identifying	identify	VERB
cana-4362	44	20	non	non	ADJ
cana-4362	44	21	-	-	ADJ
cana-4362	44	22	linear	linear	ADJ
cana-4362	44	23	patterns	pattern	NOUN
cana-4362	44	24	and	and	CCONJ
cana-4362	44	25	interactions	interaction	NOUN
cana-4362	44	26	between	between	ADP
cana-4362	44	27	features	feature	NOUN
cana-4362	44	28	that	that	SCONJ
cana-4362	44	29	traditional	traditional	ADJ
cana-4362	44	30	methods	method	NOUN
cana-4362	44	31	might	might	AUX
cana-4362	44	32	overlook	overlook	VERB
cana-4362	44	33	.	.	PUNCT
cana-4362	45	1	a	a	DET
cana-4362	45	2	cnn	cnn	PROPN
cana-4362	45	3	can	can	AUX
cana-4362	45	4	be	be	AUX
cana-4362	45	5	used	use	VERB
cana-4362	45	6	to	to	PART
cana-4362	45	7	capture	capture	VERB
cana-4362	45	8	local	local	ADJ
cana-4362	45	9	dependencies	dependency	NOUN
cana-4362	45	10	in	in	ADP
cana-4362	45	11	text	text	NOUN
cana-4362	45	12	data	datum	NOUN
cana-4362	45	13	,	,	PUNCT
cana-4362	45	14	while	while	SCONJ
cana-4362	45	15	an	an	DET
cana-4362	45	16	lstm	lstm	ADJ
cana-4362	45	17	network	network	NOUN
cana-4362	45	18	can	can	AUX
cana-4362	45	19	be	be	AUX
cana-4362	45	20	employed	employ	VERB
cana-4362	45	21	to	to	PART
cana-4362	45	22	capture	capture	VERB
cana-4362	45	23	long	long	ADJ
cana-4362	45	24	-	-	PUNCT
cana-4362	45	25	term	term	NOUN
cana-4362	45	26	dependencies	dependency	NOUN
cana-4362	45	27	.	.	PUNCT
cana-4362	46	1	when	when	SCONJ
cana-4362	46	2	these	these	DET
cana-4362	46	3	models	model	NOUN
cana-4362	46	4	are	be	AUX
cana-4362	46	5	coupled	couple	VERB
cana-4362	46	6	with	with	ADP
cana-4362	46	7	hybrid	hybrid	ADJ
cana-4362	46	8	feature	feature	NOUN
cana-4362	46	9	selection	selection	NOUN
cana-4362	46	10	methods	method	NOUN
cana-4362	46	11	,	,	PUNCT
cana-4362	46	12	the	the	DET
cana-4362	46	13	overall	overall	ADJ
cana-4362	46	14	performance	performance	NOUN
cana-4362	46	15	of	of	ADP
cana-4362	46	16	sentiment	sentiment	NOUN
cana-4362	46	17	classification	classification	NOUN
cana-4362	46	18	improves	improve	VERB
cana-4362	46	19	,	,	PUNCT
cana-4362	46	20	particularly	particularly	ADV
cana-4362	46	21	in	in	ADP
cana-4362	46	22	terms	term	NOUN
cana-4362	46	23	of	of	ADP
cana-4362	46	24	accuracy	accuracy	NOUN
cana-4362	46	25	,	,	PUNCT
cana-4362	46	26	precision	precision	NOUN
cana-4362	46	27	,	,	PUNCT
cana-4362	46	28	recall	recall	NOUN
cana-4362	46	29	,	,	PUNCT
cana-4362	46	30	and	and	CCONJ
cana-4362	46	31	f1	f1	PROPN
cana-4362	46	32	score	score	NOUN
cana-4362	46	33	.	.	PUNCT
cana-4362	47	1	hybrid	hybrid	ADJ
cana-4362	47	2	feature	feature	NOUN
cana-4362	47	3	selection	selection	NOUN
cana-4362	47	4	approaches	approach	NOUN
cana-4362	47	5	represent	represent	VERB
cana-4362	47	6	a	a	DET
cana-4362	47	7	powerful	powerful	ADJ
cana-4362	47	8	and	and	CCONJ
cana-4362	47	9	efficient	efficient	ADJ
cana-4362	47	10	solution	solution	NOUN
cana-4362	47	11	for	for	ADP
cana-4362	47	12	sentiment	sentiment	NOUN
cana-4362	47	13	classification	classification	NOUN
cana-4362	47	14	on	on	ADP
cana-4362	47	15	social	social	ADJ
cana-4362	47	16	media	medium	NOUN
cana-4362	47	17	datasets	dataset	NOUN
cana-4362	47	18	using	use	VERB
cana-4362	47	19	deep	deep	ADJ
cana-4362	47	20	learning	learning	NOUN
cana-4362	47	21	techniques	technique	NOUN
cana-4362	47	22	.	.	PUNCT
cana-4362	48	1	by	by	ADP
cana-4362	48	2	combining	combine	VERB
cana-4362	48	3	the	the	DET
cana-4362	48	4	strengths	strength	NOUN
cana-4362	48	5	of	of	ADP
cana-4362	48	6	multiple	multiple	ADJ
cana-4362	48	7	feature	feature	NOUN
cana-4362	48	8	selection	selection	NOUN
cana-4362	48	9	methods	method	NOUN
cana-4362	48	10	and	and	CCONJ
cana-4362	48	11	leveraging	leverage	VERB
cana-4362	48	12	the	the	DET
cana-4362	48	13	capabilities	capability	NOUN
cana-4362	48	14	of	of	ADP
cana-4362	48	15	deep	deep	ADJ
cana-4362	48	16	learning	learning	NOUN
cana-4362	48	17	models	model	NOUN
cana-4362	48	18	,	,	PUNCT
cana-4362	48	19	hybrid	hybrid	ADJ
cana-4362	48	20	approaches	approach	NOUN
cana-4362	48	21	can	can	AUX
cana-4362	48	22	effectively	effectively	ADV
cana-4362	48	23	handle	handle	VERB
cana-4362	48	24	the	the	DET
cana-4362	48	25	high	high	ADJ
cana-4362	48	26	dimensionality	dimensionality	NOUN
cana-4362	48	27	,	,	PUNCT
cana-4362	48	28	sparsity	sparsity	NOUN
cana-4362	48	29	,	,	PUNCT
cana-4362	48	30	and	and	CCONJ
cana-4362	48	31	noisy	noisy	ADJ
cana-4362	48	32	nature	nature	NOUN
cana-4362	48	33	of	of	ADP
cana-4362	48	34	social	social	ADJ
cana-4362	48	35	media	medium	NOUN
cana-4362	48	36	data	datum	NOUN
cana-4362	48	37	.	.	PUNCT
cana-4362	49	1	these	these	DET
cana-4362	49	2	methods	method	NOUN
cana-4362	49	3	not	not	PART
cana-4362	49	4	only	only	ADV
cana-4362	49	5	enhance	enhance	VERB
cana-4362	49	6	classification	classification	NOUN
cana-4362	49	7	accuracy	accuracy	NOUN
cana-4362	49	8	but	but	CCONJ
cana-4362	49	9	also	also	ADV
cana-4362	49	10	improve	improve	VERB
cana-4362	49	11	computational	computational	ADJ
cana-4362	49	12	efficiency	efficiency	NOUN
cana-4362	49	13	,	,	PUNCT
cana-4362	49	14	making	make	VERB
cana-4362	49	15	them	they	PRON
cana-4362	49	16	highly	highly	ADV
cana-4362	49	17	suitable	suitable	ADJ
cana-4362	49	18	for	for	ADP
cana-4362	49	19	real	real	ADJ
cana-4362	49	20	-	-	PUNCT
cana-4362	49	21	world	world	NOUN
cana-4362	49	22	applications	application	NOUN
cana-4362	49	23	in	in	ADP
cana-4362	49	24	sentiment	sentiment	NOUN
cana-4362	49	25	analysis	analysis	NOUN
cana-4362	49	26	.	.	PUNCT
cana-4362	50	1	literature	literature	NOUN
cana-4362	50	2	survey	survey	NOUN
cana-4362	50	3	zhang	zhang	PROPN
cana-4362	50	4	et	et	PROPN
cana-4362	50	5	al	al	PROPN
cana-4362	50	6	.	.	PUNCT
cana-4362	51	1	[	[	X
cana-4362	51	2	1	1	X
cana-4362	51	3	]	]	PUNCT
cana-4362	51	4	explores	explore	VERB
cana-4362	51	5	a	a	DET
cana-4362	51	6	hybrid	hybrid	ADJ
cana-4362	51	7	feature	feature	NOUN
cana-4362	51	8	selection	selection	NOUN
cana-4362	51	9	approach	approach	NOUN
cana-4362	51	10	combining	combine	VERB
cana-4362	51	11	statistical	statistical	ADJ
cana-4362	51	12	and	and	CCONJ
cana-4362	51	13	deep	deep	ADJ
cana-4362	51	14	learning	learning	NOUN
cana-4362	51	15	methods	method	NOUN
cana-4362	51	16	to	to	PART
cana-4362	51	17	enhance	enhance	VERB
cana-4362	51	18	sentiment	sentiment	NOUN
cana-4362	51	19	classification	classification	NOUN
cana-4362	51	20	of	of	ADP
cana-4362	51	21	social	social	ADJ
cana-4362	51	22	media	medium	NOUN
cana-4362	51	23	data	datum	NOUN
cana-4362	51	24	.	.	PUNCT
cana-4362	52	1	they	they	PRON
cana-4362	52	2	focus	focus	VERB
cana-4362	52	3	on	on	ADP
cana-4362	52	4	improving	improve	VERB
cana-4362	52	5	the	the	DET
cana-4362	52	6	classification	classification	NOUN
cana-4362	52	7	accuracy	accuracy	NOUN
cana-4362	52	8	by	by	ADP
cana-4362	52	9	leveraging	leverage	VERB
cana-4362	52	10	a	a	DET
cana-4362	52	11	feature	feature	NOUN
cana-4362	52	12	ranking	rank	VERB
cana-4362	52	13	mechanism	mechanism	NOUN
cana-4362	52	14	that	that	PRON
cana-4362	52	15	integrates	integrate	VERB
cana-4362	52	16	both	both	CCONJ
cana-4362	52	17	filter	filter	NOUN
cana-4362	52	18	and	and	CCONJ
cana-4362	52	19	wrapper	wrapper	NOUN
cana-4362	52	20	techniques	technique	NOUN
cana-4362	52	21	.	.	PUNCT
cana-4362	53	1	the	the	DET
cana-4362	53	2	deep	deep	ADJ
cana-4362	53	3	learning	learning	NOUN
cana-4362	53	4	models	model	NOUN
cana-4362	53	5	such	such	ADJ
cana-4362	53	6	as	as	ADP
cana-4362	53	7	cnn	cnn	PROPN
cana-4362	53	8	and	and	CCONJ
cana-4362	53	9	lstm	lstm	NOUN
cana-4362	53	10	are	be	AUX
cana-4362	53	11	used	use	VERB
cana-4362	53	12	to	to	ADP
cana-4362	53	13	fine	fine	ADJ
cana-4362	53	14	-	-	PUNCT
cana-4362	53	15	tune	tune	NOUN
cana-4362	53	16	the	the	DET
cana-4362	53	17	selected	select	VERB
cana-4362	53	18	features	feature	NOUN
cana-4362	53	19	for	for	ADP
cana-4362	53	20	optimal	optimal	ADJ
cana-4362	53	21	performance	performance	NOUN
cana-4362	53	22	.	.	PUNCT
cana-4362	54	1	their	their	PRON
cana-4362	54	2	experiments	experiment	NOUN
cana-4362	54	3	on	on	ADP
cana-4362	54	4	multiple	multiple	ADJ
cana-4362	54	5	datasets	dataset	NOUN
cana-4362	54	6	,	,	PUNCT
cana-4362	54	7	including	include	VERB
cana-4362	54	8	twitter	twitter	NOUN
cana-4362	54	9	and	and	CCONJ
cana-4362	54	10	reddit	reddit	NOUN
cana-4362	54	11	,	,	PUNCT
cana-4362	54	12	show	show	VERB
cana-4362	54	13	that	that	SCONJ
cana-4362	54	14	this	this	DET
cana-4362	54	15	hybrid	hybrid	ADJ
cana-4362	54	16	approach	approach	NOUN
cana-4362	54	17	outperforms	outperform	VERB
cana-4362	54	18	traditional	traditional	ADJ
cana-4362	54	19	feature	feature	NOUN
cana-4362	54	20	selection	selection	NOUN
cana-4362	54	21	methods	method	NOUN
cana-4362	54	22	in	in	ADP
cana-4362	54	23	terms	term	NOUN
cana-4362	54	24	of	of	ADP
cana-4362	54	25	precision	precision	NOUN
cana-4362	54	26	,	,	PUNCT
cana-4362	54	27	recall	recall	NOUN
cana-4362	54	28	,	,	PUNCT
cana-4362	54	29	and	and	CCONJ
cana-4362	54	30	f1	f1	PROPN
cana-4362	54	31	score	score	NOUN
cana-4362	54	32	.	.	PUNCT
cana-4362	55	1	the	the	DET
cana-4362	55	2	study	study	NOUN
cana-4362	55	3	highlights	highlight	VERB
cana-4362	55	4	the	the	DET
cana-4362	55	5	importance	importance	NOUN
cana-4362	55	6	of	of	ADP
cana-4362	55	7	balancing	balance	VERB
cana-4362	55	8	feature	feature	NOUN
cana-4362	55	9	dimensionality	dimensionality	NOUN
cana-4362	55	10	reduction	reduction	NOUN
cana-4362	55	11	and	and	CCONJ
cana-4362	55	12	model	model	NOUN
cana-4362	55	13	complexity	complexity	NOUN
cana-4362	55	14	to	to	PART
cana-4362	55	15	ensure	ensure	VERB
cana-4362	55	16	accurate	accurate	ADJ
cana-4362	55	17	sentiment	sentiment	NOUN
cana-4362	55	18	detection	detection	NOUN
cana-4362	55	19	in	in	ADP
cana-4362	55	20	real	real	ADJ
cana-4362	55	21	-	-	PUNCT
cana-4362	55	22	time	time	NOUN
cana-4362	55	23	social	social	ADJ
cana-4362	55	24	media	medium	NOUN
cana-4362	55	25	applications	application	NOUN
cana-4362	55	26	.	.	PUNCT
cana-4362	56	1	chen	chen	PROPN
cana-4362	56	2	et	et	PROPN
cana-4362	56	3	al	al	PROPN
cana-4362	56	4	.	.	PUNCT
cana-4362	57	1	[	[	X
cana-4362	57	2	2	2	NUM
cana-4362	57	3	]	]	PUNCT
cana-4362	57	4	presents	present	VERB
cana-4362	57	5	a	a	DET
cana-4362	57	6	hybrid	hybrid	ADJ
cana-4362	57	7	model	model	NOUN
cana-4362	57	8	combining	combine	VERB
cana-4362	57	9	deep	deep	ADJ
cana-4362	57	10	learning	learning	NOUN
cana-4362	57	11	algorithms	algorithm	NOUN
cana-4362	57	12	with	with	ADP
cana-4362	57	13	traditional	traditional	ADJ
cana-4362	57	14	feature	feature	NOUN
cana-4362	57	15	selection	selection	NOUN
cana-4362	57	16	methods	method	NOUN
cana-4362	57	17	for	for	ADP
cana-4362	57	18	sentiment	sentiment	NOUN
cana-4362	57	19	classification	classification	NOUN
cana-4362	57	20	on	on	ADP
cana-4362	57	21	large	large	ADJ
cana-4362	57	22	-	-	PUNCT
cana-4362	57	23	scale	scale	NOUN
cana-4362	57	24	social	social	ADJ
cana-4362	57	25	media	medium	NOUN
cana-4362	57	26	data	datum	NOUN
cana-4362	57	27	.	.	PUNCT
cana-4362	58	1	the	the	DET
cana-4362	58	2	authors	author	NOUN
cana-4362	58	3	propose	propose	VERB
cana-4362	58	4	using	use	VERB
cana-4362	58	5	an	an	DET
cana-4362	58	6	integration	integration	NOUN
cana-4362	58	7	of	of	ADP
cana-4362	58	8	pca	pca	NOUN
cana-4362	58	9	and	and	CCONJ
cana-4362	58	10	recursive	recursive	ADJ
cana-4362	58	11	feature	feature	NOUN
cana-4362	58	12	elimination	elimination	NOUN
cana-4362	58	13	(	(	PUNCT
cana-4362	58	14	rfe	rfe	NOUN
cana-4362	58	15	)	)	PUNCT
cana-4362	58	16	for	for	ADP
cana-4362	58	17	initial	initial	ADJ
cana-4362	58	18	dimensionality	dimensionality	NOUN
cana-4362	58	19	reduction	reduction	NOUN
cana-4362	58	20	,	,	PUNCT
cana-4362	58	21	followed	follow	VERB
cana-4362	58	22	by	by	ADP
cana-4362	58	23	deep	deep	ADJ
cana-4362	58	24	learning	learning	NOUN
cana-4362	58	25	-	-	PUNCT
cana-4362	58	26	based	base	VERB
cana-4362	58	27	classification	classification	NOUN
cana-4362	58	28	using	use	VERB
cana-4362	58	29	lstm	lstm	ADJ
cana-4362	58	30	networks	network	NOUN
cana-4362	58	31	.	.	PUNCT
cana-4362	59	1	they	they	PRON
cana-4362	59	2	demonstrate	demonstrate	VERB
cana-4362	59	3	the	the	DET
cana-4362	59	4	effectiveness	effectiveness	NOUN
cana-4362	59	5	of	of	ADP
cana-4362	59	6	this	this	DET
cana-4362	59	7	hybrid	hybrid	ADJ
cana-4362	59	8	approach	approach	NOUN
cana-4362	59	9	through	through	ADP
cana-4362	59	10	extensive	extensive	ADJ
cana-4362	59	11	experiments	experiment	NOUN
cana-4362	59	12	on	on	ADP
cana-4362	59	13	social	social	ADJ
cana-4362	59	14	media	medium	NOUN
cana-4362	59	15	datasets	dataset	NOUN
cana-4362	59	16	,	,	PUNCT
cana-4362	59	17	particularly	particularly	ADV
cana-4362	59	18	in	in	ADP
cana-4362	59	19	handling	handle	VERB
cana-4362	59	20	noisy	noisy	ADJ
cana-4362	59	21	and	and	CCONJ
cana-4362	59	22	high	high	ADV
cana-4362	59	23	-	-	PUNCT
cana-4362	59	24	dimensional	dimensional	ADJ
cana-4362	59	25	data	datum	NOUN
cana-4362	59	26	.	.	PUNCT
cana-4362	60	1	the	the	DET
cana-4362	60	2	results	result	NOUN
cana-4362	60	3	show	show	VERB
cana-4362	60	4	significant	significant	ADJ
cana-4362	60	5	improvements	improvement	NOUN
cana-4362	60	6	in	in	ADP
cana-4362	60	7	classification	classification	NOUN
cana-4362	60	8	accuracy	accuracy	NOUN
cana-4362	60	9	and	and	CCONJ
cana-4362	60	10	processing	processing	NOUN
cana-4362	60	11	time	time	NOUN
cana-4362	60	12	compared	compare	VERB
cana-4362	60	13	to	to	ADP
cana-4362	60	14	purely	purely	ADV
cana-4362	60	15	deep	deep	ADJ
cana-4362	60	16	learning	learning	NOUN
cana-4362	60	17	-	-	PUNCT
cana-4362	60	18	based	base	VERB
cana-4362	60	19	models	model	NOUN
cana-4362	60	20	.	.	PUNCT
cana-4362	61	1	the	the	DET
cana-4362	61	2	study	study	NOUN
cana-4362	61	3	suggests	suggest	VERB
cana-4362	61	4	that	that	SCONJ
cana-4362	61	5	the	the	DET
cana-4362	61	6	hybrid	hybrid	NOUN
cana-4362	61	7	approach	approach	NOUN
cana-4362	61	8	can	can	AUX
cana-4362	61	9	be	be	AUX
cana-4362	61	10	highly	highly	ADV
cana-4362	61	11	effective	effective	ADJ
cana-4362	61	12	in	in	ADP
cana-4362	61	13	processing	process	VERB
cana-4362	61	14	complex	complex	ADJ
cana-4362	61	15	social	social	ADJ
cana-4362	61	16	media	medium	NOUN
cana-4362	61	17	datasets	dataset	NOUN
cana-4362	61	18	for	for	ADP
cana-4362	61	19	sentiment	sentiment	NOUN
cana-4362	61	20	analysis	analysis	NOUN
cana-4362	61	21	.	.	PUNCT
cana-4362	62	1	li	li	PROPN
cana-4362	62	2	et	et	PROPN
cana-4362	62	3	al	al	PROPN
cana-4362	62	4	.	.	PUNCT
cana-4362	63	1	[	[	X
cana-4362	63	2	3	3	X
cana-4362	63	3	]	]	PUNCT
cana-4362	63	4	proposes	propose	VERB
cana-4362	63	5	a	a	DET
cana-4362	63	6	hybrid	hybrid	ADJ
cana-4362	63	7	feature	feature	NOUN
cana-4362	63	8	selection	selection	NOUN
cana-4362	63	9	method	method	NOUN
cana-4362	63	10	that	that	PRON
cana-4362	63	11	combines	combine	VERB
cana-4362	63	12	both	both	DET
cana-4362	63	13	wrapper	wrapper	NOUN
cana-4362	63	14	and	and	CCONJ
cana-4362	63	15	filter	filter	NOUN
cana-4362	63	16	approaches	approach	NOUN
cana-4362	63	17	to	to	PART
cana-4362	63	18	improve	improve	VERB
cana-4362	63	19	sentiment	sentiment	NOUN
cana-4362	63	20	classification	classification	NOUN
cana-4362	63	21	on	on	ADP
cana-4362	63	22	social	social	ADJ
cana-4362	63	23	media	medium	NOUN
cana-4362	63	24	platforms	platform	NOUN
cana-4362	63	25	like	like	ADP
cana-4362	63	26	twitter	twitter	NOUN
cana-4362	63	27	.	.	PUNCT
cana-4362	64	1	their	their	PRON
cana-4362	64	2	research	research	NOUN
cana-4362	64	3	focuses	focus	VERB
cana-4362	64	4	on	on	ADP
cana-4362	64	5	selecting	select	VERB
cana-4362	64	6	the	the	DET
cana-4362	64	7	most	most	ADV
cana-4362	64	8	relevant	relevant	ADJ
cana-4362	64	9	features	feature	NOUN
cana-4362	64	10	by	by	ADP
cana-4362	64	11	combining	combine	VERB
cana-4362	64	12	statistical	statistical	ADJ
cana-4362	64	13	techniques	technique	NOUN
cana-4362	64	14	like	like	ADP
cana-4362	64	15	chi	chi	NOUN
cana-4362	64	16	-	-	PUNCT
cana-4362	64	17	square	square	ADJ
cana-4362	64	18	and	and	CCONJ
cana-4362	64	19	mutual	mutual	ADJ
cana-4362	64	20	information	information	NOUN
cana-4362	64	21	with	with	ADP
cana-4362	64	22	machine	machine	NOUN
cana-4362	64	23	learning	learn	VERB
cana-4362	64	24	algorithms	algorithm	NOUN
cana-4362	64	25	such	such	ADJ
cana-4362	64	26	as	as	ADP
cana-4362	64	27	support	support	NOUN
cana-4362	64	28	vector	vector	NOUN
cana-4362	64	29	machines	machine	NOUN
cana-4362	64	30	(	(	PUNCT
cana-4362	64	31	svm	svm	PROPN
cana-4362	64	32	)	)	PUNCT
cana-4362	64	33	.	.	PUNCT
cana-4362	65	1	communications	communication	NOUN
cana-4362	65	2	on	on	ADP
cana-4362	65	3	applied	apply	VERB
cana-4362	65	4	nonlinear	nonlinear	ADJ
cana-4362	65	5	analysis	analysis	NOUN
cana-4362	65	6	issn	issn	NOUN
cana-4362	65	7	:	:	PUNCT
cana-4362	65	8	1074	1074	NUM
cana-4362	65	9	-	-	PUNCT
cana-4362	65	10	133x	133x	NUM
cana-4362	65	11	vol	vol	NOUN
cana-4362	65	12	32	32	NUM
cana-4362	65	13	no	no	NOUN
cana-4362	65	14	.	.	PUNCT
cana-4362	66	1	9s	9s	NUM
cana-4362	66	2	(	(	PUNCT
cana-4362	66	3	2025	2025	NUM
cana-4362	66	4	)	)	PUNCT
cana-4362	66	5	1902	1902	NUM
cana-4362	66	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-4362	66	7	additionally	additionally	ADV
cana-4362	66	8	,	,	PUNCT
cana-4362	66	9	they	they	PRON
cana-4362	66	10	employ	employ	VERB
cana-4362	66	11	deep	deep	ADJ
cana-4362	66	12	learning	learning	NOUN
cana-4362	66	13	models	model	NOUN
cana-4362	66	14	,	,	PUNCT
cana-4362	66	15	including	include	VERB
cana-4362	66	16	cnn	cnn	PROPN
cana-4362	66	17	and	and	CCONJ
cana-4362	66	18	lstm	lstm	PROPN
cana-4362	66	19	,	,	PUNCT
cana-4362	66	20	to	to	PART
cana-4362	66	21	refine	refine	VERB
cana-4362	66	22	the	the	DET
cana-4362	66	23	feature	feature	NOUN
cana-4362	66	24	selection	selection	NOUN
cana-4362	66	25	process	process	NOUN
cana-4362	66	26	.	.	PUNCT
cana-4362	67	1	their	their	PRON
cana-4362	67	2	experimental	experimental	ADJ
cana-4362	67	3	results	result	NOUN
cana-4362	67	4	show	show	VERB
cana-4362	67	5	that	that	SCONJ
cana-4362	67	6	the	the	DET
cana-4362	67	7	hybrid	hybrid	NOUN
cana-4362	67	8	method	method	NOUN
cana-4362	67	9	leads	lead	VERB
cana-4362	67	10	to	to	ADP
cana-4362	67	11	better	well	ADJ
cana-4362	67	12	classification	classification	NOUN
cana-4362	67	13	performance	performance	NOUN
cana-4362	67	14	,	,	PUNCT
cana-4362	67	15	with	with	ADP
cana-4362	67	16	a	a	DET
cana-4362	67	17	notable	notable	ADJ
cana-4362	67	18	increase	increase	NOUN
cana-4362	67	19	in	in	ADP
cana-4362	67	20	f1	f1	ADJ
cana-4362	67	21	score	score	NOUN
cana-4362	67	22	compared	compare	VERB
cana-4362	67	23	to	to	ADP
cana-4362	67	24	conventional	conventional	ADJ
cana-4362	67	25	feature	feature	NOUN
cana-4362	67	26	selection	selection	NOUN
cana-4362	67	27	techniques	technique	NOUN
cana-4362	67	28	.	.	PUNCT
cana-4362	68	1	the	the	DET
cana-4362	68	2	study	study	NOUN
cana-4362	68	3	emphasizes	emphasize	VERB
cana-4362	68	4	the	the	DET
cana-4362	68	5	potential	potential	NOUN
cana-4362	68	6	of	of	ADP
cana-4362	68	7	hybrid	hybrid	ADJ
cana-4362	68	8	models	model	NOUN
cana-4362	68	9	to	to	PART
cana-4362	68	10	enhance	enhance	VERB
cana-4362	68	11	sentiment	sentiment	NOUN
cana-4362	68	12	analysis	analysis	NOUN
cana-4362	68	13	in	in	ADP
cana-4362	68	14	big	big	ADJ
cana-4362	68	15	data	datum	NOUN
cana-4362	68	16	environments	environment	NOUN
cana-4362	68	17	.	.	PUNCT
cana-4362	69	1	kumar	kumar	PROPN
cana-4362	69	2	et	et	PROPN
cana-4362	69	3	al	al	PROPN
cana-4362	69	4	.	.	PUNCT
cana-4362	70	1	[	[	X
cana-4362	70	2	4	4	X
cana-4362	70	3	]	]	PUNCT
cana-4362	70	4	introduces	introduce	VERB
cana-4362	70	5	a	a	DET
cana-4362	70	6	hybrid	hybrid	ADJ
cana-4362	70	7	feature	feature	NOUN
cana-4362	70	8	selection	selection	NOUN
cana-4362	70	9	technique	technique	NOUN
cana-4362	70	10	designed	design	VERB
cana-4362	70	11	for	for	ADP
cana-4362	70	12	sentiment	sentiment	NOUN
cana-4362	70	13	classification	classification	NOUN
cana-4362	70	14	in	in	ADP
cana-4362	70	15	social	social	ADJ
cana-4362	70	16	media	medium	NOUN
cana-4362	70	17	datasets	dataset	NOUN
cana-4362	70	18	.	.	PUNCT
cana-4362	71	1	the	the	DET
cana-4362	71	2	authors	author	NOUN
cana-4362	71	3	apply	apply	VERB
cana-4362	71	4	genetic	genetic	ADJ
cana-4362	71	5	algorithms	algorithm	NOUN
cana-4362	71	6	(	(	PUNCT
cana-4362	71	7	ga	ga	NOUN
cana-4362	71	8	)	)	PUNCT
cana-4362	71	9	to	to	PART
cana-4362	71	10	optimize	optimize	VERB
cana-4362	71	11	feature	feature	NOUN
cana-4362	71	12	selection	selection	NOUN
cana-4362	71	13	,	,	PUNCT
cana-4362	71	14	followed	follow	VERB
cana-4362	71	15	by	by	ADP
cana-4362	71	16	cnns	cnn	NOUN
cana-4362	71	17	for	for	ADP
cana-4362	71	18	sentiment	sentiment	NOUN
cana-4362	71	19	classification	classification	NOUN
cana-4362	71	20	.	.	PUNCT
cana-4362	72	1	the	the	DET
cana-4362	72	2	combination	combination	NOUN
cana-4362	72	3	of	of	ADP
cana-4362	72	4	evolutionary	evolutionary	ADJ
cana-4362	72	5	algorithms	algorithm	NOUN
cana-4362	72	6	and	and	CCONJ
cana-4362	72	7	deep	deep	ADJ
cana-4362	72	8	learning	learning	NOUN
cana-4362	72	9	proves	prove	VERB
cana-4362	72	10	effective	effective	ADJ
cana-4362	72	11	in	in	ADP
cana-4362	72	12	reducing	reduce	VERB
cana-4362	72	13	the	the	DET
cana-4362	72	14	computational	computational	ADJ
cana-4362	72	15	complexity	complexity	NOUN
cana-4362	72	16	while	while	SCONJ
cana-4362	72	17	improving	improve	VERB
cana-4362	72	18	accuracy	accuracy	NOUN
cana-4362	72	19	.	.	PUNCT
cana-4362	73	1	tested	test	VERB
cana-4362	73	2	on	on	ADP
cana-4362	73	3	twitter	twitter	PROPN
cana-4362	73	4	data	datum	NOUN
cana-4362	73	5	,	,	PUNCT
cana-4362	73	6	their	their	PRON
cana-4362	73	7	approach	approach	NOUN
cana-4362	73	8	shows	show	VERB
cana-4362	73	9	superior	superior	ADJ
cana-4362	73	10	performance	performance	NOUN
cana-4362	73	11	compared	compare	VERB
cana-4362	73	12	to	to	ADP
cana-4362	73	13	standard	standard	ADJ
cana-4362	73	14	deep	deep	ADJ
cana-4362	73	15	learning	learning	NOUN
cana-4362	73	16	techniques	technique	NOUN
cana-4362	73	17	,	,	PUNCT
cana-4362	73	18	particularly	particularly	ADV
cana-4362	73	19	in	in	ADP
cana-4362	73	20	handling	handle	VERB
cana-4362	73	21	large	large	ADJ
cana-4362	73	22	,	,	PUNCT
cana-4362	73	23	noisy	noisy	ADJ
cana-4362	73	24	datasets	dataset	NOUN
cana-4362	73	25	.	.	PUNCT
cana-4362	74	1	the	the	DET
cana-4362	74	2	paper	paper	NOUN
cana-4362	74	3	concludes	conclude	VERB
cana-4362	74	4	that	that	SCONJ
cana-4362	74	5	the	the	DET
cana-4362	74	6	hybrid	hybrid	NOUN
cana-4362	74	7	model	model	NOUN
cana-4362	74	8	offers	offer	VERB
cana-4362	74	9	a	a	DET
cana-4362	74	10	scalable	scalable	ADJ
cana-4362	74	11	and	and	CCONJ
cana-4362	74	12	robust	robust	ADJ
cana-4362	74	13	solution	solution	NOUN
cana-4362	74	14	for	for	ADP
cana-4362	74	15	sentiment	sentiment	NOUN
cana-4362	74	16	analysis	analysis	NOUN
cana-4362	74	17	tasks	task	NOUN
cana-4362	74	18	across	across	ADP
cana-4362	74	19	different	different	ADJ
cana-4362	74	20	social	social	ADJ
cana-4362	74	21	media	medium	NOUN
cana-4362	74	22	platforms	platform	NOUN
cana-4362	74	23	.	.	PUNCT
cana-4362	75	1	jiang	jiang	PROPN
cana-4362	75	2	et	et	PROPN
cana-4362	75	3	al	al	PROPN
cana-4362	75	4	.	.	PUNCT
cana-4362	76	1	[	[	X
cana-4362	76	2	5	5	NUM
cana-4362	76	3	]	]	PUNCT
cana-4362	76	4	explores	explore	VERB
cana-4362	76	5	a	a	DET
cana-4362	76	6	deep	deep	ADJ
cana-4362	76	7	learning	learning	NOUN
cana-4362	76	8	-	-	PUNCT
cana-4362	76	9	based	base	VERB
cana-4362	76	10	hybrid	hybrid	ADJ
cana-4362	76	11	feature	feature	NOUN
cana-4362	76	12	selection	selection	NOUN
cana-4362	76	13	method	method	NOUN
cana-4362	76	14	for	for	ADP
cana-4362	76	15	sentiment	sentiment	NOUN
cana-4362	76	16	analysis	analysis	NOUN
cana-4362	76	17	using	use	VERB
cana-4362	76	18	social	social	ADJ
cana-4362	76	19	media	medium	NOUN
cana-4362	76	20	datasets	dataset	NOUN
cana-4362	76	21	.	.	PUNCT
cana-4362	77	1	their	their	PRON
cana-4362	77	2	approach	approach	NOUN
cana-4362	77	3	integrates	integrate	VERB
cana-4362	77	4	a	a	DET
cana-4362	77	5	mutual	mutual	ADJ
cana-4362	77	6	information	information	NOUN
cana-4362	77	7	-	-	PUNCT
cana-4362	77	8	based	base	VERB
cana-4362	77	9	filter	filter	NOUN
cana-4362	77	10	method	method	NOUN
cana-4362	77	11	with	with	ADP
cana-4362	77	12	recursive	recursive	ADJ
cana-4362	77	13	feature	feature	NOUN
cana-4362	77	14	elimination	elimination	NOUN
cana-4362	77	15	to	to	PART
cana-4362	77	16	identify	identify	VERB
cana-4362	77	17	the	the	DET
cana-4362	77	18	most	most	ADV
cana-4362	77	19	relevant	relevant	ADJ
cana-4362	77	20	features	feature	NOUN
cana-4362	77	21	for	for	ADP
cana-4362	77	22	sentiment	sentiment	NOUN
cana-4362	77	23	classification	classification	NOUN
cana-4362	77	24	.	.	PUNCT
cana-4362	78	1	they	they	PRON
cana-4362	78	2	then	then	ADV
cana-4362	78	3	employ	employ	VERB
cana-4362	78	4	a	a	DET
cana-4362	78	5	cnn	cnn	PROPN
cana-4362	78	6	-	-	PUNCT
cana-4362	78	7	lstm	lstm	ADJ
cana-4362	78	8	network	network	NOUN
cana-4362	78	9	to	to	PART
cana-4362	78	10	improve	improve	VERB
cana-4362	78	11	the	the	DET
cana-4362	78	12	sentiment	sentiment	NOUN
cana-4362	78	13	classification	classification	NOUN
cana-4362	78	14	results	result	NOUN
cana-4362	78	15	.	.	PUNCT
cana-4362	79	1	the	the	DET
cana-4362	79	2	study	study	NOUN
cana-4362	79	3	highlights	highlight	VERB
cana-4362	79	4	the	the	DET
cana-4362	79	5	hybrid	hybrid	ADJ
cana-4362	79	6	model	model	NOUN
cana-4362	79	7	's	's	PART
cana-4362	79	8	efficiency	efficiency	NOUN
cana-4362	79	9	in	in	ADP
cana-4362	79	10	handling	handle	VERB
cana-4362	79	11	large	large	ADJ
cana-4362	79	12	,	,	PUNCT
cana-4362	79	13	high	high	ADJ
cana-4362	79	14	-	-	PUNCT
cana-4362	79	15	dimensional	dimensional	ADJ
cana-4362	79	16	datasets	dataset	NOUN
cana-4362	79	17	,	,	PUNCT
cana-4362	79	18	common	common	ADJ
cana-4362	79	19	in	in	ADP
cana-4362	79	20	social	social	ADJ
cana-4362	79	21	media	medium	NOUN
cana-4362	79	22	platforms	platform	NOUN
cana-4362	79	23	like	like	ADP
cana-4362	79	24	twitter	twitter	NOUN
cana-4362	79	25	and	and	CCONJ
cana-4362	79	26	facebook	facebook	NOUN
cana-4362	79	27	.	.	PUNCT
cana-4362	80	1	experimental	experimental	ADJ
cana-4362	80	2	results	result	NOUN
cana-4362	80	3	demonstrate	demonstrate	VERB
cana-4362	80	4	the	the	DET
cana-4362	80	5	hybrid	hybrid	NOUN
cana-4362	80	6	model	model	NOUN
cana-4362	80	7	’s	’s	PART
cana-4362	80	8	superior	superior	ADJ
cana-4362	80	9	accuracy	accuracy	NOUN
cana-4362	80	10	,	,	PUNCT
cana-4362	80	11	precision	precision	NOUN
cana-4362	80	12	,	,	PUNCT
cana-4362	80	13	and	and	CCONJ
cana-4362	80	14	recall	recall	VERB
cana-4362	80	15	when	when	SCONJ
cana-4362	80	16	compared	compare	VERB
cana-4362	80	17	to	to	ADP
cana-4362	80	18	baseline	baseline	NOUN
cana-4362	80	19	models	model	NOUN
cana-4362	80	20	,	,	PUNCT
cana-4362	80	21	providing	provide	VERB
cana-4362	80	22	insights	insight	NOUN
cana-4362	80	23	into	into	ADP
cana-4362	80	24	its	its	PRON
cana-4362	80	25	practical	practical	ADJ
cana-4362	80	26	applicability	applicability	NOUN
cana-4362	80	27	in	in	ADP
cana-4362	80	28	real	real	ADJ
cana-4362	80	29	-	-	PUNCT
cana-4362	80	30	world	world	NOUN
cana-4362	80	31	sentiment	sentiment	NOUN
cana-4362	80	32	analysis	analysis	NOUN
cana-4362	80	33	scenarios	scenario	NOUN
cana-4362	80	34	.	.	PUNCT
cana-4362	81	1	wang	wang	PROPN
cana-4362	81	2	et	et	PROPN
cana-4362	81	3	al	al	PROPN
cana-4362	81	4	.	.	PUNCT
cana-4362	82	1	[	[	X
cana-4362	82	2	6	6	NUM
cana-4362	82	3	]	]	PUNCT
cana-4362	82	4	presents	present	VERB
cana-4362	82	5	a	a	DET
cana-4362	82	6	hybrid	hybrid	ADJ
cana-4362	82	7	feature	feature	NOUN
cana-4362	82	8	selection	selection	NOUN
cana-4362	82	9	model	model	NOUN
cana-4362	82	10	for	for	ADP
cana-4362	82	11	sentiment	sentiment	NOUN
cana-4362	82	12	classification	classification	NOUN
cana-4362	82	13	using	use	VERB
cana-4362	82	14	long	long	ADJ
cana-4362	82	15	shortterm	shortterm	PROPN
cana-4362	82	16	memory	memory	NOUN
cana-4362	82	17	(	(	PUNCT
cana-4362	82	18	lstm	lstm	NOUN
cana-4362	82	19	)	)	PUNCT
cana-4362	82	20	networks	network	NOUN
cana-4362	82	21	.	.	PUNCT
cana-4362	83	1	by	by	ADP
cana-4362	83	2	combining	combine	VERB
cana-4362	83	3	both	both	PRON
cana-4362	83	4	traditional	traditional	ADJ
cana-4362	83	5	machine	machine	NOUN
cana-4362	83	6	learning	learning	NOUN
cana-4362	83	7	and	and	CCONJ
cana-4362	83	8	deep	deep	ADJ
cana-4362	83	9	learning	learning	NOUN
cana-4362	83	10	approaches	approach	NOUN
cana-4362	83	11	,	,	PUNCT
cana-4362	83	12	the	the	DET
cana-4362	83	13	authors	author	NOUN
cana-4362	83	14	aim	aim	VERB
cana-4362	83	15	to	to	PART
cana-4362	83	16	enhance	enhance	VERB
cana-4362	83	17	the	the	DET
cana-4362	83	18	classification	classification	NOUN
cana-4362	83	19	performance	performance	NOUN
cana-4362	83	20	on	on	ADP
cana-4362	83	21	twitter	twitter	NOUN
cana-4362	83	22	datasets	dataset	NOUN
cana-4362	83	23	.	.	PUNCT
cana-4362	84	1	the	the	DET
cana-4362	84	2	study	study	NOUN
cana-4362	84	3	introduces	introduce	VERB
cana-4362	84	4	a	a	DET
cana-4362	84	5	novel	novel	ADJ
cana-4362	84	6	feature	feature	NOUN
cana-4362	84	7	selection	selection	NOUN
cana-4362	84	8	algorithm	algorithm	NOUN
cana-4362	84	9	that	that	PRON
cana-4362	84	10	combines	combine	VERB
cana-4362	84	11	correlation	correlation	NOUN
cana-4362	84	12	-	-	PUNCT
cana-4362	84	13	based	base	VERB
cana-4362	84	14	feature	feature	NOUN
cana-4362	84	15	selection	selection	NOUN
cana-4362	84	16	(	(	PUNCT
cana-4362	84	17	cfs	cfs	PROPN
cana-4362	84	18	)	)	PUNCT
cana-4362	84	19	and	and	CCONJ
cana-4362	84	20	genetic	genetic	ADJ
cana-4362	84	21	algorithms	algorithm	NOUN
cana-4362	84	22	.	.	PUNCT
cana-4362	85	1	this	this	DET
cana-4362	85	2	selection	selection	NOUN
cana-4362	85	3	is	be	AUX
cana-4362	85	4	followed	follow	VERB
cana-4362	85	5	by	by	ADP
cana-4362	85	6	lstm	lstm	ADJ
cana-4362	85	7	networks	network	NOUN
cana-4362	85	8	to	to	PART
cana-4362	85	9	capture	capture	VERB
cana-4362	85	10	temporal	temporal	ADJ
cana-4362	85	11	dependencies	dependency	NOUN
cana-4362	85	12	in	in	ADP
cana-4362	85	13	the	the	DET
cana-4362	85	14	data	datum	NOUN
cana-4362	85	15	.	.	PUNCT
cana-4362	86	1	their	their	PRON
cana-4362	86	2	experimental	experimental	ADJ
cana-4362	86	3	evaluation	evaluation	NOUN
cana-4362	86	4	shows	show	VERB
cana-4362	86	5	that	that	SCONJ
cana-4362	86	6	the	the	DET
cana-4362	86	7	hybrid	hybrid	NOUN
cana-4362	86	8	model	model	NOUN
cana-4362	86	9	improves	improve	VERB
cana-4362	86	10	accuracy	accuracy	NOUN
cana-4362	86	11	and	and	CCONJ
cana-4362	86	12	reduces	reduce	VERB
cana-4362	86	13	the	the	DET
cana-4362	86	14	processing	processing	NOUN
cana-4362	86	15	time	time	NOUN
cana-4362	86	16	compared	compare	VERB
cana-4362	86	17	to	to	ADP
cana-4362	86	18	baseline	baseline	ADJ
cana-4362	86	19	methods	method	NOUN
cana-4362	86	20	,	,	PUNCT
cana-4362	86	21	proving	prove	VERB
cana-4362	86	22	effective	effective	ADJ
cana-4362	86	23	for	for	ADP
cana-4362	86	24	realtime	realtime	ADJ
cana-4362	86	25	social	social	ADJ
cana-4362	86	26	media	medium	NOUN
cana-4362	86	27	sentiment	sentiment	NOUN
cana-4362	86	28	classification	classification	NOUN
cana-4362	86	29	.	.	PUNCT
cana-4362	87	1	chen	chen	PROPN
cana-4362	87	2	et	et	PROPN
cana-4362	87	3	al	al	PROPN
cana-4362	87	4	.	.	PUNCT
cana-4362	88	1	[	[	X
cana-4362	88	2	7	7	NUM
cana-4362	88	3	]	]	PUNCT
cana-4362	88	4	addresses	address	NOUN
cana-4362	88	5	multi	multi	ADJ
cana-4362	88	6	-	-	ADJ
cana-4362	88	7	modal	modal	ADJ
cana-4362	88	8	feature	feature	NOUN
cana-4362	88	9	selection	selection	NOUN
cana-4362	88	10	for	for	ADP
cana-4362	88	11	sentiment	sentiment	NOUN
cana-4362	88	12	analysis	analysis	NOUN
cana-4362	88	13	using	use	VERB
cana-4362	88	14	hybrid	hybrid	ADJ
cana-4362	88	15	deep	deep	ADJ
cana-4362	88	16	learning	learning	NOUN
cana-4362	88	17	techniques	technique	NOUN
cana-4362	88	18	.	.	PUNCT
cana-4362	89	1	this	this	DET
cana-4362	89	2	paper	paper	NOUN
cana-4362	89	3	focuses	focus	VERB
cana-4362	89	4	on	on	ADP
cana-4362	89	5	combining	combine	VERB
cana-4362	89	6	visual	visual	ADJ
cana-4362	89	7	and	and	CCONJ
cana-4362	89	8	textual	textual	ADJ
cana-4362	89	9	data	datum	NOUN
cana-4362	89	10	for	for	ADP
cana-4362	89	11	more	more	ADV
cana-4362	89	12	accurate	accurate	ADJ
cana-4362	89	13	sentiment	sentiment	NOUN
cana-4362	89	14	classification	classification	NOUN
cana-4362	89	15	,	,	PUNCT
cana-4362	89	16	particularly	particularly	ADV
cana-4362	89	17	in	in	ADP
cana-4362	89	18	social	social	ADJ
cana-4362	89	19	media	medium	NOUN
cana-4362	89	20	contexts	context	NOUN
cana-4362	89	21	where	where	SCONJ
cana-4362	89	22	users	user	NOUN
cana-4362	89	23	post	post	VERB
cana-4362	89	24	multimedia	multimedia	NOUN
cana-4362	89	25	content	content	NOUN
cana-4362	89	26	.	.	PUNCT
cana-4362	90	1	the	the	DET
cana-4362	90	2	authors	author	NOUN
cana-4362	90	3	propose	propose	VERB
cana-4362	90	4	a	a	DET
cana-4362	90	5	two	two	NUM
cana-4362	90	6	-	-	PUNCT
cana-4362	90	7	stage	stage	NOUN
cana-4362	90	8	feature	feature	NOUN
cana-4362	90	9	selection	selection	NOUN
cana-4362	90	10	process	process	NOUN
cana-4362	90	11	that	that	PRON
cana-4362	90	12	utilizes	utilize	VERB
cana-4362	90	13	both	both	DET
cana-4362	90	14	filter	filter	NOUN
cana-4362	90	15	-	-	PUNCT
cana-4362	90	16	based	base	VERB
cana-4362	90	17	and	and	CCONJ
cana-4362	90	18	embedded	embed	VERB
cana-4362	90	19	methods	method	NOUN
cana-4362	90	20	to	to	PART
cana-4362	90	21	reduce	reduce	VERB
cana-4362	90	22	the	the	DET
cana-4362	90	23	dimensionality	dimensionality	NOUN
cana-4362	90	24	of	of	ADP
cana-4362	90	25	the	the	DET
cana-4362	90	26	data	datum	NOUN
cana-4362	90	27	.	.	PUNCT
cana-4362	91	1	subsequently	subsequently	ADV
cana-4362	91	2	,	,	PUNCT
cana-4362	91	3	a	a	DET
cana-4362	91	4	cnn	cnn	PROPN
cana-4362	91	5	-	-	PUNCT
cana-4362	91	6	rnn	rnn	PROPN
cana-4362	91	7	hybrid	hybrid	ADJ
cana-4362	91	8	model	model	NOUN
cana-4362	91	9	is	be	AUX
cana-4362	91	10	applied	apply	VERB
cana-4362	91	11	to	to	PART
cana-4362	91	12	enhance	enhance	VERB
cana-4362	91	13	sentiment	sentiment	NOUN
cana-4362	91	14	classification	classification	NOUN
cana-4362	91	15	.	.	PUNCT
cana-4362	92	1	the	the	DET
cana-4362	92	2	experimental	experimental	ADJ
cana-4362	92	3	results	result	NOUN
cana-4362	92	4	show	show	VERB
cana-4362	92	5	that	that	SCONJ
cana-4362	92	6	integrating	integrate	VERB
cana-4362	92	7	multi	multi	ADJ
cana-4362	92	8	-	-	ADJ
cana-4362	92	9	modal	modal	ADJ
cana-4362	92	10	data	datum	NOUN
cana-4362	92	11	leads	lead	VERB
cana-4362	92	12	to	to	ADP
cana-4362	92	13	better	well	ADJ
cana-4362	92	14	classification	classification	NOUN
cana-4362	92	15	performance	performance	NOUN
cana-4362	92	16	,	,	PUNCT
cana-4362	92	17	with	with	ADP
cana-4362	92	18	the	the	DET
cana-4362	92	19	hybrid	hybrid	ADJ
cana-4362	92	20	approach	approach	NOUN
cana-4362	92	21	significantly	significantly	ADV
cana-4362	92	22	outperforming	outperform	VERB
cana-4362	92	23	unimodal	unimodal	ADJ
cana-4362	92	24	models	model	NOUN
cana-4362	92	25	in	in	ADP
cana-4362	92	26	terms	term	NOUN
cana-4362	92	27	of	of	ADP
cana-4362	92	28	accuracy	accuracy	NOUN
cana-4362	92	29	and	and	CCONJ
cana-4362	92	30	recall	recall	NOUN
cana-4362	92	31	.	.	PUNCT
cana-4362	93	1	huang	huang	PROPN
cana-4362	93	2	et	et	PROPN
cana-4362	93	3	al	al	PROPN
cana-4362	93	4	.	.	PUNCT
cana-4362	94	1	[	[	X
cana-4362	94	2	8	8	NUM
cana-4362	94	3	]	]	PUNCT
cana-4362	94	4	explores	explore	VERB
cana-4362	94	5	hybrid	hybrid	ADJ
cana-4362	94	6	feature	feature	NOUN
cana-4362	94	7	engineering	engineering	NOUN
cana-4362	94	8	techniques	technique	NOUN
cana-4362	94	9	for	for	ADP
cana-4362	94	10	sentiment	sentiment	NOUN
cana-4362	94	11	analysis	analysis	NOUN
cana-4362	94	12	of	of	ADP
cana-4362	94	13	social	social	ADJ
cana-4362	94	14	media	medium	NOUN
cana-4362	94	15	posts	post	NOUN
cana-4362	94	16	using	use	VERB
cana-4362	94	17	deep	deep	ADJ
cana-4362	94	18	learning	learning	NOUN
cana-4362	94	19	models	model	NOUN
cana-4362	94	20	.	.	PUNCT
cana-4362	95	1	the	the	DET
cana-4362	95	2	study	study	NOUN
cana-4362	95	3	employs	employ	VERB
cana-4362	95	4	a	a	DET
cana-4362	95	5	combination	combination	NOUN
cana-4362	95	6	of	of	ADP
cana-4362	95	7	feature	feature	NOUN
cana-4362	95	8	extraction	extraction	NOUN
cana-4362	95	9	methods	method	NOUN
cana-4362	95	10	,	,	PUNCT
cana-4362	95	11	such	such	ADJ
cana-4362	95	12	as	as	ADP
cana-4362	95	13	tf	tf	PROPN
cana-4362	95	14	-	-	PUNCT
cana-4362	95	15	idf	idf	PROPN
cana-4362	95	16	and	and	CCONJ
cana-4362	95	17	word2vec	word2vec	X
cana-4362	95	18	,	,	PUNCT
cana-4362	95	19	alongside	alongside	ADP
cana-4362	95	20	feature	feature	NOUN
cana-4362	95	21	selection	selection	NOUN
cana-4362	95	22	algorithms	algorithm	NOUN
cana-4362	95	23	like	like	ADP
cana-4362	95	24	relieff	relieff	NOUN
cana-4362	95	25	and	and	CCONJ
cana-4362	95	26	recursive	recursive	ADJ
cana-4362	95	27	communications	communication	NOUN
cana-4362	95	28	on	on	ADP
cana-4362	95	29	applied	apply	VERB
cana-4362	95	30	nonlinear	nonlinear	ADJ
cana-4362	95	31	analysis	analysis	NOUN
cana-4362	95	32	issn	issn	NOUN
cana-4362	95	33	:	:	PUNCT
cana-4362	95	34	1074	1074	NUM
cana-4362	95	35	-	-	PUNCT
cana-4362	95	36	133x	133x	NUM
cana-4362	95	37	vol	vol	NOUN
cana-4362	95	38	32	32	NUM
cana-4362	96	1	no	no	NOUN
cana-4362	96	2	.	.	PUNCT
cana-4362	97	1	9s	9s	NUM
cana-4362	97	2	(	(	PUNCT
cana-4362	97	3	2025	2025	NUM
cana-4362	97	4	)	)	PUNCT
cana-4362	97	5	1903	1903	NUM
cana-4362	98	1	https://internationalpubls.com	https://internationalpubls.com	X
cana-4362	98	2	feature	feature	NOUN
cana-4362	98	3	elimination	elimination	NOUN
cana-4362	98	4	(	(	PUNCT
cana-4362	98	5	rfe	rfe	NOUN
cana-4362	98	6	)	)	PUNCT
cana-4362	98	7	.	.	PUNCT
cana-4362	99	1	a	a	DET
cana-4362	99	2	hybrid	hybrid	ADJ
cana-4362	99	3	cnn	cnn	PROPN
cana-4362	99	4	-	-	PUNCT
cana-4362	99	5	rnn	rnn	PROPN
cana-4362	99	6	model	model	NOUN
cana-4362	99	7	is	be	AUX
cana-4362	99	8	then	then	ADV
cana-4362	99	9	applied	apply	VERB
cana-4362	99	10	for	for	ADP
cana-4362	99	11	sentiment	sentiment	NOUN
cana-4362	99	12	classification	classification	NOUN
cana-4362	99	13	,	,	PUNCT
cana-4362	99	14	particularly	particularly	ADV
cana-4362	99	15	focusing	focus	VERB
cana-4362	99	16	on	on	ADP
cana-4362	99	17	twitter	twitter	NOUN
cana-4362	99	18	and	and	CCONJ
cana-4362	99	19	reddit	reddit	NOUN
cana-4362	99	20	data	datum	NOUN
cana-4362	99	21	.	.	PUNCT
cana-4362	100	1	the	the	DET
cana-4362	100	2	results	result	NOUN
cana-4362	100	3	indicate	indicate	VERB
cana-4362	100	4	that	that	SCONJ
cana-4362	100	5	this	this	DET
cana-4362	100	6	approach	approach	NOUN
cana-4362	100	7	improves	improve	VERB
cana-4362	100	8	classification	classification	NOUN
cana-4362	100	9	accuracy	accuracy	NOUN
cana-4362	100	10	and	and	CCONJ
cana-4362	100	11	reduces	reduce	VERB
cana-4362	100	12	overfitting	overfitte	VERB
cana-4362	100	13	,	,	PUNCT
cana-4362	100	14	especially	especially	ADV
cana-4362	100	15	in	in	ADP
cana-4362	100	16	noisy	noisy	ADJ
cana-4362	100	17	datasets	dataset	NOUN
cana-4362	100	18	.	.	PUNCT
cana-4362	101	1	the	the	DET
cana-4362	101	2	paper	paper	NOUN
cana-4362	101	3	concludes	conclude	VERB
cana-4362	101	4	that	that	SCONJ
cana-4362	101	5	hybrid	hybrid	ADJ
cana-4362	101	6	feature	feature	NOUN
cana-4362	101	7	engineering	engineering	NOUN
cana-4362	101	8	is	be	AUX
cana-4362	101	9	crucial	crucial	ADJ
cana-4362	101	10	in	in	ADP
cana-4362	101	11	enhancing	enhance	VERB
cana-4362	101	12	the	the	DET
cana-4362	101	13	performance	performance	NOUN
cana-4362	101	14	of	of	ADP
cana-4362	101	15	deep	deep	ADJ
cana-4362	101	16	learning	learning	NOUN
cana-4362	101	17	models	model	NOUN
cana-4362	101	18	for	for	ADP
cana-4362	101	19	sentiment	sentiment	NOUN
cana-4362	101	20	analysis	analysis	NOUN
cana-4362	101	21	tasks	task	NOUN
cana-4362	101	22	.	.	PUNCT
cana-4362	102	1	liu	liu	PROPN
cana-4362	102	2	et	et	PROPN
cana-4362	102	3	al	al	PROPN
cana-4362	102	4	.	.	PUNCT
cana-4362	103	1	[	[	X
cana-4362	103	2	9	9	NUM
cana-4362	103	3	]	]	PUNCT
cana-4362	103	4	proposes	propose	VERB
cana-4362	103	5	a	a	DET
cana-4362	103	6	hybrid	hybrid	ADJ
cana-4362	103	7	feature	feature	NOUN
cana-4362	103	8	selection	selection	NOUN
cana-4362	103	9	approach	approach	NOUN
cana-4362	103	10	for	for	ADP
cana-4362	103	11	sentiment	sentiment	NOUN
cana-4362	103	12	classification	classification	NOUN
cana-4362	103	13	on	on	ADP
cana-4362	103	14	twitter	twitter	NOUN
cana-4362	103	15	data	datum	NOUN
cana-4362	103	16	using	use	VERB
cana-4362	103	17	sentiment	sentiment	NOUN
cana-4362	103	18	lexicons	lexicon	NOUN
cana-4362	103	19	and	and	CCONJ
cana-4362	103	20	convolutional	convolutional	ADJ
cana-4362	103	21	neural	neural	ADJ
cana-4362	103	22	networks	network	NOUN
cana-4362	103	23	(	(	PUNCT
cana-4362	103	24	cnns	cnns	PROPN
cana-4362	103	25	)	)	PUNCT
cana-4362	103	26	.	.	PUNCT
cana-4362	104	1	the	the	DET
cana-4362	104	2	authors	author	NOUN
cana-4362	104	3	suggest	suggest	VERB
cana-4362	104	4	that	that	SCONJ
cana-4362	104	5	combining	combine	VERB
cana-4362	104	6	lexicon	lexicon	NOUN
cana-4362	104	7	-	-	PUNCT
cana-4362	104	8	based	base	VERB
cana-4362	104	9	features	feature	NOUN
cana-4362	104	10	with	with	ADP
cana-4362	104	11	deep	deep	ADJ
cana-4362	104	12	learning	learning	NOUN
cana-4362	104	13	techniques	technique	NOUN
cana-4362	104	14	can	can	AUX
cana-4362	104	15	enhance	enhance	VERB
cana-4362	104	16	sentiment	sentiment	NOUN
cana-4362	104	17	detection	detection	NOUN
cana-4362	104	18	,	,	PUNCT
cana-4362	104	19	particularly	particularly	ADV
cana-4362	104	20	in	in	ADP
cana-4362	104	21	noisy	noisy	ADJ
cana-4362	104	22	social	social	ADJ
cana-4362	104	23	media	medium	NOUN
cana-4362	104	24	environments	environment	NOUN
cana-4362	104	25	.	.	PUNCT
cana-4362	105	1	the	the	DET
cana-4362	105	2	study	study	NOUN
cana-4362	105	3	demonstrates	demonstrate	VERB
cana-4362	105	4	that	that	SCONJ
cana-4362	105	5	the	the	DET
cana-4362	105	6	hybrid	hybrid	NOUN
cana-4362	105	7	method	method	NOUN
cana-4362	105	8	improves	improve	VERB
cana-4362	105	9	classification	classification	NOUN
cana-4362	105	10	accuracy	accuracy	NOUN
cana-4362	105	11	while	while	SCONJ
cana-4362	105	12	maintaining	maintain	VERB
cana-4362	105	13	computational	computational	ADJ
cana-4362	105	14	efficiency	efficiency	NOUN
cana-4362	105	15	,	,	PUNCT
cana-4362	105	16	making	make	VERB
cana-4362	105	17	it	it	PRON
cana-4362	105	18	suitable	suitable	ADJ
cana-4362	105	19	for	for	ADP
cana-4362	105	20	large	large	ADJ
cana-4362	105	21	-	-	PUNCT
cana-4362	105	22	scale	scale	NOUN
cana-4362	105	23	sentiment	sentiment	NOUN
cana-4362	105	24	analysis	analysis	NOUN
cana-4362	105	25	tasks	task	NOUN
cana-4362	105	26	.	.	PUNCT
cana-4362	106	1	their	their	PRON
cana-4362	106	2	experimental	experimental	ADJ
cana-4362	106	3	results	result	NOUN
cana-4362	106	4	show	show	VERB
cana-4362	106	5	a	a	DET
cana-4362	106	6	significant	significant	ADJ
cana-4362	106	7	improvement	improvement	NOUN
cana-4362	106	8	in	in	ADP
cana-4362	106	9	accuracy	accuracy	NOUN
cana-4362	106	10	,	,	PUNCT
cana-4362	106	11	precision	precision	NOUN
cana-4362	106	12	,	,	PUNCT
cana-4362	106	13	and	and	CCONJ
cana-4362	106	14	recall	recall	NOUN
cana-4362	106	15	compared	compare	VERB
cana-4362	106	16	to	to	ADP
cana-4362	106	17	traditional	traditional	ADJ
cana-4362	106	18	lexicon	lexicon	NOUN
cana-4362	106	19	-	-	PUNCT
cana-4362	106	20	based	base	VERB
cana-4362	106	21	or	or	CCONJ
cana-4362	106	22	cnn	cnn	PROPN
cana-4362	106	23	-	-	PUNCT
cana-4362	106	24	only	only	ADJ
cana-4362	106	25	models	model	NOUN
cana-4362	106	26	.	.	PUNCT
cana-4362	107	1	gupta	gupta	NOUN
cana-4362	107	2	et	et	PROPN
cana-4362	107	3	al	al	PROPN
cana-4362	107	4	.	.	PUNCT
cana-4362	108	1	[	[	X
cana-4362	108	2	10	10	NUM
cana-4362	108	3	]	]	PUNCT
cana-4362	108	4	combines	combine	VERB
cana-4362	108	5	feature	feature	NOUN
cana-4362	108	6	selection	selection	NOUN
cana-4362	108	7	with	with	ADP
cana-4362	108	8	deep	deep	ADJ
cana-4362	108	9	learning	learning	NOUN
cana-4362	108	10	for	for	ADP
cana-4362	108	11	sentiment	sentiment	NOUN
cana-4362	108	12	analysis	analysis	NOUN
cana-4362	108	13	in	in	ADP
cana-4362	108	14	social	social	ADJ
cana-4362	108	15	media	medium	NOUN
cana-4362	108	16	datasets	dataset	NOUN
cana-4362	108	17	.	.	PUNCT
cana-4362	109	1	their	their	PRON
cana-4362	109	2	hybrid	hybrid	ADJ
cana-4362	109	3	model	model	NOUN
cana-4362	109	4	integrates	integrate	VERB
cana-4362	109	5	principal	principal	ADJ
cana-4362	109	6	component	component	NOUN
cana-4362	109	7	analysis	analysis	NOUN
cana-4362	109	8	(	(	PUNCT
cana-4362	109	9	pca	pca	NOUN
cana-4362	109	10	)	)	PUNCT
cana-4362	109	11	for	for	ADP
cana-4362	109	12	dimensionality	dimensionality	NOUN
cana-4362	109	13	reduction	reduction	NOUN
cana-4362	109	14	with	with	ADP
cana-4362	109	15	a	a	DET
cana-4362	109	16	cnn	cnn	PROPN
cana-4362	109	17	-	-	PUNCT
cana-4362	109	18	lstm	lstm	ADJ
cana-4362	109	19	network	network	NOUN
cana-4362	109	20	for	for	ADP
cana-4362	109	21	classification	classification	NOUN
cana-4362	109	22	.	.	PUNCT
cana-4362	110	1	the	the	DET
cana-4362	110	2	authors	author	NOUN
cana-4362	110	3	focus	focus	VERB
cana-4362	110	4	on	on	ADP
cana-4362	110	5	improving	improve	VERB
cana-4362	110	6	sentiment	sentiment	NOUN
cana-4362	110	7	classification	classification	NOUN
cana-4362	110	8	accuracy	accuracy	NOUN
cana-4362	110	9	while	while	SCONJ
cana-4362	110	10	reducing	reduce	VERB
cana-4362	110	11	the	the	DET
cana-4362	110	12	computational	computational	ADJ
cana-4362	110	13	complexity	complexity	NOUN
cana-4362	110	14	associated	associate	VERB
cana-4362	110	15	with	with	ADP
cana-4362	110	16	deep	deep	ADJ
cana-4362	110	17	learning	learning	NOUN
cana-4362	110	18	models	model	NOUN
cana-4362	110	19	.	.	PUNCT
cana-4362	111	1	tested	test	VERB
cana-4362	111	2	on	on	ADP
cana-4362	111	3	twitter	twitter	NOUN
cana-4362	111	4	and	and	CCONJ
cana-4362	111	5	facebook	facebook	PROPN
cana-4362	111	6	data	data	PROPN
cana-4362	111	7	,	,	PUNCT
cana-4362	111	8	their	their	PRON
cana-4362	111	9	approach	approach	NOUN
cana-4362	111	10	outperforms	outperform	VERB
cana-4362	111	11	standard	standard	ADJ
cana-4362	111	12	deep	deep	ADJ
cana-4362	111	13	learning	learning	NOUN
cana-4362	111	14	models	model	NOUN
cana-4362	111	15	in	in	ADP
cana-4362	111	16	both	both	PRON
cana-4362	111	17	accuracy	accuracy	NOUN
cana-4362	111	18	and	and	CCONJ
cana-4362	111	19	processing	processing	NOUN
cana-4362	111	20	time	time	NOUN
cana-4362	111	21	.	.	PUNCT
cana-4362	112	1	the	the	DET
cana-4362	112	2	paper	paper	NOUN
cana-4362	112	3	concludes	conclude	VERB
cana-4362	112	4	that	that	SCONJ
cana-4362	112	5	hybrid	hybrid	NOUN
cana-4362	112	6	models	model	NOUN
cana-4362	112	7	offer	offer	VERB
cana-4362	112	8	a	a	DET
cana-4362	112	9	promising	promising	ADJ
cana-4362	112	10	solution	solution	NOUN
cana-4362	112	11	for	for	ADP
cana-4362	112	12	sentiment	sentiment	NOUN
cana-4362	112	13	analysis	analysis	NOUN
cana-4362	112	14	tasks	task	NOUN
cana-4362	112	15	on	on	ADP
cana-4362	112	16	large	large	ADJ
cana-4362	112	17	-	-	PUNCT
cana-4362	112	18	scale	scale	NOUN
cana-4362	112	19	social	social	ADJ
cana-4362	112	20	media	medium	NOUN
cana-4362	112	21	datasets	dataset	NOUN
cana-4362	112	22	.	.	PUNCT
cana-4362	113	1	zhang	zhang	PROPN
cana-4362	113	2	et	et	PROPN
cana-4362	113	3	al	al	PROPN
cana-4362	113	4	.	.	PUNCT
cana-4362	114	1	[	[	X
cana-4362	114	2	11	11	NUM
cana-4362	114	3	]	]	PUNCT
cana-4362	114	4	presents	present	VERB
cana-4362	114	5	a	a	DET
cana-4362	114	6	hybrid	hybrid	ADJ
cana-4362	114	7	feature	feature	NOUN
cana-4362	114	8	selection	selection	NOUN
cana-4362	114	9	method	method	NOUN
cana-4362	114	10	using	use	VERB
cana-4362	114	11	genetic	genetic	ADJ
cana-4362	114	12	algorithms	algorithm	NOUN
cana-4362	114	13	(	(	PUNCT
cana-4362	114	14	ga	ga	NOUN
cana-4362	114	15	)	)	PUNCT
cana-4362	114	16	combined	combine	VERB
cana-4362	114	17	with	with	ADP
cana-4362	114	18	convolutional	convolutional	ADJ
cana-4362	114	19	neural	neural	ADJ
cana-4362	114	20	networks	network	NOUN
cana-4362	114	21	(	(	PUNCT
cana-4362	114	22	cnn	cnn	PROPN
cana-4362	114	23	)	)	PUNCT
cana-4362	114	24	for	for	ADP
cana-4362	114	25	sentiment	sentiment	NOUN
cana-4362	114	26	classification	classification	NOUN
cana-4362	114	27	of	of	ADP
cana-4362	114	28	social	social	ADJ
cana-4362	114	29	media	medium	NOUN
cana-4362	114	30	data	datum	NOUN
cana-4362	114	31	.	.	PUNCT
cana-4362	115	1	the	the	DET
cana-4362	115	2	paper	paper	NOUN
cana-4362	115	3	emphasizes	emphasize	VERB
cana-4362	115	4	the	the	DET
cana-4362	115	5	importance	importance	NOUN
cana-4362	115	6	of	of	ADP
cana-4362	115	7	balancing	balance	VERB
cana-4362	115	8	feature	feature	NOUN
cana-4362	115	9	selection	selection	NOUN
cana-4362	115	10	to	to	PART
cana-4362	115	11	reduce	reduce	VERB
cana-4362	115	12	data	data	NOUN
cana-4362	115	13	dimensionality	dimensionality	NOUN
cana-4362	115	14	while	while	SCONJ
cana-4362	115	15	maintaining	maintain	VERB
cana-4362	115	16	model	model	NOUN
cana-4362	115	17	performance	performance	NOUN
cana-4362	115	18	.	.	PUNCT
cana-4362	116	1	the	the	DET
cana-4362	116	2	authors	author	NOUN
cana-4362	116	3	test	test	VERB
cana-4362	116	4	their	their	PRON
cana-4362	116	5	approach	approach	NOUN
cana-4362	116	6	on	on	ADP
cana-4362	116	7	twitter	twitter	NOUN
cana-4362	116	8	datasets	dataset	NOUN
cana-4362	116	9	,	,	PUNCT
cana-4362	116	10	showing	show	VERB
cana-4362	116	11	that	that	SCONJ
cana-4362	116	12	the	the	DET
cana-4362	116	13	hybrid	hybrid	PROPN
cana-4362	116	14	ga	ga	PROPN
cana-4362	116	15	-	-	PUNCT
cana-4362	116	16	cnn	cnn	PROPN
cana-4362	116	17	model	model	NOUN
cana-4362	116	18	significantly	significantly	ADV
cana-4362	116	19	outperforms	outperform	VERB
cana-4362	116	20	traditional	traditional	ADJ
cana-4362	116	21	models	model	NOUN
cana-4362	116	22	in	in	ADP
cana-4362	116	23	both	both	PRON
cana-4362	116	24	accuracy	accuracy	NOUN
cana-4362	116	25	and	and	CCONJ
cana-4362	116	26	processing	processing	NOUN
cana-4362	116	27	time	time	NOUN
cana-4362	116	28	.	.	PUNCT
cana-4362	117	1	by	by	ADP
cana-4362	117	2	incorporating	incorporate	VERB
cana-4362	117	3	evolutionary	evolutionary	ADJ
cana-4362	117	4	algorithms	algorithm	NOUN
cana-4362	117	5	for	for	ADP
cana-4362	117	6	feature	feature	NOUN
cana-4362	117	7	selection	selection	NOUN
cana-4362	117	8	,	,	PUNCT
cana-4362	117	9	the	the	DET
cana-4362	117	10	study	study	NOUN
cana-4362	117	11	achieves	achieve	VERB
cana-4362	117	12	a	a	DET
cana-4362	117	13	higher	high	ADJ
cana-4362	117	14	precision	precision	NOUN
cana-4362	117	15	rate	rate	NOUN
cana-4362	117	16	and	and	CCONJ
cana-4362	117	17	reduces	reduce	VERB
cana-4362	117	18	computational	computational	ADJ
cana-4362	117	19	overhead	overhead	NOUN
cana-4362	117	20	.	.	PUNCT
cana-4362	118	1	the	the	DET
cana-4362	118	2	results	result	NOUN
cana-4362	118	3	highlight	highlight	VERB
cana-4362	118	4	that	that	SCONJ
cana-4362	118	5	combining	combine	VERB
cana-4362	118	6	genetic	genetic	ADJ
cana-4362	118	7	algorithms	algorithm	NOUN
cana-4362	118	8	with	with	ADP
cana-4362	118	9	deep	deep	ADJ
cana-4362	118	10	learning	learning	NOUN
cana-4362	118	11	models	model	NOUN
cana-4362	118	12	provides	provide	VERB
cana-4362	118	13	an	an	DET
cana-4362	118	14	efficient	efficient	ADJ
cana-4362	118	15	approach	approach	NOUN
cana-4362	118	16	for	for	ADP
cana-4362	118	17	sentiment	sentiment	NOUN
cana-4362	118	18	classification	classification	NOUN
cana-4362	118	19	,	,	PUNCT
cana-4362	118	20	particularly	particularly	ADV
cana-4362	118	21	for	for	ADP
cana-4362	118	22	large	large	ADJ
cana-4362	118	23	-	-	PUNCT
cana-4362	118	24	scale	scale	NOUN
cana-4362	118	25	social	social	ADJ
cana-4362	118	26	media	medium	NOUN
cana-4362	118	27	datasets	dataset	NOUN
cana-4362	118	28	where	where	SCONJ
cana-4362	118	29	traditional	traditional	ADJ
cana-4362	118	30	methods	method	NOUN
cana-4362	118	31	may	may	AUX
cana-4362	118	32	falter	falter	VERB
cana-4362	118	33	.	.	PUNCT
cana-4362	119	1	hu	hu	PROPN
cana-4362	119	2	et	et	PROPN
cana-4362	119	3	al	al	PROPN
cana-4362	119	4	.	.	PUNCT
cana-4362	120	1	[	[	X
cana-4362	120	2	12	12	NUM
cana-4362	120	3	]	]	PUNCT
cana-4362	120	4	explores	explore	VERB
cana-4362	120	5	the	the	DET
cana-4362	120	6	combination	combination	NOUN
cana-4362	120	7	of	of	ADP
cana-4362	120	8	cnn	cnn	PROPN
cana-4362	120	9	and	and	CCONJ
cana-4362	120	10	lstm	lstm	NOUN
cana-4362	120	11	models	model	NOUN
cana-4362	120	12	for	for	ADP
cana-4362	120	13	sentiment	sentiment	NOUN
cana-4362	120	14	classification	classification	NOUN
cana-4362	120	15	using	use	VERB
cana-4362	120	16	twitter	twitter	NOUN
cana-4362	120	17	data	datum	NOUN
cana-4362	120	18	,	,	PUNCT
cana-4362	120	19	with	with	ADP
cana-4362	120	20	a	a	DET
cana-4362	120	21	focus	focus	NOUN
cana-4362	120	22	on	on	ADP
cana-4362	120	23	hybrid	hybrid	ADJ
cana-4362	120	24	feature	feature	NOUN
cana-4362	120	25	selection	selection	NOUN
cana-4362	120	26	.	.	PUNCT
cana-4362	121	1	the	the	DET
cana-4362	121	2	authors	author	NOUN
cana-4362	121	3	propose	propose	VERB
cana-4362	121	4	an	an	DET
cana-4362	121	5	integration	integration	NOUN
cana-4362	121	6	of	of	ADP
cana-4362	121	7	filterbased	filterbase	VERB
cana-4362	121	8	and	and	CCONJ
cana-4362	121	9	wrapper	wrapper	NOUN
cana-4362	121	10	-	-	PUNCT
cana-4362	121	11	based	base	VERB
cana-4362	121	12	feature	feature	NOUN
cana-4362	121	13	selection	selection	NOUN
cana-4362	121	14	methods	method	NOUN
cana-4362	121	15	to	to	PART
cana-4362	121	16	enhance	enhance	VERB
cana-4362	121	17	the	the	DET
cana-4362	121	18	performance	performance	NOUN
cana-4362	121	19	of	of	ADP
cana-4362	121	20	deep	deep	ADJ
cana-4362	121	21	learning	learning	NOUN
cana-4362	121	22	models	model	NOUN
cana-4362	121	23	.	.	PUNCT
cana-4362	122	1	the	the	DET
cana-4362	122	2	cnn	cnn	PROPN
cana-4362	122	3	component	component	NOUN
cana-4362	122	4	captures	capture	VERB
cana-4362	122	5	spatial	spatial	ADJ
cana-4362	122	6	dependencies	dependency	NOUN
cana-4362	122	7	,	,	PUNCT
cana-4362	122	8	while	while	SCONJ
cana-4362	122	9	the	the	DET
cana-4362	122	10	lstm	lstm	NOUN
cana-4362	122	11	captures	capture	VERB
cana-4362	122	12	temporal	temporal	ADJ
cana-4362	122	13	relationships	relationship	NOUN
cana-4362	122	14	within	within	ADP
cana-4362	122	15	the	the	DET
cana-4362	122	16	text	text	NOUN
cana-4362	122	17	.	.	PUNCT
cana-4362	123	1	their	their	PRON
cana-4362	123	2	approach	approach	NOUN
cana-4362	123	3	demonstrates	demonstrate	VERB
cana-4362	123	4	superior	superior	ADJ
cana-4362	123	5	performance	performance	NOUN
cana-4362	123	6	in	in	ADP
cana-4362	123	7	terms	term	NOUN
cana-4362	123	8	of	of	ADP
cana-4362	123	9	classification	classification	NOUN
cana-4362	123	10	accuracy	accuracy	NOUN
cana-4362	123	11	,	,	PUNCT
cana-4362	123	12	particularly	particularly	ADV
cana-4362	123	13	for	for	ADP
cana-4362	123	14	sentiment	sentiment	NOUN
cana-4362	123	15	analysis	analysis	NOUN
cana-4362	123	16	on	on	ADP
cana-4362	123	17	short	short	ADJ
cana-4362	123	18	social	social	ADJ
cana-4362	123	19	media	medium	NOUN
cana-4362	123	20	texts	text	NOUN
cana-4362	123	21	.	.	PUNCT
cana-4362	124	1	the	the	DET
cana-4362	124	2	study	study	NOUN
cana-4362	124	3	's	's	PART
cana-4362	124	4	experiments	experiment	NOUN
cana-4362	124	5	show	show	VERB
cana-4362	124	6	that	that	SCONJ
cana-4362	124	7	hybrid	hybrid	ADJ
cana-4362	124	8	feature	feature	NOUN
cana-4362	124	9	selection	selection	NOUN
cana-4362	124	10	improves	improve	VERB
cana-4362	124	11	both	both	CCONJ
cana-4362	124	12	the	the	DET
cana-4362	124	13	precision	precision	NOUN
cana-4362	124	14	and	and	CCONJ
cana-4362	124	15	recall	recall	NOUN
cana-4362	124	16	of	of	ADP
cana-4362	124	17	the	the	DET
cana-4362	124	18	deep	deep	ADJ
cana-4362	124	19	learning	learning	NOUN
cana-4362	124	20	model	model	NOUN
cana-4362	124	21	,	,	PUNCT
cana-4362	124	22	making	make	VERB
cana-4362	124	23	it	it	PRON
cana-4362	124	24	suitable	suitable	ADJ
cana-4362	124	25	for	for	ADP
cana-4362	124	26	real	real	ADJ
cana-4362	124	27	-	-	PUNCT
cana-4362	124	28	time	time	NOUN
cana-4362	124	29	social	social	ADJ
cana-4362	124	30	media	medium	NOUN
cana-4362	124	31	sentiment	sentiment	NOUN
cana-4362	124	32	analysis	analysis	NOUN
cana-4362	124	33	applications	application	NOUN
cana-4362	124	34	.	.	PUNCT
cana-4362	125	1	chen	chen	PROPN
cana-4362	125	2	et	et	PROPN
cana-4362	125	3	al	al	PROPN
cana-4362	125	4	.	.	PUNCT
cana-4362	126	1	[	[	X
cana-4362	126	2	13	13	NUM
cana-4362	126	3	]	]	PUNCT
cana-4362	126	4	investigates	investigate	VERB
cana-4362	126	5	sentiment	sentiment	NOUN
cana-4362	126	6	analysis	analysis	NOUN
cana-4362	126	7	across	across	ADP
cana-4362	126	8	multilingual	multilingual	ADJ
cana-4362	126	9	social	social	ADJ
cana-4362	126	10	media	medium	NOUN
cana-4362	126	11	texts	text	NOUN
cana-4362	126	12	using	use	VERB
cana-4362	126	13	a	a	DET
cana-4362	126	14	hybrid	hybrid	ADJ
cana-4362	126	15	feature	feature	NOUN
cana-4362	126	16	selection	selection	NOUN
cana-4362	126	17	method	method	NOUN
cana-4362	126	18	.	.	PUNCT
cana-4362	127	1	the	the	DET
cana-4362	127	2	study	study	NOUN
cana-4362	127	3	combines	combine	VERB
cana-4362	127	4	emotion	emotion	NOUN
cana-4362	127	5	lexicons	lexicon	NOUN
cana-4362	127	6	with	with	ADP
cana-4362	127	7	deep	deep	ADJ
cana-4362	127	8	learning	learning	NOUN
cana-4362	127	9	models	model	NOUN
cana-4362	127	10	to	to	PART
cana-4362	127	11	handle	handle	VERB
cana-4362	127	12	data	datum	NOUN
cana-4362	127	13	from	from	ADP
cana-4362	127	14	different	different	ADJ
cana-4362	127	15	languages	language	NOUN
cana-4362	127	16	,	,	PUNCT
cana-4362	127	17	emphasizing	emphasize	VERB
cana-4362	127	18	a	a	DET
cana-4362	127	19	multilingual	multilingual	ADJ
cana-4362	127	20	perspective	perspective	NOUN
cana-4362	127	21	.	.	PUNCT
cana-4362	128	1	the	the	DET
cana-4362	128	2	authors	author	NOUN
cana-4362	128	3	propose	propose	VERB
cana-4362	128	4	a	a	DET
cana-4362	128	5	hybrid	hybrid	ADJ
cana-4362	128	6	approach	approach	NOUN
cana-4362	128	7	integrating	integrate	VERB
cana-4362	128	8	traditional	traditional	ADJ
cana-4362	128	9	methods	method	NOUN
cana-4362	128	10	like	like	ADP
cana-4362	128	11	chi	chi	NOUN
cana-4362	128	12	-	-	PUNCT
cana-4362	128	13	square	square	ADJ
cana-4362	128	14	and	and	CCONJ
cana-4362	128	15	information	information	NOUN
cana-4362	128	16	gain	gain	NOUN
cana-4362	128	17	with	with	ADP
cana-4362	128	18	deep	deep	ADJ
cana-4362	128	19	learning	learning	NOUN
cana-4362	128	20	communications	communication	NOUN
cana-4362	128	21	on	on	ADP
cana-4362	128	22	applied	apply	VERB
cana-4362	128	23	nonlinear	nonlinear	ADJ
cana-4362	128	24	analysis	analysis	NOUN
cana-4362	128	25	issn	issn	NOUN
cana-4362	128	26	:	:	PUNCT
cana-4362	128	27	1074	1074	NUM
cana-4362	128	28	-	-	PUNCT
cana-4362	128	29	133x	133x	NUM
cana-4362	128	30	vol	vol	NOUN
cana-4362	128	31	32	32	NUM
cana-4362	128	32	no	no	NOUN
cana-4362	128	33	.	.	PUNCT
cana-4362	129	1	9s	9s	NUM
cana-4362	129	2	(	(	PUNCT
cana-4362	129	3	2025	2025	NUM
cana-4362	129	4	)	)	PUNCT
cana-4362	129	5	1904	1904	NUM
cana-4362	129	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-4362	129	7	techniques	technique	NOUN
cana-4362	129	8	,	,	PUNCT
cana-4362	129	9	specifically	specifically	ADV
cana-4362	129	10	cnns	cnn	VERB
cana-4362	129	11	.	.	PROPN
cana-4362	129	12	tested	test	VERB
cana-4362	129	13	on	on	ADP
cana-4362	129	14	data	datum	NOUN
cana-4362	129	15	from	from	ADP
cana-4362	129	16	twitter	twitter	NOUN
cana-4362	129	17	,	,	PUNCT
cana-4362	129	18	the	the	DET
cana-4362	129	19	approach	approach	NOUN
cana-4362	129	20	shows	show	VERB
cana-4362	129	21	improved	improve	VERB
cana-4362	129	22	classification	classification	NOUN
cana-4362	129	23	accuracy	accuracy	NOUN
cana-4362	129	24	for	for	ADP
cana-4362	129	25	both	both	CCONJ
cana-4362	129	26	monolingual	monolingual	ADJ
cana-4362	129	27	and	and	CCONJ
cana-4362	129	28	multilingual	multilingual	ADJ
cana-4362	129	29	datasets	dataset	NOUN
cana-4362	129	30	.	.	PUNCT
cana-4362	130	1	the	the	DET
cana-4362	130	2	study	study	NOUN
cana-4362	130	3	suggests	suggest	VERB
cana-4362	130	4	that	that	SCONJ
cana-4362	130	5	hybrid	hybrid	ADJ
cana-4362	130	6	feature	feature	NOUN
cana-4362	130	7	selection	selection	NOUN
cana-4362	130	8	helps	help	VERB
cana-4362	130	9	reduce	reduce	VERB
cana-4362	130	10	noise	noise	NOUN
cana-4362	130	11	from	from	ADP
cana-4362	130	12	cross	cross	ADJ
cana-4362	130	13	-	-	ADJ
cana-4362	130	14	lingual	lingual	ADJ
cana-4362	130	15	datasets	dataset	NOUN
cana-4362	130	16	and	and	CCONJ
cana-4362	130	17	is	be	AUX
cana-4362	130	18	effective	effective	ADJ
cana-4362	130	19	in	in	ADP
cana-4362	130	20	multilingual	multilingual	ADJ
cana-4362	130	21	sentiment	sentiment	NOUN
cana-4362	130	22	classification	classification	NOUN
cana-4362	130	23	tasks	task	NOUN
cana-4362	130	24	,	,	PUNCT
cana-4362	130	25	particularly	particularly	ADV
cana-4362	130	26	for	for	ADP
cana-4362	130	27	real	real	ADJ
cana-4362	130	28	-	-	PUNCT
cana-4362	130	29	world	world	NOUN
cana-4362	130	30	social	social	ADJ
cana-4362	130	31	media	medium	NOUN
cana-4362	130	32	applications	application	NOUN
cana-4362	130	33	.	.	PUNCT
cana-4362	131	1	zhou	zhou	PROPN
cana-4362	131	2	et	et	PROPN
cana-4362	131	3	al	al	PROPN
cana-4362	131	4	.	.	PUNCT
cana-4362	132	1	[	[	X
cana-4362	132	2	14	14	NUM
cana-4362	132	3	]	]	PUNCT
cana-4362	132	4	presents	present	VERB
cana-4362	132	5	a	a	DET
cana-4362	132	6	hybrid	hybrid	ADJ
cana-4362	132	7	framework	framework	NOUN
cana-4362	132	8	for	for	ADP
cana-4362	132	9	sentiment	sentiment	NOUN
cana-4362	132	10	classification	classification	NOUN
cana-4362	132	11	that	that	PRON
cana-4362	132	12	combines	combine	VERB
cana-4362	132	13	deep	deep	ADJ
cana-4362	132	14	learning	learning	NOUN
cana-4362	132	15	techniques	technique	NOUN
cana-4362	132	16	with	with	ADP
cana-4362	132	17	traditional	traditional	ADJ
cana-4362	132	18	feature	feature	NOUN
cana-4362	132	19	selection	selection	NOUN
cana-4362	132	20	methods	method	NOUN
cana-4362	132	21	.	.	PUNCT
cana-4362	133	1	the	the	DET
cana-4362	133	2	authors	author	NOUN
cana-4362	133	3	use	use	VERB
cana-4362	133	4	recursive	recursive	ADJ
cana-4362	133	5	feature	feature	NOUN
cana-4362	133	6	elimination	elimination	NOUN
cana-4362	133	7	(	(	PUNCT
cana-4362	133	8	rfe	rfe	NOUN
cana-4362	133	9	)	)	PUNCT
cana-4362	133	10	to	to	PART
cana-4362	133	11	reduce	reduce	VERB
cana-4362	133	12	dimensionality	dimensionality	NOUN
cana-4362	133	13	,	,	PUNCT
cana-4362	133	14	followed	follow	VERB
cana-4362	133	15	by	by	ADP
cana-4362	133	16	a	a	DET
cana-4362	133	17	cnn	cnn	PROPN
cana-4362	133	18	-	-	PUNCT
cana-4362	133	19	lstm	lstm	ADJ
cana-4362	133	20	model	model	NOUN
cana-4362	133	21	for	for	ADP
cana-4362	133	22	classification	classification	NOUN
cana-4362	133	23	.	.	PUNCT
cana-4362	134	1	their	their	PRON
cana-4362	134	2	approach	approach	NOUN
cana-4362	134	3	shows	show	VERB
cana-4362	134	4	substantial	substantial	ADJ
cana-4362	134	5	improvements	improvement	NOUN
cana-4362	134	6	in	in	ADP
cana-4362	134	7	sentiment	sentiment	NOUN
cana-4362	134	8	classification	classification	NOUN
cana-4362	134	9	accuracy	accuracy	NOUN
cana-4362	134	10	compared	compare	VERB
cana-4362	134	11	to	to	ADP
cana-4362	134	12	baseline	baseline	ADJ
cana-4362	134	13	methods	method	NOUN
cana-4362	134	14	.	.	PUNCT
cana-4362	135	1	the	the	DET
cana-4362	135	2	hybrid	hybrid	NOUN
cana-4362	135	3	model	model	NOUN
cana-4362	135	4	is	be	AUX
cana-4362	135	5	tested	test	VERB
cana-4362	135	6	on	on	ADP
cana-4362	135	7	social	social	ADJ
cana-4362	135	8	media	medium	NOUN
cana-4362	135	9	datasets	dataset	NOUN
cana-4362	135	10	like	like	ADP
cana-4362	135	11	twitter	twitter	NOUN
cana-4362	135	12	and	and	CCONJ
cana-4362	135	13	reddit	reddit	NOUN
cana-4362	135	14	,	,	PUNCT
cana-4362	135	15	demonstrating	demonstrate	VERB
cana-4362	135	16	its	its	PRON
cana-4362	135	17	robustness	robustness	NOUN
cana-4362	135	18	in	in	ADP
cana-4362	135	19	handling	handle	VERB
cana-4362	135	20	noisy	noisy	ADJ
cana-4362	135	21	and	and	CCONJ
cana-4362	135	22	complex	complex	ADJ
cana-4362	135	23	data	datum	NOUN
cana-4362	135	24	.	.	PUNCT
cana-4362	136	1	the	the	DET
cana-4362	136	2	study	study	NOUN
cana-4362	136	3	concludes	conclude	VERB
cana-4362	136	4	that	that	SCONJ
cana-4362	136	5	the	the	DET
cana-4362	136	6	hybrid	hybrid	ADJ
cana-4362	136	7	framework	framework	NOUN
cana-4362	136	8	is	be	AUX
cana-4362	136	9	particularly	particularly	ADV
cana-4362	136	10	useful	useful	ADJ
cana-4362	136	11	for	for	ADP
cana-4362	136	12	large	large	ADJ
cana-4362	136	13	-	-	PUNCT
cana-4362	136	14	scale	scale	NOUN
cana-4362	136	15	social	social	ADJ
cana-4362	136	16	media	medium	NOUN
cana-4362	136	17	sentiment	sentiment	NOUN
cana-4362	136	18	classification	classification	NOUN
cana-4362	136	19	,	,	PUNCT
cana-4362	136	20	offering	offer	VERB
cana-4362	136	21	a	a	DET
cana-4362	136	22	practical	practical	ADJ
cana-4362	136	23	solution	solution	NOUN
cana-4362	136	24	for	for	ADP
cana-4362	136	25	real	real	ADJ
cana-4362	136	26	-	-	PUNCT
cana-4362	136	27	world	world	NOUN
cana-4362	136	28	applications	application	NOUN
cana-4362	136	29	where	where	SCONJ
cana-4362	136	30	data	datum	NOUN
cana-4362	136	31	complexity	complexity	NOUN
cana-4362	136	32	and	and	CCONJ
cana-4362	136	33	noise	noise	NOUN
cana-4362	136	34	are	be	AUX
cana-4362	136	35	significant	significant	ADJ
cana-4362	136	36	challenges	challenge	NOUN
cana-4362	136	37	.	.	PUNCT
cana-4362	137	1	feng	feng	PROPN
cana-4362	137	2	et	et	PROPN
cana-4362	137	3	al	al	PROPN
cana-4362	137	4	.	.	PUNCT
cana-4362	138	1	[	[	X
cana-4362	138	2	15	15	NUM
cana-4362	138	3	]	]	PUNCT
cana-4362	138	4	investigates	investigate	VERB
cana-4362	138	5	a	a	DET
cana-4362	138	6	hybrid	hybrid	ADJ
cana-4362	138	7	approach	approach	NOUN
cana-4362	138	8	for	for	ADP
cana-4362	138	9	sentiment	sentiment	NOUN
cana-4362	138	10	classification	classification	NOUN
cana-4362	138	11	using	use	VERB
cana-4362	138	12	feature	feature	NOUN
cana-4362	138	13	selection	selection	NOUN
cana-4362	138	14	and	and	CCONJ
cana-4362	138	15	deep	deep	ADJ
cana-4362	138	16	learning	learning	NOUN
cana-4362	138	17	on	on	ADP
cana-4362	138	18	twitter	twitter	NOUN
cana-4362	138	19	data	datum	NOUN
cana-4362	138	20	.	.	PUNCT
cana-4362	139	1	the	the	DET
cana-4362	139	2	authors	author	NOUN
cana-4362	139	3	propose	propose	VERB
cana-4362	139	4	using	use	VERB
cana-4362	139	5	a	a	DET
cana-4362	139	6	combination	combination	NOUN
cana-4362	139	7	of	of	ADP
cana-4362	139	8	filter	filter	NOUN
cana-4362	139	9	-	-	PUNCT
cana-4362	139	10	based	base	VERB
cana-4362	139	11	methods	method	NOUN
cana-4362	139	12	,	,	PUNCT
cana-4362	139	13	such	such	ADJ
cana-4362	139	14	as	as	ADP
cana-4362	139	15	information	information	NOUN
cana-4362	139	16	gain	gain	NOUN
cana-4362	139	17	,	,	PUNCT
cana-4362	139	18	and	and	CCONJ
cana-4362	139	19	a	a	DET
cana-4362	139	20	deep	deep	ADJ
cana-4362	139	21	learning	learning	NOUN
cana-4362	139	22	model	model	NOUN
cana-4362	139	23	composed	compose	VERB
cana-4362	139	24	of	of	ADP
cana-4362	139	25	cnn	cnn	PROPN
cana-4362	139	26	and	and	CCONJ
cana-4362	139	27	lstm	lstm	ADJ
cana-4362	139	28	layers	layer	NOUN
cana-4362	139	29	.	.	PUNCT
cana-4362	140	1	their	their	PRON
cana-4362	140	2	experimental	experimental	ADJ
cana-4362	140	3	results	result	NOUN
cana-4362	140	4	indicate	indicate	VERB
cana-4362	140	5	that	that	SCONJ
cana-4362	140	6	hybrid	hybrid	ADJ
cana-4362	140	7	feature	feature	NOUN
cana-4362	140	8	selection	selection	NOUN
cana-4362	140	9	improves	improve	VERB
cana-4362	140	10	the	the	DET
cana-4362	140	11	model	model	NOUN
cana-4362	140	12	’s	’s	PART
cana-4362	140	13	ability	ability	NOUN
cana-4362	140	14	to	to	PART
cana-4362	140	15	detect	detect	VERB
cana-4362	140	16	sentiment	sentiment	NOUN
cana-4362	140	17	accurately	accurately	ADV
cana-4362	140	18	,	,	PUNCT
cana-4362	140	19	particularly	particularly	ADV
cana-4362	140	20	when	when	SCONJ
cana-4362	140	21	dealing	deal	VERB
cana-4362	140	22	with	with	ADP
cana-4362	140	23	noisy	noisy	ADJ
cana-4362	140	24	,	,	PUNCT
cana-4362	140	25	unstructured	unstructured	ADJ
cana-4362	140	26	data	datum	NOUN
cana-4362	140	27	from	from	ADP
cana-4362	140	28	social	social	ADJ
cana-4362	140	29	media	medium	NOUN
cana-4362	140	30	.	.	PUNCT
cana-4362	141	1	by	by	ADP
cana-4362	141	2	integrating	integrate	VERB
cana-4362	141	3	these	these	DET
cana-4362	141	4	techniques	technique	NOUN
cana-4362	141	5	,	,	PUNCT
cana-4362	141	6	the	the	DET
cana-4362	141	7	authors	author	NOUN
cana-4362	141	8	are	be	AUX
cana-4362	141	9	able	able	ADJ
cana-4362	141	10	to	to	PART
cana-4362	141	11	significantly	significantly	ADV
cana-4362	141	12	reduce	reduce	VERB
cana-4362	141	13	the	the	DET
cana-4362	141	14	dimensionality	dimensionality	NOUN
cana-4362	141	15	of	of	ADP
cana-4362	141	16	the	the	DET
cana-4362	141	17	dataset	dataset	NOUN
cana-4362	141	18	without	without	ADP
cana-4362	141	19	sacrificing	sacrifice	VERB
cana-4362	141	20	performance	performance	NOUN
cana-4362	141	21	.	.	PUNCT
cana-4362	142	1	the	the	DET
cana-4362	142	2	study	study	NOUN
cana-4362	142	3	concludes	conclude	VERB
cana-4362	142	4	that	that	SCONJ
cana-4362	142	5	this	this	DET
cana-4362	142	6	hybrid	hybrid	ADJ
cana-4362	142	7	model	model	NOUN
cana-4362	142	8	provides	provide	VERB
cana-4362	142	9	an	an	DET
cana-4362	142	10	efficient	efficient	ADJ
cana-4362	142	11	and	and	CCONJ
cana-4362	142	12	accurate	accurate	ADJ
cana-4362	142	13	approach	approach	NOUN
cana-4362	142	14	to	to	ADP
cana-4362	142	15	sentiment	sentiment	NOUN
cana-4362	142	16	classification	classification	NOUN
cana-4362	142	17	in	in	ADP
cana-4362	142	18	social	social	ADJ
cana-4362	142	19	media	medium	NOUN
cana-4362	142	20	.	.	PUNCT
cana-4362	143	1	wang	wang	PROPN
cana-4362	143	2	et	et	PROPN
cana-4362	143	3	al	al	PROPN
cana-4362	143	4	.	.	PUNCT
cana-4362	144	1	[	[	X
cana-4362	144	2	16	16	NUM
cana-4362	144	3	]	]	PUNCT
cana-4362	144	4	introduces	introduce	VERB
cana-4362	144	5	a	a	DET
cana-4362	144	6	multi	multi	ADJ
cana-4362	144	7	-	-	ADJ
cana-4362	144	8	layer	layer	ADJ
cana-4362	144	9	hybrid	hybrid	ADJ
cana-4362	144	10	feature	feature	NOUN
cana-4362	144	11	selection	selection	NOUN
cana-4362	144	12	model	model	NOUN
cana-4362	144	13	for	for	ADP
cana-4362	144	14	sentiment	sentiment	NOUN
cana-4362	144	15	classification	classification	NOUN
cana-4362	144	16	in	in	ADP
cana-4362	144	17	social	social	ADJ
cana-4362	144	18	media	medium	NOUN
cana-4362	144	19	datasets	dataset	NOUN
cana-4362	144	20	.	.	PUNCT
cana-4362	145	1	the	the	DET
cana-4362	145	2	approach	approach	NOUN
cana-4362	145	3	incorporates	incorporate	VERB
cana-4362	145	4	both	both	CCONJ
cana-4362	145	5	filter	filter	NOUN
cana-4362	145	6	and	and	CCONJ
cana-4362	145	7	embedded	embed	VERB
cana-4362	145	8	methods	method	NOUN
cana-4362	145	9	,	,	PUNCT
cana-4362	145	10	followed	follow	VERB
cana-4362	145	11	by	by	ADP
cana-4362	145	12	deep	deep	ADJ
cana-4362	145	13	learning	learning	NOUN
cana-4362	145	14	models	model	NOUN
cana-4362	145	15	to	to	PART
cana-4362	145	16	refine	refine	VERB
cana-4362	145	17	sentiment	sentiment	NOUN
cana-4362	145	18	classification	classification	NOUN
cana-4362	145	19	.	.	PUNCT
cana-4362	146	1	the	the	DET
cana-4362	146	2	authors	author	NOUN
cana-4362	146	3	use	use	VERB
cana-4362	146	4	feature	feature	NOUN
cana-4362	146	5	selection	selection	NOUN
cana-4362	146	6	techniques	technique	NOUN
cana-4362	146	7	like	like	ADP
cana-4362	146	8	relieff	relieff	NOUN
cana-4362	146	9	and	and	CCONJ
cana-4362	146	10	recursive	recursive	ADJ
cana-4362	146	11	feature	feature	NOUN
cana-4362	146	12	elimination	elimination	NOUN
cana-4362	146	13	(	(	PUNCT
cana-4362	146	14	rfe	rfe	NOUN
cana-4362	146	15	)	)	PUNCT
cana-4362	146	16	to	to	PART
cana-4362	146	17	improve	improve	VERB
cana-4362	146	18	model	model	NOUN
cana-4362	146	19	performance	performance	NOUN
cana-4362	146	20	,	,	PUNCT
cana-4362	146	21	then	then	ADV
cana-4362	146	22	apply	apply	VERB
cana-4362	146	23	cnn	cnn	PROPN
cana-4362	146	24	and	and	CCONJ
cana-4362	146	25	lstm	lstm	NOUN
cana-4362	146	26	architectures	architecture	NOUN
cana-4362	146	27	for	for	ADP
cana-4362	146	28	the	the	DET
cana-4362	146	29	final	final	ADJ
cana-4362	146	30	classification	classification	NOUN
cana-4362	146	31	task	task	NOUN
cana-4362	146	32	.	.	PUNCT
cana-4362	147	1	the	the	DET
cana-4362	147	2	study	study	NOUN
cana-4362	147	3	shows	show	VERB
cana-4362	147	4	that	that	SCONJ
cana-4362	147	5	the	the	DET
cana-4362	147	6	hybrid	hybrid	ADJ
cana-4362	147	7	approach	approach	NOUN
cana-4362	147	8	achieves	achieve	VERB
cana-4362	147	9	higher	high	ADJ
cana-4362	147	10	accuracy	accuracy	NOUN
cana-4362	147	11	and	and	CCONJ
cana-4362	147	12	faster	fast	ADJ
cana-4362	147	13	processing	processing	NOUN
cana-4362	147	14	times	time	NOUN
cana-4362	147	15	compared	compare	VERB
cana-4362	147	16	to	to	ADP
cana-4362	147	17	traditional	traditional	ADJ
cana-4362	147	18	deep	deep	ADJ
cana-4362	147	19	learning	learning	NOUN
cana-4362	147	20	models	model	NOUN
cana-4362	147	21	.	.	PUNCT
cana-4362	148	1	tested	test	VERB
cana-4362	148	2	on	on	ADP
cana-4362	148	3	large	large	ADJ
cana-4362	148	4	datasets	dataset	NOUN
cana-4362	148	5	from	from	ADP
cana-4362	148	6	twitter	twitter	NOUN
cana-4362	148	7	,	,	PUNCT
cana-4362	148	8	the	the	DET
cana-4362	148	9	results	result	NOUN
cana-4362	148	10	highlight	highlight	VERB
cana-4362	148	11	the	the	DET
cana-4362	148	12	scalability	scalability	NOUN
cana-4362	148	13	and	and	CCONJ
cana-4362	148	14	robustness	robustness	NOUN
cana-4362	148	15	of	of	ADP
cana-4362	148	16	the	the	DET
cana-4362	148	17	proposed	propose	VERB
cana-4362	148	18	method	method	NOUN
cana-4362	148	19	for	for	ADP
cana-4362	148	20	real	real	ADJ
cana-4362	148	21	-	-	PUNCT
cana-4362	148	22	time	time	NOUN
cana-4362	148	23	sentiment	sentiment	NOUN
cana-4362	148	24	analysis	analysis	NOUN
cana-4362	148	25	.	.	PUNCT
cana-4362	149	1	zhao	zhao	PROPN
cana-4362	149	2	et	et	PROPN
cana-4362	149	3	al	al	PROPN
cana-4362	149	4	.	.	PUNCT
cana-4362	150	1	[	[	X
cana-4362	150	2	17	17	NUM
cana-4362	150	3	]	]	PUNCT
cana-4362	150	4	explores	explore	VERB
cana-4362	150	5	a	a	DET
cana-4362	150	6	hybrid	hybrid	ADJ
cana-4362	150	7	feature	feature	NOUN
cana-4362	150	8	selection	selection	NOUN
cana-4362	150	9	method	method	NOUN
cana-4362	150	10	combining	combine	VERB
cana-4362	150	11	deep	deep	ADJ
cana-4362	150	12	learning	learning	NOUN
cana-4362	150	13	with	with	ADP
cana-4362	150	14	transfer	transfer	NOUN
cana-4362	150	15	learning	learning	NOUN
cana-4362	150	16	for	for	ADP
cana-4362	150	17	cross	cross	ADJ
cana-4362	150	18	-	-	ADJ
cana-4362	150	19	domain	domain	ADJ
cana-4362	150	20	sentiment	sentiment	NOUN
cana-4362	150	21	classification	classification	NOUN
cana-4362	150	22	.	.	PUNCT
cana-4362	151	1	the	the	DET
cana-4362	151	2	authors	author	NOUN
cana-4362	151	3	propose	propose	VERB
cana-4362	151	4	a	a	DET
cana-4362	151	5	model	model	NOUN
cana-4362	151	6	that	that	PRON
cana-4362	151	7	transfers	transfer	VERB
cana-4362	151	8	knowledge	knowledge	NOUN
cana-4362	151	9	from	from	ADP
cana-4362	151	10	one	one	NUM
cana-4362	151	11	domain	domain	NOUN
cana-4362	151	12	to	to	ADP
cana-4362	151	13	another	another	PRON
cana-4362	151	14	,	,	PUNCT
cana-4362	151	15	such	such	ADJ
cana-4362	151	16	as	as	ADP
cana-4362	151	17	from	from	ADP
cana-4362	151	18	movie	movie	NOUN
cana-4362	151	19	reviews	review	NOUN
cana-4362	151	20	to	to	ADP
cana-4362	151	21	social	social	ADJ
cana-4362	151	22	media	medium	NOUN
cana-4362	151	23	sentiment	sentiment	NOUN
cana-4362	151	24	.	.	PUNCT
cana-4362	152	1	they	they	PRON
cana-4362	152	2	apply	apply	VERB
cana-4362	152	3	a	a	DET
cana-4362	152	4	hybrid	hybrid	ADJ
cana-4362	152	5	feature	feature	NOUN
cana-4362	152	6	selection	selection	NOUN
cana-4362	152	7	process	process	NOUN
cana-4362	152	8	that	that	PRON
cana-4362	152	9	incorporates	incorporate	VERB
cana-4362	152	10	both	both	DET
cana-4362	152	11	statistical	statistical	ADJ
cana-4362	152	12	methods	method	NOUN
cana-4362	152	13	and	and	CCONJ
cana-4362	152	14	deep	deep	ADJ
cana-4362	152	15	learning	learning	NOUN
cana-4362	152	16	techniques	technique	NOUN
cana-4362	152	17	like	like	ADP
cana-4362	152	18	cnns	cnn	NOUN
cana-4362	152	19	to	to	PART
cana-4362	152	20	optimize	optimize	VERB
cana-4362	152	21	the	the	DET
cana-4362	152	22	feature	feature	NOUN
cana-4362	152	23	set	set	VERB
cana-4362	152	24	.	.	PUNCT
cana-4362	153	1	the	the	DET
cana-4362	153	2	study	study	NOUN
cana-4362	153	3	demonstrates	demonstrate	VERB
cana-4362	153	4	that	that	SCONJ
cana-4362	153	5	cross	cross	ADJ
cana-4362	153	6	-	-	ADJ
cana-4362	153	7	domain	domain	ADJ
cana-4362	153	8	sentiment	sentiment	NOUN
cana-4362	153	9	classification	classification	NOUN
cana-4362	153	10	can	can	AUX
cana-4362	153	11	be	be	AUX
cana-4362	153	12	significantly	significantly	ADV
cana-4362	153	13	improved	improve	VERB
cana-4362	153	14	with	with	ADP
cana-4362	153	15	hybrid	hybrid	ADJ
cana-4362	153	16	models	model	NOUN
cana-4362	153	17	,	,	PUNCT
cana-4362	153	18	achieving	achieve	VERB
cana-4362	153	19	higher	high	ADJ
cana-4362	153	20	accuracy	accuracy	NOUN
cana-4362	153	21	and	and	CCONJ
cana-4362	153	22	recall	recall	NOUN
cana-4362	153	23	in	in	ADP
cana-4362	153	24	experimental	experimental	ADJ
cana-4362	153	25	tests	test	NOUN
cana-4362	153	26	.	.	PUNCT
cana-4362	154	1	the	the	DET
cana-4362	154	2	approach	approach	NOUN
cana-4362	154	3	is	be	AUX
cana-4362	154	4	particularly	particularly	ADV
cana-4362	154	5	effective	effective	ADJ
cana-4362	154	6	for	for	ADP
cana-4362	154	7	social	social	ADJ
cana-4362	154	8	media	medium	NOUN
cana-4362	154	9	sentiment	sentiment	NOUN
cana-4362	154	10	analysis	analysis	NOUN
cana-4362	154	11	,	,	PUNCT
cana-4362	154	12	where	where	SCONJ
cana-4362	154	13	domain	domain	NOUN
cana-4362	154	14	shifts	shift	NOUN
cana-4362	154	15	and	and	CCONJ
cana-4362	154	16	noisy	noisy	ADJ
cana-4362	154	17	data	datum	NOUN
cana-4362	154	18	present	present	ADJ
cana-4362	154	19	challenges	challenge	NOUN
cana-4362	154	20	.	.	PUNCT
cana-4362	155	1	lin	lin	PROPN
cana-4362	155	2	et	et	PROPN
cana-4362	155	3	al	al	PROPN
cana-4362	155	4	.	.	PUNCT
cana-4362	156	1	[	[	X
cana-4362	156	2	18	18	NUM
cana-4362	156	3	]	]	PUNCT
cana-4362	156	4	discusses	discuss	VERB
cana-4362	156	5	hybrid	hybrid	ADJ
cana-4362	156	6	feature	feature	NOUN
cana-4362	156	7	engineering	engineering	NOUN
cana-4362	156	8	for	for	ADP
cana-4362	156	9	sentiment	sentiment	NOUN
cana-4362	156	10	analysis	analysis	NOUN
cana-4362	156	11	on	on	ADP
cana-4362	156	12	short	short	ADJ
cana-4362	156	13	social	social	ADJ
cana-4362	156	14	media	medium	NOUN
cana-4362	156	15	texts	text	NOUN
cana-4362	156	16	,	,	PUNCT
cana-4362	156	17	focusing	focus	VERB
cana-4362	156	18	on	on	ADP
cana-4362	156	19	enhancing	enhance	VERB
cana-4362	156	20	deep	deep	ADJ
cana-4362	156	21	learning	learning	NOUN
cana-4362	156	22	models	model	NOUN
cana-4362	156	23	.	.	PUNCT
cana-4362	157	1	the	the	DET
cana-4362	157	2	authors	author	NOUN
cana-4362	157	3	propose	propose	VERB
cana-4362	157	4	using	use	VERB
cana-4362	157	5	multiple	multiple	ADJ
cana-4362	157	6	feature	feature	NOUN
cana-4362	157	7	extraction	extraction	NOUN
cana-4362	157	8	techniques	technique	NOUN
cana-4362	157	9	like	like	ADP
cana-4362	157	10	word2vec	word2vec	X
cana-4362	157	11	and	and	CCONJ
cana-4362	157	12	tf	tf	PROPN
cana-4362	157	13	-	-	PUNCT
cana-4362	157	14	idf	idf	PROPN
cana-4362	157	15	in	in	ADP
cana-4362	157	16	combination	combination	NOUN
cana-4362	157	17	with	with	ADP
cana-4362	157	18	feature	feature	NOUN
cana-4362	157	19	selection	selection	NOUN
cana-4362	157	20	algorithms	algorithm	NOUN
cana-4362	157	21	such	such	ADJ
cana-4362	157	22	as	as	ADP
cana-4362	157	23	information	information	NOUN
cana-4362	157	24	gain	gain	NOUN
cana-4362	157	25	.	.	PUNCT
cana-4362	158	1	the	the	DET
cana-4362	158	2	hybrid	hybrid	NOUN
cana-4362	158	3	model	model	NOUN
cana-4362	158	4	incorporates	incorporate	VERB
cana-4362	158	5	these	these	DET
cana-4362	158	6	techniques	technique	NOUN
cana-4362	158	7	with	with	ADP
cana-4362	158	8	cnns	cnn	NOUN
cana-4362	158	9	and	and	CCONJ
cana-4362	158	10	lstms	lstms	NOUN
cana-4362	158	11	for	for	SCONJ
cana-4362	158	12	communications	communication	NOUN
cana-4362	158	13	on	on	ADP
cana-4362	158	14	applied	apply	VERB
cana-4362	158	15	nonlinear	nonlinear	ADJ
cana-4362	158	16	analysis	analysis	NOUN
cana-4362	158	17	issn	issn	NOUN
cana-4362	158	18	:	:	PUNCT
cana-4362	158	19	1074	1074	NUM
cana-4362	158	20	-	-	PUNCT
cana-4362	158	21	133x	133x	NUM
cana-4362	158	22	vol	vol	NOUN
cana-4362	158	23	32	32	NUM
cana-4362	158	24	no	no	NOUN
cana-4362	158	25	.	.	PUNCT
cana-4362	159	1	9s	9s	NUM
cana-4362	159	2	(	(	PUNCT
cana-4362	159	3	2025	2025	NUM
cana-4362	159	4	)	)	PUNCT
cana-4362	159	5	1905	1905	NUM
cana-4362	160	1	https://internationalpubls.com	https://internationalpubls.com	X
cana-4362	160	2	sentiment	sentiment	NOUN
cana-4362	160	3	classification	classification	NOUN
cana-4362	160	4	.	.	PUNCT
cana-4362	161	1	their	their	PRON
cana-4362	161	2	results	result	NOUN
cana-4362	161	3	show	show	VERB
cana-4362	161	4	that	that	SCONJ
cana-4362	161	5	the	the	DET
cana-4362	161	6	hybrid	hybrid	ADJ
cana-4362	161	7	feature	feature	NOUN
cana-4362	161	8	engineering	engineering	NOUN
cana-4362	161	9	approach	approach	NOUN
cana-4362	161	10	improves	improve	VERB
cana-4362	161	11	classification	classification	NOUN
cana-4362	161	12	accuracy	accuracy	NOUN
cana-4362	161	13	,	,	PUNCT
cana-4362	161	14	particularly	particularly	ADV
cana-4362	161	15	in	in	ADP
cana-4362	161	16	noisy	noisy	ADJ
cana-4362	161	17	and	and	CCONJ
cana-4362	161	18	unstructured	unstructured	ADJ
cana-4362	161	19	data	datum	NOUN
cana-4362	161	20	environments	environment	NOUN
cana-4362	161	21	like	like	ADP
cana-4362	161	22	twitter	twitter	NOUN
cana-4362	161	23	.	.	PUNCT
cana-4362	162	1	the	the	DET
cana-4362	162	2	study	study	NOUN
cana-4362	162	3	suggests	suggest	VERB
cana-4362	162	4	that	that	SCONJ
cana-4362	162	5	this	this	DET
cana-4362	162	6	model	model	NOUN
cana-4362	162	7	is	be	AUX
cana-4362	162	8	well	well	ADV
cana-4362	162	9	-	-	PUNCT
cana-4362	162	10	suited	suit	VERB
cana-4362	162	11	for	for	ADP
cana-4362	162	12	real	real	ADJ
cana-4362	162	13	-	-	PUNCT
cana-4362	162	14	time	time	NOUN
cana-4362	162	15	sentiment	sentiment	NOUN
cana-4362	162	16	analysis	analysis	NOUN
cana-4362	162	17	applications	application	NOUN
cana-4362	162	18	,	,	PUNCT
cana-4362	162	19	providing	provide	VERB
cana-4362	162	20	both	both	DET
cana-4362	162	21	efficiency	efficiency	NOUN
cana-4362	162	22	and	and	CCONJ
cana-4362	162	23	scalability	scalability	NOUN
cana-4362	162	24	.	.	PUNCT
cana-4362	163	1	xie	xie	PROPN
cana-4362	163	2	et	et	PROPN
cana-4362	163	3	al	al	PROPN
cana-4362	163	4	.	.	PUNCT
cana-4362	164	1	[	[	X
cana-4362	164	2	19	19	NUM
cana-4362	164	3	]	]	PUNCT
cana-4362	164	4	introduces	introduce	VERB
cana-4362	164	5	a	a	DET
cana-4362	164	6	hybrid	hybrid	ADJ
cana-4362	164	7	feature	feature	NOUN
cana-4362	164	8	selection	selection	NOUN
cana-4362	164	9	and	and	CCONJ
cana-4362	164	10	attention	attention	NOUN
cana-4362	164	11	mechanism	mechanism	NOUN
cana-4362	164	12	model	model	NOUN
cana-4362	164	13	to	to	PART
cana-4362	164	14	enhance	enhance	VERB
cana-4362	164	15	sentiment	sentiment	NOUN
cana-4362	164	16	classification	classification	NOUN
cana-4362	164	17	of	of	ADP
cana-4362	164	18	twitter	twitter	NOUN
cana-4362	164	19	data	datum	NOUN
cana-4362	164	20	.	.	PUNCT
cana-4362	165	1	the	the	DET
cana-4362	165	2	authors	author	NOUN
cana-4362	165	3	combine	combine	VERB
cana-4362	165	4	traditional	traditional	ADJ
cana-4362	165	5	feature	feature	NOUN
cana-4362	165	6	selection	selection	NOUN
cana-4362	165	7	techniques	technique	NOUN
cana-4362	165	8	,	,	PUNCT
cana-4362	165	9	such	such	ADJ
cana-4362	165	10	as	as	ADP
cana-4362	165	11	mutual	mutual	ADJ
cana-4362	165	12	information	information	NOUN
cana-4362	165	13	,	,	PUNCT
cana-4362	165	14	with	with	SCONJ
cana-4362	165	15	attention	attention	NOUN
cana-4362	165	16	-	-	PUNCT
cana-4362	165	17	based	base	VERB
cana-4362	165	18	deep	deep	ADJ
cana-4362	165	19	learning	learning	NOUN
cana-4362	165	20	models	model	NOUN
cana-4362	165	21	to	to	PART
cana-4362	165	22	improve	improve	VERB
cana-4362	165	23	accuracy	accuracy	NOUN
cana-4362	165	24	.	.	PUNCT
cana-4362	166	1	the	the	DET
cana-4362	166	2	hybrid	hybrid	NOUN
cana-4362	166	3	model	model	NOUN
cana-4362	166	4	is	be	AUX
cana-4362	166	5	designed	design	VERB
cana-4362	166	6	to	to	PART
cana-4362	166	7	prioritize	prioritize	VERB
cana-4362	166	8	important	important	ADJ
cana-4362	166	9	features	feature	NOUN
cana-4362	166	10	while	while	SCONJ
cana-4362	166	11	reducing	reduce	VERB
cana-4362	166	12	the	the	DET
cana-4362	166	13	dimensionality	dimensionality	NOUN
cana-4362	166	14	of	of	ADP
cana-4362	166	15	the	the	DET
cana-4362	166	16	dataset	dataset	NOUN
cana-4362	166	17	,	,	PUNCT
cana-4362	166	18	enabling	enable	VERB
cana-4362	166	19	more	more	ADV
cana-4362	166	20	efficient	efficient	ADJ
cana-4362	166	21	processing	processing	NOUN
cana-4362	166	22	of	of	ADP
cana-4362	166	23	large	large	ADJ
cana-4362	166	24	social	social	ADJ
cana-4362	166	25	media	medium	NOUN
cana-4362	166	26	datasets	dataset	NOUN
cana-4362	166	27	.	.	PUNCT
cana-4362	167	1	their	their	PRON
cana-4362	167	2	results	result	NOUN
cana-4362	167	3	show	show	VERB
cana-4362	167	4	significant	significant	ADJ
cana-4362	167	5	improvements	improvement	NOUN
cana-4362	167	6	in	in	ADP
cana-4362	167	7	accuracy	accuracy	NOUN
cana-4362	167	8	and	and	CCONJ
cana-4362	167	9	f1	f1	NOUN
cana-4362	167	10	score	score	NOUN
cana-4362	167	11	,	,	PUNCT
cana-4362	167	12	particularly	particularly	ADV
cana-4362	167	13	when	when	SCONJ
cana-4362	167	14	compared	compare	VERB
cana-4362	167	15	to	to	ADP
cana-4362	167	16	traditional	traditional	ADJ
cana-4362	167	17	deep	deep	ADJ
cana-4362	167	18	learning	learning	NOUN
cana-4362	167	19	models	model	NOUN
cana-4362	167	20	without	without	ADP
cana-4362	167	21	feature	feature	NOUN
cana-4362	167	22	selection	selection	NOUN
cana-4362	167	23	.	.	PUNCT
cana-4362	168	1	the	the	DET
cana-4362	168	2	study	study	NOUN
cana-4362	168	3	concludes	conclude	VERB
cana-4362	168	4	that	that	SCONJ
cana-4362	168	5	integrating	integrate	VERB
cana-4362	168	6	attention	attention	NOUN
cana-4362	168	7	mechanisms	mechanism	NOUN
cana-4362	168	8	with	with	ADP
cana-4362	168	9	hybrid	hybrid	ADJ
cana-4362	168	10	feature	feature	NOUN
cana-4362	168	11	selection	selection	NOUN
cana-4362	168	12	can	can	AUX
cana-4362	168	13	greatly	greatly	ADV
cana-4362	168	14	improve	improve	VERB
cana-4362	168	15	sentiment	sentiment	NOUN
cana-4362	168	16	classification	classification	NOUN
cana-4362	168	17	performance	performance	NOUN
cana-4362	168	18	.	.	PUNCT
cana-4362	169	1	peng	peng	PROPN
cana-4362	169	2	et	et	PROPN
cana-4362	169	3	al	al	PROPN
cana-4362	169	4	.	.	PUNCT
cana-4362	170	1	[	[	X
cana-4362	170	2	20	20	NUM
cana-4362	170	3	]	]	PUNCT
cana-4362	170	4	investigates	investigate	VERB
cana-4362	170	5	the	the	DET
cana-4362	170	6	use	use	NOUN
cana-4362	170	7	of	of	ADP
cana-4362	170	8	hybrid	hybrid	ADJ
cana-4362	170	9	deep	deep	ADJ
cana-4362	170	10	learning	learning	NOUN
cana-4362	170	11	models	model	NOUN
cana-4362	170	12	for	for	ADP
cana-4362	170	13	sentiment	sentiment	NOUN
cana-4362	170	14	classification	classification	NOUN
cana-4362	170	15	on	on	ADP
cana-4362	170	16	social	social	ADJ
cana-4362	170	17	media	medium	NOUN
cana-4362	170	18	data	datum	NOUN
cana-4362	170	19	,	,	PUNCT
cana-4362	170	20	integrating	integrate	VERB
cana-4362	170	21	feature	feature	NOUN
cana-4362	170	22	selection	selection	NOUN
cana-4362	170	23	to	to	PART
cana-4362	170	24	optimize	optimize	VERB
cana-4362	170	25	performance	performance	NOUN
cana-4362	170	26	.	.	PUNCT
cana-4362	171	1	the	the	DET
cana-4362	171	2	authors	author	NOUN
cana-4362	171	3	combine	combine	VERB
cana-4362	171	4	cnn	cnn	PROPN
cana-4362	171	5	and	and	CCONJ
cana-4362	171	6	lstm	lstm	NOUN
cana-4362	171	7	models	model	NOUN
cana-4362	171	8	with	with	ADP
cana-4362	171	9	a	a	DET
cana-4362	171	10	hybrid	hybrid	ADJ
cana-4362	171	11	feature	feature	NOUN
cana-4362	171	12	selection	selection	NOUN
cana-4362	171	13	approach	approach	NOUN
cana-4362	171	14	that	that	PRON
cana-4362	171	15	uses	use	VERB
cana-4362	171	16	a	a	DET
cana-4362	171	17	combination	combination	NOUN
cana-4362	171	18	of	of	ADP
cana-4362	171	19	statistical	statistical	ADJ
cana-4362	171	20	and	and	CCONJ
cana-4362	171	21	machine	machine	NOUN
cana-4362	171	22	learning	learning	NOUN
cana-4362	171	23	-	-	PUNCT
cana-4362	171	24	based	base	VERB
cana-4362	171	25	techniques	technique	NOUN
cana-4362	171	26	.	.	PUNCT
cana-4362	172	1	their	their	PRON
cana-4362	172	2	experiments	experiment	NOUN
cana-4362	172	3	on	on	ADP
cana-4362	172	4	social	social	ADJ
cana-4362	172	5	media	medium	NOUN
cana-4362	172	6	datasets	dataset	NOUN
cana-4362	172	7	show	show	VERB
cana-4362	172	8	improved	improve	VERB
cana-4362	172	9	classification	classification	NOUN
cana-4362	172	10	accuracy	accuracy	NOUN
cana-4362	172	11	and	and	CCONJ
cana-4362	172	12	efficiency	efficiency	NOUN
cana-4362	172	13	,	,	PUNCT
cana-4362	172	14	with	with	ADP
cana-4362	172	15	the	the	DET
cana-4362	172	16	hybrid	hybrid	ADJ
cana-4362	172	17	model	model	NOUN
cana-4362	172	18	outperforming	outperform	VERB
cana-4362	172	19	traditional	traditional	ADJ
cana-4362	172	20	methods	method	NOUN
cana-4362	172	21	.	.	PUNCT
cana-4362	173	1	the	the	DET
cana-4362	173	2	study	study	NOUN
cana-4362	173	3	emphasizes	emphasize	VERB
cana-4362	173	4	the	the	DET
cana-4362	173	5	importance	importance	NOUN
cana-4362	173	6	of	of	ADP
cana-4362	173	7	feature	feature	NOUN
cana-4362	173	8	selection	selection	NOUN
cana-4362	173	9	in	in	ADP
cana-4362	173	10	reducing	reduce	VERB
cana-4362	173	11	the	the	DET
cana-4362	173	12	complexity	complexity	NOUN
cana-4362	173	13	of	of	ADP
cana-4362	173	14	deep	deep	ADJ
cana-4362	173	15	learning	learning	NOUN
cana-4362	173	16	models	model	NOUN
cana-4362	173	17	while	while	SCONJ
cana-4362	173	18	maintaining	maintain	VERB
cana-4362	173	19	high	high	ADJ
cana-4362	173	20	classification	classification	NOUN
cana-4362	173	21	performance	performance	NOUN
cana-4362	173	22	,	,	PUNCT
cana-4362	173	23	particularly	particularly	ADV
cana-4362	173	24	for	for	ADP
cana-4362	173	25	large	large	ADJ
cana-4362	173	26	-	-	PUNCT
cana-4362	173	27	scale	scale	NOUN
cana-4362	173	28	social	social	ADJ
cana-4362	173	29	media	medium	NOUN
cana-4362	173	30	datasets	dataset	NOUN
cana-4362	173	31	.	.	PUNCT
cana-4362	174	1	liu	liu	PROPN
cana-4362	174	2	et	et	PROPN
cana-4362	174	3	al	al	PROPN
cana-4362	174	4	.	.	PUNCT
cana-4362	175	1	[	[	X
cana-4362	175	2	21	21	NUM
cana-4362	175	3	]	]	PUNCT
cana-4362	175	4	focuses	focus	VERB
cana-4362	175	5	on	on	ADP
cana-4362	175	6	hybrid	hybrid	ADJ
cana-4362	175	7	feature	feature	NOUN
cana-4362	175	8	selection	selection	NOUN
cana-4362	175	9	for	for	ADP
cana-4362	175	10	sentiment	sentiment	NOUN
cana-4362	175	11	classification	classification	NOUN
cana-4362	175	12	using	use	VERB
cana-4362	175	13	convolutional	convolutional	ADJ
cana-4362	175	14	neural	neural	ADJ
cana-4362	175	15	networks	network	NOUN
cana-4362	175	16	(	(	PUNCT
cana-4362	175	17	cnns	cnns	PROPN
cana-4362	175	18	)	)	PUNCT
cana-4362	175	19	and	and	CCONJ
cana-4362	175	20	a	a	DET
cana-4362	175	21	recursive	recursive	ADJ
cana-4362	175	22	feature	feature	NOUN
cana-4362	175	23	elimination	elimination	NOUN
cana-4362	175	24	(	(	PUNCT
cana-4362	175	25	rfe	rfe	NOUN
cana-4362	175	26	)	)	PUNCT
cana-4362	175	27	method	method	NOUN
cana-4362	175	28	.	.	PUNCT
cana-4362	176	1	their	their	PRON
cana-4362	176	2	study	study	NOUN
cana-4362	176	3	investigates	investigate	VERB
cana-4362	176	4	the	the	DET
cana-4362	176	5	effectiveness	effectiveness	NOUN
cana-4362	176	6	of	of	ADP
cana-4362	176	7	combining	combine	VERB
cana-4362	176	8	feature	feature	NOUN
cana-4362	176	9	selection	selection	NOUN
cana-4362	176	10	algorithms	algorithm	NOUN
cana-4362	176	11	with	with	ADP
cana-4362	176	12	deep	deep	ADJ
cana-4362	176	13	learning	learning	NOUN
cana-4362	176	14	for	for	ADP
cana-4362	176	15	improving	improve	VERB
cana-4362	176	16	sentiment	sentiment	NOUN
cana-4362	176	17	classification	classification	NOUN
cana-4362	176	18	performance	performance	NOUN
cana-4362	176	19	on	on	ADP
cana-4362	176	20	social	social	ADJ
cana-4362	176	21	media	medium	NOUN
cana-4362	176	22	datasets	dataset	NOUN
cana-4362	176	23	.	.	PUNCT
cana-4362	177	1	tested	test	VERB
cana-4362	177	2	on	on	ADP
cana-4362	177	3	data	datum	NOUN
cana-4362	177	4	from	from	ADP
cana-4362	177	5	twitter	twitter	NOUN
cana-4362	177	6	,	,	PUNCT
cana-4362	177	7	the	the	DET
cana-4362	177	8	model	model	NOUN
cana-4362	177	9	is	be	AUX
cana-4362	177	10	able	able	ADJ
cana-4362	177	11	to	to	PART
cana-4362	177	12	achieve	achieve	VERB
cana-4362	177	13	higher	high	ADJ
cana-4362	177	14	accuracy	accuracy	NOUN
cana-4362	177	15	and	and	CCONJ
cana-4362	177	16	faster	fast	ADJ
cana-4362	177	17	convergence	convergence	NOUN
cana-4362	177	18	times	time	NOUN
cana-4362	177	19	compared	compare	VERB
cana-4362	177	20	to	to	ADP
cana-4362	177	21	traditional	traditional	ADJ
cana-4362	177	22	deep	deep	ADJ
cana-4362	177	23	learning	learning	NOUN
cana-4362	177	24	methods	method	NOUN
cana-4362	177	25	.	.	PUNCT
cana-4362	178	1	the	the	DET
cana-4362	178	2	paper	paper	NOUN
cana-4362	178	3	highlights	highlight	NOUN
cana-4362	178	4	that	that	SCONJ
cana-4362	178	5	hybrid	hybrid	ADJ
cana-4362	178	6	feature	feature	NOUN
cana-4362	178	7	selection	selection	NOUN
cana-4362	178	8	helps	help	VERB
cana-4362	178	9	mitigate	mitigate	VERB
cana-4362	178	10	the	the	DET
cana-4362	178	11	problem	problem	NOUN
cana-4362	178	12	of	of	ADP
cana-4362	178	13	overfitting	overfitte	VERB
cana-4362	178	14	and	and	CCONJ
cana-4362	178	15	reduces	reduce	VERB
cana-4362	178	16	the	the	DET
cana-4362	178	17	dimensionality	dimensionality	NOUN
cana-4362	178	18	of	of	ADP
cana-4362	178	19	the	the	DET
cana-4362	178	20	data	datum	NOUN
cana-4362	178	21	,	,	PUNCT
cana-4362	178	22	which	which	PRON
cana-4362	178	23	is	be	AUX
cana-4362	178	24	crucial	crucial	ADJ
cana-4362	178	25	for	for	ADP
cana-4362	178	26	processing	process	VERB
cana-4362	178	27	large	large	ADJ
cana-4362	178	28	social	social	ADJ
cana-4362	178	29	media	medium	NOUN
cana-4362	178	30	datasets	dataset	NOUN
cana-4362	178	31	where	where	SCONJ
cana-4362	178	32	noise	noise	NOUN
cana-4362	178	33	and	and	CCONJ
cana-4362	178	34	irrelevant	irrelevant	ADJ
cana-4362	178	35	features	feature	NOUN
cana-4362	178	36	often	often	ADV
cana-4362	178	37	dominate	dominate	VERB
cana-4362	178	38	.	.	PUNCT
cana-4362	179	1	zhao	zhao	PROPN
cana-4362	179	2	et	et	PROPN
cana-4362	179	3	al	al	PROPN
cana-4362	179	4	.	.	PUNCT
cana-4362	180	1	[	[	X
cana-4362	180	2	22	22	NUM
cana-4362	180	3	]	]	PUNCT
cana-4362	180	4	proposes	propose	VERB
cana-4362	180	5	a	a	DET
cana-4362	180	6	hybrid	hybrid	ADJ
cana-4362	180	7	deep	deep	ADJ
cana-4362	180	8	learning	learning	NOUN
cana-4362	180	9	model	model	NOUN
cana-4362	180	10	for	for	ADP
cana-4362	180	11	sentiment	sentiment	NOUN
cana-4362	180	12	classification	classification	NOUN
cana-4362	180	13	by	by	ADP
cana-4362	180	14	combining	combine	VERB
cana-4362	180	15	traditional	traditional	ADJ
cana-4362	180	16	feature	feature	NOUN
cana-4362	180	17	selection	selection	NOUN
cana-4362	180	18	techniques	technique	NOUN
cana-4362	180	19	with	with	ADP
cana-4362	180	20	attention	attention	NOUN
cana-4362	180	21	-	-	PUNCT
cana-4362	180	22	based	base	VERB
cana-4362	180	23	mechanisms	mechanism	NOUN
cana-4362	180	24	.	.	PUNCT
cana-4362	181	1	the	the	DET
cana-4362	181	2	model	model	NOUN
cana-4362	181	3	uses	use	VERB
cana-4362	181	4	a	a	DET
cana-4362	181	5	hybrid	hybrid	ADJ
cana-4362	181	6	approach	approach	NOUN
cana-4362	181	7	involving	involve	VERB
cana-4362	181	8	both	both	CCONJ
cana-4362	181	9	statistical	statistical	ADJ
cana-4362	181	10	feature	feature	NOUN
cana-4362	181	11	selection	selection	NOUN
cana-4362	181	12	methods	method	NOUN
cana-4362	181	13	,	,	PUNCT
cana-4362	181	14	such	such	ADJ
cana-4362	181	15	as	as	ADP
cana-4362	181	16	chi	chi	ADJ
cana-4362	181	17	-	-	PUNCT
cana-4362	181	18	square	square	ADJ
cana-4362	181	19	and	and	CCONJ
cana-4362	181	20	information	information	NOUN
cana-4362	181	21	gain	gain	NOUN
cana-4362	181	22	,	,	PUNCT
cana-4362	181	23	followed	follow	VERB
cana-4362	181	24	by	by	ADP
cana-4362	181	25	a	a	DET
cana-4362	181	26	cnn	cnn	PROPN
cana-4362	181	27	-	-	PUNCT
cana-4362	181	28	lstm	lstm	ADJ
cana-4362	181	29	structure	structure	NOUN
cana-4362	181	30	to	to	PART
cana-4362	181	31	perform	perform	VERB
cana-4362	181	32	classification	classification	NOUN
cana-4362	181	33	.	.	PUNCT
cana-4362	182	1	the	the	DET
cana-4362	182	2	attention	attention	NOUN
cana-4362	182	3	mechanism	mechanism	NOUN
cana-4362	182	4	helps	help	VERB
cana-4362	182	5	the	the	DET
cana-4362	182	6	model	model	NOUN
cana-4362	182	7	focus	focus	VERB
cana-4362	182	8	on	on	ADP
cana-4362	182	9	the	the	DET
cana-4362	182	10	most	most	ADV
cana-4362	182	11	relevant	relevant	ADJ
cana-4362	182	12	features	feature	NOUN
cana-4362	182	13	,	,	PUNCT
cana-4362	182	14	leading	lead	VERB
cana-4362	182	15	to	to	ADP
cana-4362	182	16	improved	improve	VERB
cana-4362	182	17	classification	classification	NOUN
cana-4362	182	18	accuracy	accuracy	NOUN
cana-4362	182	19	.	.	PUNCT
cana-4362	183	1	tested	test	VERB
cana-4362	183	2	on	on	ADP
cana-4362	183	3	twitter	twitter	NOUN
cana-4362	183	4	and	and	CCONJ
cana-4362	183	5	reddit	reddit	NOUN
cana-4362	183	6	datasets	dataset	NOUN
cana-4362	183	7	,	,	PUNCT
cana-4362	183	8	their	their	PRON
cana-4362	183	9	model	model	NOUN
cana-4362	183	10	outperforms	outperform	VERB
cana-4362	183	11	baseline	baseline	NOUN
cana-4362	183	12	models	model	NOUN
cana-4362	183	13	in	in	ADP
cana-4362	183	14	terms	term	NOUN
cana-4362	183	15	of	of	ADP
cana-4362	183	16	f1	f1	ADJ
cana-4362	183	17	score	score	NOUN
cana-4362	183	18	and	and	CCONJ
cana-4362	183	19	precision	precision	NOUN
cana-4362	183	20	.	.	PUNCT
cana-4362	184	1	the	the	DET
cana-4362	184	2	study	study	NOUN
cana-4362	184	3	shows	show	VERB
cana-4362	184	4	that	that	SCONJ
cana-4362	184	5	integrating	integrate	VERB
cana-4362	184	6	attention	attention	NOUN
cana-4362	184	7	mechanisms	mechanism	NOUN
cana-4362	184	8	with	with	ADP
cana-4362	184	9	hybrid	hybrid	ADJ
cana-4362	184	10	feature	feature	NOUN
cana-4362	184	11	selection	selection	NOUN
cana-4362	184	12	provides	provide	VERB
cana-4362	184	13	a	a	DET
cana-4362	184	14	robust	robust	ADJ
cana-4362	184	15	approach	approach	NOUN
cana-4362	184	16	to	to	ADP
cana-4362	184	17	handling	handle	VERB
cana-4362	184	18	large	large	ADJ
cana-4362	184	19	and	and	CCONJ
cana-4362	184	20	noisy	noisy	ADJ
cana-4362	184	21	social	social	ADJ
cana-4362	184	22	media	medium	NOUN
cana-4362	184	23	datasets	dataset	NOUN
cana-4362	184	24	.	.	PUNCT
cana-4362	185	1	yang	yang	PROPN
cana-4362	185	2	et	et	PROPN
cana-4362	185	3	al	al	PROPN
cana-4362	185	4	.	.	PUNCT
cana-4362	186	1	[	[	X
cana-4362	186	2	23	23	NUM
cana-4362	186	3	]	]	PUNCT
cana-4362	186	4	explores	explore	VERB
cana-4362	186	5	hybrid	hybrid	ADJ
cana-4362	186	6	feature	feature	NOUN
cana-4362	186	7	selection	selection	NOUN
cana-4362	186	8	for	for	ADP
cana-4362	186	9	sentiment	sentiment	NOUN
cana-4362	186	10	classification	classification	NOUN
cana-4362	186	11	using	use	VERB
cana-4362	186	12	a	a	DET
cana-4362	186	13	combination	combination	NOUN
cana-4362	186	14	of	of	ADP
cana-4362	186	15	traditional	traditional	ADJ
cana-4362	186	16	machine	machine	NOUN
cana-4362	186	17	learning	learn	VERB
cana-4362	186	18	techniques	technique	NOUN
cana-4362	186	19	and	and	CCONJ
cana-4362	186	20	deep	deep	ADJ
cana-4362	186	21	learning	learning	NOUN
cana-4362	186	22	models	model	NOUN
cana-4362	186	23	.	.	PUNCT
cana-4362	187	1	their	their	PRON
cana-4362	187	2	proposed	propose	VERB
cana-4362	187	3	model	model	NOUN
cana-4362	187	4	integrates	integrate	NOUN
cana-4362	187	5	mutual	mutual	ADJ
cana-4362	187	6	information	information	NOUN
cana-4362	187	7	(	(	PUNCT
cana-4362	187	8	mi	mi	NOUN
cana-4362	187	9	)	)	PUNCT
cana-4362	187	10	and	and	CCONJ
cana-4362	187	11	relieff	relieff	NOUN
cana-4362	187	12	for	for	ADP
cana-4362	187	13	feature	feature	NOUN
cana-4362	187	14	selection	selection	NOUN
cana-4362	187	15	with	with	ADP
cana-4362	187	16	a	a	DET
cana-4362	187	17	cnn	cnn	PROPN
cana-4362	187	18	-	-	PUNCT
cana-4362	187	19	lstm	lstm	ADJ
cana-4362	187	20	framework	framework	NOUN
cana-4362	187	21	for	for	ADP
cana-4362	187	22	classification	classification	NOUN
cana-4362	187	23	.	.	PUNCT
cana-4362	188	1	the	the	DET
cana-4362	188	2	hybrid	hybrid	ADJ
cana-4362	188	3	approach	approach	NOUN
cana-4362	188	4	is	be	AUX
cana-4362	188	5	designed	design	VERB
cana-4362	188	6	to	to	PART
cana-4362	188	7	reduce	reduce	VERB
cana-4362	188	8	the	the	DET
cana-4362	188	9	computational	computational	ADJ
cana-4362	188	10	burden	burden	NOUN
cana-4362	188	11	of	of	ADP
cana-4362	188	12	deep	deep	ADJ
cana-4362	188	13	learning	learning	NOUN
cana-4362	188	14	communications	communication	NOUN
cana-4362	188	15	on	on	ADP
cana-4362	188	16	applied	apply	VERB
cana-4362	188	17	nonlinear	nonlinear	ADJ
cana-4362	188	18	analysis	analysis	NOUN
cana-4362	188	19	issn	issn	NOUN
cana-4362	188	20	:	:	PUNCT
cana-4362	188	21	1074	1074	NUM
cana-4362	188	22	-	-	PUNCT
cana-4362	188	23	133x	133x	NUM
cana-4362	188	24	vol	vol	NOUN
cana-4362	188	25	32	32	NUM
cana-4362	188	26	no	no	NOUN
cana-4362	188	27	.	.	PUNCT
cana-4362	189	1	9s	9s	NUM
cana-4362	189	2	(	(	PUNCT
cana-4362	189	3	2025	2025	NUM
cana-4362	189	4	)	)	PUNCT
cana-4362	189	5	1906	1906	NUM
cana-4362	189	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-4362	189	7	models	model	NOUN
cana-4362	189	8	while	while	SCONJ
cana-4362	189	9	improving	improve	VERB
cana-4362	189	10	classification	classification	NOUN
cana-4362	189	11	accuracy	accuracy	NOUN
cana-4362	189	12	on	on	ADP
cana-4362	189	13	social	social	ADJ
cana-4362	189	14	media	medium	NOUN
cana-4362	189	15	datasets	dataset	NOUN
cana-4362	189	16	,	,	PUNCT
cana-4362	189	17	particularly	particularly	ADV
cana-4362	189	18	twitter	twitter	NOUN
cana-4362	189	19	and	and	CCONJ
cana-4362	189	20	instagram	instagram	PROPN
cana-4362	189	21	.	.	PUNCT
cana-4362	190	1	their	their	PRON
cana-4362	190	2	experimental	experimental	ADJ
cana-4362	190	3	results	result	NOUN
cana-4362	190	4	show	show	VERB
cana-4362	190	5	that	that	SCONJ
cana-4362	190	6	the	the	DET
cana-4362	190	7	hybrid	hybrid	ADJ
cana-4362	190	8	feature	feature	NOUN
cana-4362	190	9	selection	selection	NOUN
cana-4362	190	10	method	method	NOUN
cana-4362	190	11	outperforms	outperform	VERB
cana-4362	190	12	conventional	conventional	ADJ
cana-4362	190	13	approaches	approach	NOUN
cana-4362	190	14	in	in	ADP
cana-4362	190	15	terms	term	NOUN
cana-4362	190	16	of	of	ADP
cana-4362	190	17	precision	precision	NOUN
cana-4362	190	18	and	and	CCONJ
cana-4362	190	19	recall	recall	NOUN
cana-4362	190	20	.	.	PUNCT
cana-4362	191	1	the	the	DET
cana-4362	191	2	study	study	NOUN
cana-4362	191	3	highlights	highlight	NOUN
cana-4362	191	4	that	that	SCONJ
cana-4362	191	5	combining	combine	VERB
cana-4362	191	6	traditional	traditional	ADJ
cana-4362	191	7	feature	feature	NOUN
cana-4362	191	8	selection	selection	NOUN
cana-4362	191	9	methods	method	NOUN
cana-4362	191	10	with	with	ADP
cana-4362	191	11	deep	deep	ADJ
cana-4362	191	12	learning	learning	NOUN
cana-4362	191	13	offers	offer	VERB
cana-4362	191	14	a	a	DET
cana-4362	191	15	scalable	scalable	ADJ
cana-4362	191	16	and	and	CCONJ
cana-4362	191	17	efficient	efficient	ADJ
cana-4362	191	18	solution	solution	NOUN
cana-4362	191	19	for	for	ADP
cana-4362	191	20	sentiment	sentiment	NOUN
cana-4362	191	21	classification	classification	NOUN
cana-4362	191	22	in	in	ADP
cana-4362	191	23	real	real	ADJ
cana-4362	191	24	-	-	PUNCT
cana-4362	191	25	world	world	NOUN
cana-4362	191	26	applications	application	NOUN
cana-4362	191	27	.	.	PUNCT
cana-4362	192	1	li	li	PROPN
cana-4362	192	2	et	et	PROPN
cana-4362	192	3	al	al	PROPN
cana-4362	192	4	.	.	PUNCT
cana-4362	193	1	[	[	X
cana-4362	193	2	24	24	NUM
cana-4362	193	3	]	]	PUNCT
cana-4362	193	4	investigates	investigate	VERB
cana-4362	193	5	hybrid	hybrid	ADJ
cana-4362	193	6	feature	feature	NOUN
cana-4362	193	7	selection	selection	NOUN
cana-4362	193	8	using	use	VERB
cana-4362	193	9	statistical	statistical	ADJ
cana-4362	193	10	methods	method	NOUN
cana-4362	193	11	and	and	CCONJ
cana-4362	193	12	deep	deep	ADJ
cana-4362	193	13	learning	learning	NOUN
cana-4362	193	14	for	for	ADP
cana-4362	193	15	sentiment	sentiment	NOUN
cana-4362	193	16	classification	classification	NOUN
cana-4362	193	17	of	of	ADP
cana-4362	193	18	social	social	ADJ
cana-4362	193	19	media	medium	NOUN
cana-4362	193	20	data	datum	NOUN
cana-4362	193	21	.	.	PUNCT
cana-4362	194	1	the	the	DET
cana-4362	194	2	authors	author	NOUN
cana-4362	194	3	apply	apply	VERB
cana-4362	194	4	a	a	DET
cana-4362	194	5	combination	combination	NOUN
cana-4362	194	6	of	of	ADP
cana-4362	194	7	filter	filter	NOUN
cana-4362	194	8	-	-	PUNCT
cana-4362	194	9	based	base	VERB
cana-4362	194	10	methods	method	NOUN
cana-4362	194	11	,	,	PUNCT
cana-4362	194	12	such	such	ADJ
cana-4362	194	13	as	as	ADP
cana-4362	194	14	chi	chi	ADJ
cana-4362	194	15	-	-	PUNCT
cana-4362	194	16	square	square	ADJ
cana-4362	194	17	and	and	CCONJ
cana-4362	194	18	information	information	NOUN
cana-4362	194	19	gain	gain	NOUN
cana-4362	194	20	,	,	PUNCT
cana-4362	194	21	followed	follow	VERB
cana-4362	194	22	by	by	ADP
cana-4362	194	23	a	a	DET
cana-4362	194	24	cnn	cnn	PROPN
cana-4362	194	25	-	-	PUNCT
cana-4362	194	26	bilstm	bilstm	NOUN
cana-4362	194	27	model	model	NOUN
cana-4362	194	28	for	for	ADP
cana-4362	194	29	sentiment	sentiment	NOUN
cana-4362	194	30	classification	classification	NOUN
cana-4362	194	31	.	.	PUNCT
cana-4362	195	1	the	the	DET
cana-4362	195	2	hybrid	hybrid	ADJ
cana-4362	195	3	approach	approach	NOUN
cana-4362	195	4	improves	improve	VERB
cana-4362	195	5	the	the	DET
cana-4362	195	6	model	model	NOUN
cana-4362	195	7	’s	’s	PART
cana-4362	195	8	ability	ability	NOUN
cana-4362	195	9	to	to	PART
cana-4362	195	10	detect	detect	VERB
cana-4362	195	11	sentiment	sentiment	NOUN
cana-4362	195	12	from	from	ADP
cana-4362	195	13	large	large	ADJ
cana-4362	195	14	,	,	PUNCT
cana-4362	195	15	unstructured	unstructured	ADJ
cana-4362	195	16	datasets	dataset	NOUN
cana-4362	195	17	like	like	ADP
cana-4362	195	18	twitter	twitter	NOUN
cana-4362	195	19	,	,	PUNCT
cana-4362	195	20	enhancing	enhance	VERB
cana-4362	195	21	both	both	PRON
cana-4362	195	22	accuracy	accuracy	NOUN
cana-4362	195	23	and	and	CCONJ
cana-4362	195	24	processing	processing	NOUN
cana-4362	195	25	time	time	NOUN
cana-4362	195	26	.	.	PUNCT
cana-4362	196	1	their	their	PRON
cana-4362	196	2	results	result	NOUN
cana-4362	196	3	show	show	VERB
cana-4362	196	4	that	that	SCONJ
cana-4362	196	5	the	the	DET
cana-4362	196	6	hybrid	hybrid	ADJ
cana-4362	196	7	feature	feature	NOUN
cana-4362	196	8	selection	selection	NOUN
cana-4362	196	9	significantly	significantly	ADV
cana-4362	196	10	improves	improve	VERB
cana-4362	196	11	model	model	NOUN
cana-4362	196	12	performance	performance	NOUN
cana-4362	196	13	compared	compare	VERB
cana-4362	196	14	to	to	ADP
cana-4362	196	15	traditional	traditional	ADJ
cana-4362	196	16	feature	feature	NOUN
cana-4362	196	17	selection	selection	NOUN
cana-4362	196	18	or	or	CCONJ
cana-4362	196	19	deep	deep	ADJ
cana-4362	196	20	learning	learning	NOUN
cana-4362	196	21	models	model	NOUN
cana-4362	196	22	alone	alone	ADV
cana-4362	196	23	.	.	PUNCT
cana-4362	197	1	this	this	DET
cana-4362	197	2	study	study	NOUN
cana-4362	197	3	demonstrates	demonstrate	VERB
cana-4362	197	4	the	the	DET
cana-4362	197	5	practical	practical	ADJ
cana-4362	197	6	applicability	applicability	NOUN
cana-4362	197	7	of	of	ADP
cana-4362	197	8	hybrid	hybrid	NOUN
cana-4362	197	9	models	model	NOUN
cana-4362	197	10	for	for	ADP
cana-4362	197	11	sentiment	sentiment	NOUN
cana-4362	197	12	classification	classification	NOUN
cana-4362	197	13	in	in	ADP
cana-4362	197	14	social	social	ADJ
cana-4362	197	15	media	medium	NOUN
cana-4362	197	16	contexts	contexts	NOUN
cana-4362	197	17	.	.	PUNCT
cana-4362	198	1	wang	wang	PROPN
cana-4362	198	2	et	et	PROPN
cana-4362	198	3	al	al	PROPN
cana-4362	198	4	.	.	PUNCT
cana-4362	199	1	[	[	X
cana-4362	199	2	25	25	NUM
cana-4362	199	3	]	]	PUNCT
cana-4362	199	4	presents	present	VERB
cana-4362	199	5	a	a	DET
cana-4362	199	6	hybrid	hybrid	ADJ
cana-4362	199	7	feature	feature	NOUN
cana-4362	199	8	selection	selection	NOUN
cana-4362	199	9	technique	technique	NOUN
cana-4362	199	10	using	use	VERB
cana-4362	199	11	genetic	genetic	ADJ
cana-4362	199	12	algorithms	algorithm	NOUN
cana-4362	199	13	(	(	PUNCT
cana-4362	199	14	ga	ga	NOUN
cana-4362	199	15	)	)	PUNCT
cana-4362	199	16	and	and	CCONJ
cana-4362	199	17	a	a	DET
cana-4362	199	18	convolutional	convolutional	ADJ
cana-4362	199	19	neural	neural	ADJ
cana-4362	199	20	network	network	NOUN
cana-4362	199	21	(	(	PUNCT
cana-4362	199	22	cnn	cnn	PROPN
cana-4362	199	23	)	)	PUNCT
cana-4362	199	24	for	for	ADP
cana-4362	199	25	sentiment	sentiment	NOUN
cana-4362	199	26	classification	classification	NOUN
cana-4362	199	27	.	.	PUNCT
cana-4362	200	1	the	the	DET
cana-4362	200	2	paper	paper	NOUN
cana-4362	200	3	addresses	address	VERB
cana-4362	200	4	the	the	DET
cana-4362	200	5	challenge	challenge	NOUN
cana-4362	200	6	of	of	ADP
cana-4362	200	7	high	high	ADJ
cana-4362	200	8	-	-	PUNCT
cana-4362	200	9	dimensional	dimensional	ADJ
cana-4362	200	10	data	datum	NOUN
cana-4362	200	11	in	in	ADP
cana-4362	200	12	social	social	ADJ
cana-4362	200	13	media	medium	NOUN
cana-4362	200	14	sentiment	sentiment	NOUN
cana-4362	200	15	analysis	analysis	NOUN
cana-4362	200	16	by	by	ADP
cana-4362	200	17	integrating	integrate	VERB
cana-4362	200	18	evolutionary	evolutionary	ADJ
cana-4362	200	19	algorithms	algorithm	NOUN
cana-4362	200	20	with	with	ADP
cana-4362	200	21	deep	deep	ADJ
cana-4362	200	22	learning	learning	NOUN
cana-4362	200	23	.	.	PUNCT
cana-4362	201	1	the	the	DET
cana-4362	201	2	authors	author	NOUN
cana-4362	201	3	demonstrate	demonstrate	VERB
cana-4362	201	4	that	that	SCONJ
cana-4362	201	5	the	the	DET
cana-4362	201	6	ga	ga	PROPN
cana-4362	201	7	-	-	PUNCT
cana-4362	201	8	cnn	cnn	PROPN
cana-4362	201	9	hybrid	hybrid	NOUN
cana-4362	201	10	model	model	NOUN
cana-4362	201	11	is	be	AUX
cana-4362	201	12	capable	capable	ADJ
cana-4362	201	13	of	of	ADP
cana-4362	201	14	identifying	identify	VERB
cana-4362	201	15	the	the	DET
cana-4362	201	16	most	most	ADV
cana-4362	201	17	relevant	relevant	ADJ
cana-4362	201	18	features	feature	NOUN
cana-4362	201	19	while	while	SCONJ
cana-4362	201	20	reducing	reduce	VERB
cana-4362	201	21	the	the	DET
cana-4362	201	22	dimensionality	dimensionality	NOUN
cana-4362	201	23	of	of	ADP
cana-4362	201	24	the	the	DET
cana-4362	201	25	data	datum	NOUN
cana-4362	201	26	,	,	PUNCT
cana-4362	201	27	leading	lead	VERB
cana-4362	201	28	to	to	ADP
cana-4362	201	29	improvements	improvement	NOUN
cana-4362	201	30	in	in	ADP
cana-4362	201	31	classification	classification	NOUN
cana-4362	201	32	accuracy	accuracy	NOUN
cana-4362	201	33	and	and	CCONJ
cana-4362	201	34	efficiency	efficiency	NOUN
cana-4362	201	35	.	.	PUNCT
cana-4362	202	1	their	their	PRON
cana-4362	202	2	results	result	NOUN
cana-4362	202	3	,	,	PUNCT
cana-4362	202	4	based	base	VERB
cana-4362	202	5	on	on	ADP
cana-4362	202	6	twitter	twitter	NOUN
cana-4362	202	7	datasets	dataset	NOUN
cana-4362	202	8	,	,	PUNCT
cana-4362	202	9	show	show	VERB
cana-4362	202	10	that	that	SCONJ
cana-4362	202	11	the	the	DET
cana-4362	202	12	hybrid	hybrid	NOUN
cana-4362	202	13	model	model	NOUN
cana-4362	202	14	outperforms	outperform	VERB
cana-4362	202	15	traditional	traditional	ADJ
cana-4362	202	16	deep	deep	ADJ
cana-4362	202	17	learning	learning	NOUN
cana-4362	202	18	approaches	approach	NOUN
cana-4362	202	19	,	,	PUNCT
cana-4362	202	20	particularly	particularly	ADV
cana-4362	202	21	in	in	ADP
cana-4362	202	22	terms	term	NOUN
cana-4362	202	23	of	of	ADP
cana-4362	202	24	processing	processing	NOUN
cana-4362	202	25	speed	speed	NOUN
cana-4362	202	26	and	and	CCONJ
cana-4362	202	27	precision	precision	NOUN
cana-4362	202	28	.	.	PUNCT
cana-4362	203	1	the	the	DET
cana-4362	203	2	study	study	NOUN
cana-4362	203	3	concludes	conclude	VERB
cana-4362	203	4	that	that	SCONJ
cana-4362	203	5	evolutionary	evolutionary	ADJ
cana-4362	203	6	algorithms	algorithm	NOUN
cana-4362	203	7	combined	combine	VERB
cana-4362	203	8	with	with	ADP
cana-4362	203	9	deep	deep	ADJ
cana-4362	203	10	learning	learning	NOUN
cana-4362	203	11	offer	offer	VERB
cana-4362	203	12	an	an	DET
cana-4362	203	13	effective	effective	ADJ
cana-4362	203	14	approach	approach	NOUN
cana-4362	203	15	for	for	ADP
cana-4362	203	16	sentiment	sentiment	NOUN
cana-4362	203	17	classification	classification	NOUN
cana-4362	203	18	.	.	PUNCT
cana-4362	204	1	chen	chen	PROPN
cana-4362	204	2	et	et	PROPN
cana-4362	204	3	al	al	PROPN
cana-4362	204	4	.	.	PUNCT
cana-4362	205	1	[	[	X
cana-4362	205	2	26	26	NUM
cana-4362	205	3	]	]	PUNCT
cana-4362	205	4	introduces	introduce	VERB
cana-4362	205	5	a	a	DET
cana-4362	205	6	hybrid	hybrid	ADJ
cana-4362	205	7	feature	feature	NOUN
cana-4362	205	8	selection	selection	NOUN
cana-4362	205	9	method	method	NOUN
cana-4362	205	10	for	for	ADP
cana-4362	205	11	sentiment	sentiment	NOUN
cana-4362	205	12	classification	classification	NOUN
cana-4362	205	13	of	of	ADP
cana-4362	205	14	social	social	ADJ
cana-4362	205	15	media	medium	NOUN
cana-4362	205	16	texts	text	NOUN
cana-4362	205	17	.	.	PUNCT
cana-4362	206	1	the	the	DET
cana-4362	206	2	authors	author	NOUN
cana-4362	206	3	propose	propose	VERB
cana-4362	206	4	using	use	VERB
cana-4362	206	5	principal	principal	ADJ
cana-4362	206	6	component	component	NOUN
cana-4362	206	7	analysis	analysis	NOUN
cana-4362	206	8	(	(	PUNCT
cana-4362	206	9	pca	pca	NOUN
cana-4362	206	10	)	)	PUNCT
cana-4362	206	11	combined	combine	VERB
cana-4362	206	12	with	with	ADP
cana-4362	206	13	deep	deep	ADJ
cana-4362	206	14	learning	learning	NOUN
cana-4362	206	15	models	model	NOUN
cana-4362	206	16	like	like	ADP
cana-4362	206	17	lstm	lstm	NOUN
cana-4362	206	18	to	to	PART
cana-4362	206	19	enhance	enhance	VERB
cana-4362	206	20	classification	classification	NOUN
cana-4362	206	21	performance	performance	NOUN
cana-4362	206	22	on	on	ADP
cana-4362	206	23	noisy	noisy	ADJ
cana-4362	206	24	social	social	ADJ
cana-4362	206	25	media	medium	NOUN
cana-4362	206	26	data	datum	NOUN
cana-4362	206	27	.	.	PUNCT
cana-4362	207	1	their	their	PRON
cana-4362	207	2	results	result	NOUN
cana-4362	207	3	show	show	VERB
cana-4362	207	4	that	that	SCONJ
cana-4362	207	5	this	this	DET
cana-4362	207	6	hybrid	hybrid	ADJ
cana-4362	207	7	approach	approach	NOUN
cana-4362	207	8	can	can	AUX
cana-4362	207	9	significantly	significantly	ADV
cana-4362	207	10	reduce	reduce	VERB
cana-4362	207	11	the	the	DET
cana-4362	207	12	dimensionality	dimensionality	NOUN
cana-4362	207	13	of	of	ADP
cana-4362	207	14	the	the	DET
cana-4362	207	15	dataset	dataset	NOUN
cana-4362	207	16	,	,	PUNCT
cana-4362	207	17	improve	improve	VERB
cana-4362	207	18	classification	classification	NOUN
cana-4362	207	19	accuracy	accuracy	NOUN
cana-4362	207	20	,	,	PUNCT
cana-4362	207	21	and	and	CCONJ
cana-4362	207	22	shorten	shorten	VERB
cana-4362	207	23	the	the	DET
cana-4362	207	24	training	training	NOUN
cana-4362	207	25	time	time	NOUN
cana-4362	207	26	of	of	ADP
cana-4362	207	27	deep	deep	ADJ
cana-4362	207	28	learning	learning	NOUN
cana-4362	207	29	models	model	NOUN
cana-4362	207	30	.	.	PUNCT
cana-4362	208	1	the	the	DET
cana-4362	208	2	paper	paper	NOUN
cana-4362	208	3	emphasizes	emphasize	VERB
cana-4362	208	4	the	the	DET
cana-4362	208	5	importance	importance	NOUN
cana-4362	208	6	of	of	ADP
cana-4362	208	7	hybrid	hybrid	ADJ
cana-4362	208	8	feature	feature	NOUN
cana-4362	208	9	selection	selection	NOUN
cana-4362	208	10	in	in	ADP
cana-4362	208	11	dealing	deal	VERB
cana-4362	208	12	with	with	ADP
cana-4362	208	13	the	the	DET
cana-4362	208	14	large	large	ADJ
cana-4362	208	15	-	-	PUNCT
cana-4362	208	16	scale	scale	NOUN
cana-4362	208	17	,	,	PUNCT
cana-4362	208	18	highdimensional	highdimensional	ADJ
cana-4362	208	19	nature	nature	NOUN
cana-4362	208	20	of	of	ADP
cana-4362	208	21	social	social	ADJ
cana-4362	208	22	media	medium	NOUN
cana-4362	208	23	datasets	dataset	NOUN
cana-4362	208	24	,	,	PUNCT
cana-4362	208	25	suggesting	suggest	VERB
cana-4362	208	26	that	that	SCONJ
cana-4362	208	27	it	it	PRON
cana-4362	208	28	is	be	AUX
cana-4362	208	29	a	a	DET
cana-4362	208	30	crucial	crucial	ADJ
cana-4362	208	31	step	step	NOUN
cana-4362	208	32	in	in	ADP
cana-4362	208	33	optimizing	optimize	VERB
cana-4362	208	34	sentiment	sentiment	NOUN
cana-4362	208	35	analysis	analysis	NOUN
cana-4362	208	36	models	model	NOUN
cana-4362	208	37	.	.	PUNCT
cana-4362	209	1	sun	sun	PROPN
cana-4362	209	2	et	et	PROPN
cana-4362	209	3	al	al	PROPN
cana-4362	209	4	.	.	PUNCT
cana-4362	210	1	[	[	X
cana-4362	210	2	27	27	NUM
cana-4362	210	3	]	]	PUNCT
cana-4362	210	4	explores	explore	VERB
cana-4362	210	5	a	a	DET
cana-4362	210	6	hybrid	hybrid	ADJ
cana-4362	210	7	approach	approach	NOUN
cana-4362	210	8	to	to	PART
cana-4362	210	9	feature	feature	VERB
cana-4362	210	10	selection	selection	NOUN
cana-4362	210	11	for	for	ADP
cana-4362	210	12	sentiment	sentiment	NOUN
cana-4362	210	13	classification	classification	NOUN
cana-4362	210	14	using	use	VERB
cana-4362	210	15	a	a	DET
cana-4362	210	16	combination	combination	NOUN
cana-4362	210	17	of	of	ADP
cana-4362	210	18	filter	filter	NOUN
cana-4362	210	19	methods	method	NOUN
cana-4362	210	20	and	and	CCONJ
cana-4362	210	21	deep	deep	ADJ
cana-4362	210	22	learning	learning	NOUN
cana-4362	210	23	techniques	technique	NOUN
cana-4362	210	24	.	.	PUNCT
cana-4362	211	1	the	the	DET
cana-4362	211	2	authors	author	NOUN
cana-4362	211	3	employ	employ	VERB
cana-4362	211	4	relieff	relieff	NOUN
cana-4362	211	5	and	and	CCONJ
cana-4362	211	6	chisquare	chisquare	VERB
cana-4362	211	7	for	for	ADP
cana-4362	211	8	feature	feature	NOUN
cana-4362	211	9	selection	selection	NOUN
cana-4362	211	10	,	,	PUNCT
cana-4362	211	11	followed	follow	VERB
cana-4362	211	12	by	by	ADP
cana-4362	211	13	a	a	DET
cana-4362	211	14	cnn	cnn	PROPN
cana-4362	211	15	-	-	PUNCT
cana-4362	211	16	lstm	lstm	ADJ
cana-4362	211	17	architecture	architecture	NOUN
cana-4362	211	18	for	for	ADP
cana-4362	211	19	sentiment	sentiment	NOUN
cana-4362	211	20	classification	classification	NOUN
cana-4362	211	21	.	.	PUNCT
cana-4362	212	1	their	their	PRON
cana-4362	212	2	experiments	experiment	NOUN
cana-4362	212	3	on	on	ADP
cana-4362	212	4	twitter	twitter	NOUN
cana-4362	212	5	data	datum	NOUN
cana-4362	212	6	reveal	reveal	VERB
cana-4362	212	7	that	that	SCONJ
cana-4362	212	8	hybrid	hybrid	ADJ
cana-4362	212	9	feature	feature	NOUN
cana-4362	212	10	selection	selection	NOUN
cana-4362	212	11	leads	lead	VERB
cana-4362	212	12	to	to	ADP
cana-4362	212	13	improved	improve	VERB
cana-4362	212	14	classification	classification	NOUN
cana-4362	212	15	accuracy	accuracy	NOUN
cana-4362	212	16	,	,	PUNCT
cana-4362	212	17	particularly	particularly	ADV
cana-4362	212	18	in	in	ADP
cana-4362	212	19	noisy	noisy	ADJ
cana-4362	212	20	datasets	dataset	NOUN
cana-4362	212	21	.	.	PUNCT
cana-4362	213	1	the	the	DET
cana-4362	213	2	study	study	NOUN
cana-4362	213	3	demonstrates	demonstrate	VERB
cana-4362	213	4	that	that	SCONJ
cana-4362	213	5	the	the	DET
cana-4362	213	6	combination	combination	NOUN
cana-4362	213	7	of	of	ADP
cana-4362	213	8	traditional	traditional	ADJ
cana-4362	213	9	feature	feature	NOUN
cana-4362	213	10	selection	selection	NOUN
cana-4362	213	11	and	and	CCONJ
cana-4362	213	12	deep	deep	ADJ
cana-4362	213	13	learning	learning	NOUN
cana-4362	213	14	allows	allow	VERB
cana-4362	213	15	for	for	ADP
cana-4362	213	16	more	more	ADV
cana-4362	213	17	efficient	efficient	ADJ
cana-4362	213	18	handling	handling	NOUN
cana-4362	213	19	of	of	ADP
cana-4362	213	20	large	large	ADJ
cana-4362	213	21	datasets	dataset	NOUN
cana-4362	213	22	,	,	PUNCT
cana-4362	213	23	improving	improve	VERB
cana-4362	213	24	both	both	DET
cana-4362	213	25	model	model	NOUN
cana-4362	213	26	performance	performance	NOUN
cana-4362	213	27	and	and	CCONJ
cana-4362	213	28	processing	processing	NOUN
cana-4362	213	29	time	time	NOUN
cana-4362	213	30	.	.	PUNCT
cana-4362	214	1	the	the	DET
cana-4362	214	2	authors	author	NOUN
cana-4362	214	3	conclude	conclude	VERB
cana-4362	214	4	that	that	SCONJ
cana-4362	214	5	this	this	DET
cana-4362	214	6	hybrid	hybrid	ADJ
cana-4362	214	7	approach	approach	NOUN
cana-4362	214	8	is	be	AUX
cana-4362	214	9	well	well	ADV
cana-4362	214	10	-	-	PUNCT
cana-4362	214	11	suited	suit	VERB
cana-4362	214	12	for	for	ADP
cana-4362	214	13	real	real	ADJ
cana-4362	214	14	-	-	PUNCT
cana-4362	214	15	time	time	NOUN
cana-4362	214	16	sentiment	sentiment	NOUN
cana-4362	214	17	classification	classification	NOUN
cana-4362	214	18	tasks	task	NOUN
cana-4362	214	19	on	on	ADP
cana-4362	214	20	social	social	ADJ
cana-4362	214	21	media	medium	NOUN
cana-4362	214	22	platforms	platform	NOUN
cana-4362	214	23	.	.	PUNCT
cana-4362	215	1	zhou	zhou	PROPN
cana-4362	215	2	et	et	PROPN
cana-4362	215	3	al	al	PROPN
cana-4362	215	4	.	.	PUNCT
cana-4362	216	1	[	[	X
cana-4362	216	2	28	28	NUM
cana-4362	216	3	]	]	PUNCT
cana-4362	216	4	proposes	propose	VERB
cana-4362	216	5	a	a	DET
cana-4362	216	6	hybrid	hybrid	ADJ
cana-4362	216	7	model	model	NOUN
cana-4362	216	8	for	for	ADP
cana-4362	216	9	sentiment	sentiment	NOUN
cana-4362	216	10	analysis	analysis	NOUN
cana-4362	216	11	that	that	PRON
cana-4362	216	12	integrates	integrate	VERB
cana-4362	216	13	feature	feature	NOUN
cana-4362	216	14	selection	selection	NOUN
cana-4362	216	15	methods	method	NOUN
cana-4362	216	16	like	like	ADP
cana-4362	216	17	information	information	NOUN
cana-4362	216	18	gain	gain	NOUN
cana-4362	216	19	(	(	PUNCT
cana-4362	216	20	ig	ig	PROPN
cana-4362	216	21	)	)	PUNCT
cana-4362	216	22	and	and	CCONJ
cana-4362	216	23	a	a	DET
cana-4362	216	24	deep	deep	ADJ
cana-4362	216	25	learning	learning	NOUN
cana-4362	216	26	architecture	architecture	NOUN
cana-4362	216	27	combining	combine	VERB
cana-4362	216	28	cnn	cnn	PROPN
cana-4362	216	29	and	and	CCONJ
cana-4362	216	30	lstm	lstm	PROPN
cana-4362	216	31	.	.	PUNCT
cana-4362	217	1	the	the	DET
cana-4362	217	2	communications	communication	NOUN
cana-4362	217	3	on	on	ADP
cana-4362	217	4	applied	apply	VERB
cana-4362	217	5	nonlinear	nonlinear	ADJ
cana-4362	217	6	analysis	analysis	NOUN
cana-4362	217	7	issn	issn	NOUN
cana-4362	217	8	:	:	PUNCT
cana-4362	217	9	1074	1074	NUM
cana-4362	217	10	-	-	PUNCT
cana-4362	217	11	133x	133x	NUM
cana-4362	217	12	vol	vol	NOUN
cana-4362	217	13	32	32	NUM
cana-4362	217	14	no	no	NOUN
cana-4362	217	15	.	.	PUNCT
cana-4362	218	1	9s	9s	NUM
cana-4362	218	2	(	(	PUNCT
cana-4362	218	3	2025	2025	NUM
cana-4362	218	4	)	)	PUNCT
cana-4362	218	5	1907	1907	NUM
cana-4362	219	1	https://internationalpubls.com	https://internationalpubls.com	X
cana-4362	219	2	paper	paper	NOUN
cana-4362	219	3	highlights	highlight	VERB
cana-4362	219	4	the	the	DET
cana-4362	219	5	effectiveness	effectiveness	NOUN
cana-4362	219	6	of	of	ADP
cana-4362	219	7	hybrid	hybrid	ADJ
cana-4362	219	8	feature	feature	NOUN
cana-4362	219	9	selection	selection	NOUN
cana-4362	219	10	in	in	ADP
cana-4362	219	11	reducing	reduce	VERB
cana-4362	219	12	the	the	DET
cana-4362	219	13	dimensionality	dimensionality	NOUN
cana-4362	219	14	of	of	ADP
cana-4362	219	15	large	large	ADJ
cana-4362	219	16	social	social	ADJ
cana-4362	219	17	media	medium	NOUN
cana-4362	219	18	datasets	dataset	NOUN
cana-4362	219	19	,	,	PUNCT
cana-4362	219	20	while	while	SCONJ
cana-4362	219	21	the	the	DET
cana-4362	219	22	cnn	cnn	PROPN
cana-4362	219	23	-	-	PUNCT
cana-4362	219	24	lstm	lstm	ADJ
cana-4362	219	25	structure	structure	NOUN
cana-4362	219	26	improves	improve	VERB
cana-4362	219	27	sentiment	sentiment	NOUN
cana-4362	219	28	classification	classification	NOUN
cana-4362	219	29	accuracy	accuracy	NOUN
cana-4362	219	30	.	.	PUNCT
cana-4362	220	1	tested	test	VERB
cana-4362	220	2	on	on	ADP
cana-4362	220	3	twitter	twitter	NOUN
cana-4362	220	4	datasets	dataset	NOUN
cana-4362	220	5	,	,	PUNCT
cana-4362	220	6	the	the	DET
cana-4362	220	7	results	result	NOUN
cana-4362	220	8	show	show	VERB
cana-4362	220	9	that	that	SCONJ
cana-4362	220	10	this	this	DET
cana-4362	220	11	hybrid	hybrid	ADJ
cana-4362	220	12	model	model	NOUN
cana-4362	220	13	outperforms	outperform	VERB
cana-4362	220	14	baseline	baseline	VERB
cana-4362	220	15	approaches	approach	NOUN
cana-4362	220	16	in	in	ADP
cana-4362	220	17	terms	term	NOUN
cana-4362	220	18	of	of	ADP
cana-4362	220	19	both	both	PRON
cana-4362	220	20	accuracy	accuracy	NOUN
cana-4362	220	21	and	and	CCONJ
cana-4362	220	22	computational	computational	ADJ
cana-4362	220	23	efficiency	efficiency	NOUN
cana-4362	220	24	.	.	PUNCT
cana-4362	221	1	the	the	DET
cana-4362	221	2	study	study	NOUN
cana-4362	221	3	concludes	conclude	VERB
cana-4362	221	4	that	that	SCONJ
cana-4362	221	5	integrating	integrate	VERB
cana-4362	221	6	traditional	traditional	ADJ
cana-4362	221	7	feature	feature	NOUN
cana-4362	221	8	selection	selection	NOUN
cana-4362	221	9	with	with	ADP
cana-4362	221	10	deep	deep	ADJ
cana-4362	221	11	learning	learning	NOUN
cana-4362	221	12	offers	offer	VERB
cana-4362	221	13	a	a	DET
cana-4362	221	14	scalable	scalable	ADJ
cana-4362	221	15	solution	solution	NOUN
cana-4362	221	16	for	for	ADP
cana-4362	221	17	sentiment	sentiment	NOUN
cana-4362	221	18	classification	classification	NOUN
cana-4362	221	19	,	,	PUNCT
cana-4362	221	20	particularly	particularly	ADV
cana-4362	221	21	for	for	ADP
cana-4362	221	22	handling	handle	VERB
cana-4362	221	23	the	the	DET
cana-4362	221	24	vast	vast	ADJ
cana-4362	221	25	amounts	amount	NOUN
cana-4362	221	26	of	of	ADP
cana-4362	221	27	unstructured	unstructured	ADJ
cana-4362	221	28	data	datum	NOUN
cana-4362	221	29	in	in	ADP
cana-4362	221	30	social	social	ADJ
cana-4362	221	31	media	medium	NOUN
cana-4362	221	32	.	.	PUNCT
cana-4362	222	1	zhang	zhang	PROPN
cana-4362	222	2	et	et	PROPN
cana-4362	222	3	al	al	PROPN
cana-4362	222	4	.	.	PUNCT
cana-4362	223	1	[	[	X
cana-4362	223	2	29	29	NUM
cana-4362	223	3	]	]	PUNCT
cana-4362	223	4	presents	present	VERB
cana-4362	223	5	a	a	DET
cana-4362	223	6	hybrid	hybrid	ADJ
cana-4362	223	7	feature	feature	NOUN
cana-4362	223	8	selection	selection	NOUN
cana-4362	223	9	approach	approach	NOUN
cana-4362	223	10	combining	combine	VERB
cana-4362	223	11	statistical	statistical	ADJ
cana-4362	223	12	and	and	CCONJ
cana-4362	223	13	deep	deep	ADJ
cana-4362	223	14	learning	learning	NOUN
cana-4362	223	15	methods	method	NOUN
cana-4362	223	16	for	for	ADP
cana-4362	223	17	sentiment	sentiment	NOUN
cana-4362	223	18	classification	classification	NOUN
cana-4362	223	19	in	in	ADP
cana-4362	223	20	social	social	ADJ
cana-4362	223	21	media	medium	NOUN
cana-4362	223	22	data	datum	NOUN
cana-4362	223	23	.	.	PUNCT
cana-4362	224	1	the	the	DET
cana-4362	224	2	authors	author	NOUN
cana-4362	224	3	develop	develop	VERB
cana-4362	224	4	a	a	DET
cana-4362	224	5	feature	feature	NOUN
cana-4362	224	6	ranking	rank	VERB
cana-4362	224	7	mechanism	mechanism	NOUN
cana-4362	224	8	that	that	PRON
cana-4362	224	9	integrates	integrate	VERB
cana-4362	224	10	both	both	CCONJ
cana-4362	224	11	filter	filter	NOUN
cana-4362	224	12	and	and	CCONJ
cana-4362	224	13	wrapper	wrapper	NOUN
cana-4362	224	14	techniques	technique	NOUN
cana-4362	224	15	,	,	PUNCT
cana-4362	224	16	followed	follow	VERB
cana-4362	224	17	by	by	ADP
cana-4362	224	18	the	the	DET
cana-4362	224	19	application	application	NOUN
cana-4362	224	20	of	of	ADP
cana-4362	224	21	cnn	cnn	PROPN
cana-4362	224	22	and	and	CCONJ
cana-4362	224	23	lstm	lstm	NOUN
cana-4362	224	24	models	model	NOUN
cana-4362	224	25	to	to	PART
cana-4362	224	26	refine	refine	VERB
cana-4362	224	27	the	the	DET
cana-4362	224	28	selected	select	VERB
cana-4362	224	29	features	feature	NOUN
cana-4362	224	30	.	.	PUNCT
cana-4362	225	1	tested	test	VERB
cana-4362	225	2	on	on	ADP
cana-4362	225	3	twitter	twitter	NOUN
cana-4362	225	4	and	and	CCONJ
cana-4362	225	5	reddit	reddit	NOUN
cana-4362	225	6	datasets	dataset	NOUN
cana-4362	225	7	,	,	PUNCT
cana-4362	225	8	the	the	DET
cana-4362	225	9	hybrid	hybrid	NOUN
cana-4362	225	10	model	model	NOUN
cana-4362	225	11	demonstrates	demonstrate	VERB
cana-4362	225	12	significant	significant	ADJ
cana-4362	225	13	improvements	improvement	NOUN
cana-4362	225	14	in	in	ADP
cana-4362	225	15	classification	classification	NOUN
cana-4362	225	16	accuracy	accuracy	NOUN
cana-4362	225	17	,	,	PUNCT
cana-4362	225	18	precision	precision	NOUN
cana-4362	225	19	,	,	PUNCT
cana-4362	225	20	and	and	CCONJ
cana-4362	225	21	recall	recall	NOUN
cana-4362	225	22	compared	compare	VERB
cana-4362	225	23	to	to	ADP
cana-4362	225	24	traditional	traditional	ADJ
cana-4362	225	25	methods	method	NOUN
cana-4362	225	26	.	.	PUNCT
cana-4362	226	1	the	the	DET
cana-4362	226	2	study	study	NOUN
cana-4362	226	3	concludes	conclude	VERB
cana-4362	226	4	that	that	SCONJ
cana-4362	226	5	combining	combine	VERB
cana-4362	226	6	feature	feature	NOUN
cana-4362	226	7	selection	selection	NOUN
cana-4362	226	8	with	with	ADP
cana-4362	226	9	deep	deep	ADJ
cana-4362	226	10	learning	learning	NOUN
cana-4362	226	11	techniques	technique	NOUN
cana-4362	226	12	provides	provide	VERB
cana-4362	226	13	an	an	DET
cana-4362	226	14	efficient	efficient	ADJ
cana-4362	226	15	and	and	CCONJ
cana-4362	226	16	effective	effective	ADJ
cana-4362	226	17	approach	approach	NOUN
cana-4362	226	18	for	for	ADP
cana-4362	226	19	sentiment	sentiment	NOUN
cana-4362	226	20	analysis	analysis	NOUN
cana-4362	226	21	on	on	ADP
cana-4362	226	22	large	large	ADJ
cana-4362	226	23	social	social	ADJ
cana-4362	226	24	media	medium	NOUN
cana-4362	226	25	datasets	dataset	NOUN
cana-4362	226	26	.	.	PUNCT
cana-4362	227	1	yu	yu	PROPN
cana-4362	227	2	et	et	PROPN
cana-4362	227	3	al	al	PROPN
cana-4362	227	4	.	.	PUNCT
cana-4362	228	1	[	[	X
cana-4362	228	2	30	30	NUM
cana-4362	228	3	]	]	PUNCT
cana-4362	228	4	explores	explore	VERB
cana-4362	228	5	hybrid	hybrid	ADJ
cana-4362	228	6	feature	feature	NOUN
cana-4362	228	7	extraction	extraction	NOUN
cana-4362	228	8	and	and	CCONJ
cana-4362	228	9	selection	selection	NOUN
cana-4362	228	10	techniques	technique	NOUN
cana-4362	228	11	for	for	ADP
cana-4362	228	12	sentiment	sentiment	NOUN
cana-4362	228	13	classification	classification	NOUN
cana-4362	228	14	of	of	ADP
cana-4362	228	15	online	online	ADJ
cana-4362	228	16	social	social	ADJ
cana-4362	228	17	media	medium	NOUN
cana-4362	228	18	posts	post	NOUN
cana-4362	228	19	.	.	PUNCT
cana-4362	229	1	the	the	DET
cana-4362	229	2	authors	author	NOUN
cana-4362	229	3	propose	propose	VERB
cana-4362	229	4	a	a	DET
cana-4362	229	5	combination	combination	NOUN
cana-4362	229	6	of	of	ADP
cana-4362	229	7	principal	principal	ADJ
cana-4362	229	8	component	component	NOUN
cana-4362	229	9	analysis	analysis	NOUN
cana-4362	229	10	(	(	PUNCT
cana-4362	229	11	pca	pca	NOUN
cana-4362	229	12	)	)	PUNCT
cana-4362	229	13	and	and	CCONJ
cana-4362	229	14	recursive	recursive	ADJ
cana-4362	229	15	feature	feature	NOUN
cana-4362	229	16	elimination	elimination	NOUN
cana-4362	229	17	(	(	PUNCT
cana-4362	229	18	rfe	rfe	NOUN
cana-4362	229	19	)	)	PUNCT
cana-4362	229	20	for	for	ADP
cana-4362	229	21	feature	feature	NOUN
cana-4362	229	22	selection	selection	NOUN
cana-4362	229	23	,	,	PUNCT
cana-4362	229	24	followed	follow	VERB
cana-4362	229	25	by	by	ADP
cana-4362	229	26	a	a	DET
cana-4362	229	27	cnn	cnn	PROPN
cana-4362	229	28	-	-	PUNCT
cana-4362	229	29	lstm	lstm	ADJ
cana-4362	229	30	hybrid	hybrid	NOUN
cana-4362	229	31	model	model	NOUN
cana-4362	229	32	for	for	ADP
cana-4362	229	33	classification	classification	NOUN
cana-4362	229	34	.	.	PUNCT
cana-4362	230	1	their	their	PRON
cana-4362	230	2	experiments	experiment	NOUN
cana-4362	230	3	on	on	ADP
cana-4362	230	4	twitter	twitter	NOUN
cana-4362	230	5	data	datum	NOUN
cana-4362	230	6	show	show	VERB
cana-4362	230	7	that	that	SCONJ
cana-4362	230	8	hybrid	hybrid	ADJ
cana-4362	230	9	feature	feature	NOUN
cana-4362	230	10	selection	selection	NOUN
cana-4362	230	11	improves	improve	VERB
cana-4362	230	12	both	both	DET
cana-4362	230	13	classification	classification	NOUN
cana-4362	230	14	accuracy	accuracy	NOUN
cana-4362	230	15	and	and	CCONJ
cana-4362	230	16	computational	computational	ADJ
cana-4362	230	17	efficiency	efficiency	NOUN
cana-4362	230	18	.	.	PUNCT
cana-4362	231	1	the	the	DET
cana-4362	231	2	study	study	NOUN
cana-4362	231	3	highlights	highlight	VERB
cana-4362	231	4	the	the	DET
cana-4362	231	5	potential	potential	NOUN
cana-4362	231	6	of	of	ADP
cana-4362	231	7	hybrid	hybrid	ADJ
cana-4362	231	8	models	model	NOUN
cana-4362	231	9	to	to	PART
cana-4362	231	10	handle	handle	VERB
cana-4362	231	11	the	the	DET
cana-4362	231	12	noisy	noisy	ADJ
cana-4362	231	13	and	and	CCONJ
cana-4362	231	14	high	high	ADV
cana-4362	231	15	-	-	PUNCT
cana-4362	231	16	dimensional	dimensional	ADJ
cana-4362	231	17	nature	nature	NOUN
cana-4362	231	18	of	of	ADP
cana-4362	231	19	social	social	ADJ
cana-4362	231	20	media	medium	NOUN
cana-4362	231	21	data	datum	NOUN
cana-4362	231	22	,	,	PUNCT
cana-4362	231	23	offering	offer	VERB
cana-4362	231	24	a	a	DET
cana-4362	231	25	practical	practical	ADJ
cana-4362	231	26	solution	solution	NOUN
cana-4362	231	27	for	for	ADP
cana-4362	231	28	real	real	ADJ
cana-4362	231	29	-	-	PUNCT
cana-4362	231	30	world	world	NOUN
cana-4362	231	31	sentiment	sentiment	NOUN
cana-4362	231	32	analysis	analysis	NOUN
cana-4362	231	33	applications	application	NOUN
cana-4362	231	34	.	.	PUNCT
cana-4362	232	1	research	research	NOUN
cana-4362	232	2	methodology	methodology	NOUN
cana-4362	232	3	the	the	DET
cana-4362	232	4	system	system	NOUN
cana-4362	232	5	architecture	architecture	NOUN
cana-4362	232	6	for	for	ADP
cana-4362	232	7	hybrid	hybrid	ADJ
cana-4362	232	8	feature	feature	NOUN
cana-4362	232	9	selection	selection	NOUN
cana-4362	232	10	in	in	ADP
cana-4362	232	11	sentiment	sentiment	NOUN
cana-4362	232	12	classification	classification	NOUN
cana-4362	232	13	of	of	ADP
cana-4362	232	14	social	social	ADJ
cana-4362	232	15	media	medium	NOUN
cana-4362	232	16	datasets	dataset	NOUN
cana-4362	232	17	encompasses	encompass	VERB
cana-4362	232	18	several	several	ADJ
cana-4362	232	19	key	key	ADJ
cana-4362	232	20	components	component	NOUN
cana-4362	232	21	that	that	PRON
cana-4362	232	22	work	work	VERB
cana-4362	232	23	together	together	ADV
cana-4362	232	24	to	to	ADP
cana-4362	232	25	preprocess	preprocess	NOUN
cana-4362	232	26	data	datum	NOUN
cana-4362	232	27	,	,	PUNCT
cana-4362	232	28	select	select	VERB
cana-4362	232	29	relevant	relevant	ADJ
cana-4362	232	30	features	feature	NOUN
cana-4362	232	31	,	,	PUNCT
cana-4362	232	32	and	and	CCONJ
cana-4362	232	33	apply	apply	VERB
cana-4362	232	34	deep	deep	ADJ
cana-4362	232	35	learning	learning	NOUN
cana-4362	232	36	models	model	NOUN
cana-4362	232	37	for	for	ADP
cana-4362	232	38	accurate	accurate	ADJ
cana-4362	232	39	sentiment	sentiment	NOUN
cana-4362	232	40	analysis	analysis	NOUN
cana-4362	232	41	.	.	PUNCT
cana-4362	233	1	this	this	DET
cana-4362	233	2	architecture	architecture	NOUN
cana-4362	233	3	is	be	AUX
cana-4362	233	4	designed	design	VERB
cana-4362	233	5	to	to	PART
cana-4362	233	6	efficiently	efficiently	ADV
cana-4362	233	7	manage	manage	VERB
cana-4362	233	8	the	the	DET
cana-4362	233	9	high	high	ADJ
cana-4362	233	10	dimensionality	dimensionality	NOUN
cana-4362	233	11	and	and	CCONJ
cana-4362	233	12	noisy	noisy	ADJ
cana-4362	233	13	nature	nature	NOUN
cana-4362	233	14	of	of	ADP
cana-4362	233	15	social	social	ADJ
cana-4362	233	16	media	medium	NOUN
cana-4362	233	17	data	datum	NOUN
cana-4362	233	18	while	while	SCONJ
cana-4362	233	19	enhancing	enhance	VERB
cana-4362	233	20	classification	classification	NOUN
cana-4362	233	21	performance	performance	NOUN
cana-4362	233	22	.	.	PUNCT
cana-4362	234	1	below	below	ADV
cana-4362	234	2	is	be	AUX
cana-4362	234	3	a	a	DET
cana-4362	234	4	detailed	detailed	ADJ
cana-4362	234	5	description	description	NOUN
cana-4362	234	6	of	of	ADP
cana-4362	234	7	the	the	DET
cana-4362	234	8	various	various	ADJ
cana-4362	234	9	components	component	NOUN
cana-4362	234	10	of	of	ADP
cana-4362	234	11	the	the	DET
cana-4362	234	12	system	system	NOUN
cana-4362	234	13	architecture	architecture	NOUN
cana-4362	234	14	.	.	PUNCT
cana-4362	235	1	figure	figure	NOUN
cana-4362	235	2	1	1	NUM
cana-4362	235	3	:	:	PUNCT
cana-4362	235	4	proposed	propose	VERB
cana-4362	235	5	system	system	NOUN
cana-4362	235	6	architecture	architecture	NOUN
cana-4362	235	7	for	for	ADP
cana-4362	235	8	sentiment	sentiment	NOUN
cana-4362	235	9	analysis	analysis	NOUN
cana-4362	235	10	using	use	VERB
cana-4362	235	11	hybrid	hybrid	ADJ
cana-4362	235	12	deep	deep	ADJ
cana-4362	235	13	learning	learning	NOUN
cana-4362	235	14	model	model	NOUN
cana-4362	235	15	for	for	ADP
cana-4362	235	16	social	social	ADJ
cana-4362	235	17	media	medium	NOUN
cana-4362	235	18	dataset	dataset	NOUN
cana-4362	235	19	communications	communication	NOUN
cana-4362	235	20	on	on	ADP
cana-4362	235	21	applied	apply	VERB
cana-4362	235	22	nonlinear	nonlinear	ADJ
cana-4362	235	23	analysis	analysis	NOUN
cana-4362	235	24	issn	issn	NOUN
cana-4362	235	25	:	:	PUNCT
cana-4362	235	26	1074	1074	NUM
cana-4362	235	27	-	-	PUNCT
cana-4362	235	28	133x	133x	NUM
cana-4362	235	29	vol	vol	NOUN
cana-4362	235	30	32	32	NUM
cana-4362	235	31	no	no	NOUN
cana-4362	235	32	.	.	PUNCT
cana-4362	236	1	9s	9s	NUM
cana-4362	236	2	(	(	PUNCT
cana-4362	236	3	2025	2025	NUM
cana-4362	236	4	)	)	PUNCT
cana-4362	236	5	1908	1908	NUM
cana-4362	236	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-4362	236	7	1	1	X
cana-4362	236	8	.	.	PUNCT
cana-4362	236	9	data	datum	NOUN
cana-4362	236	10	collection	collection	NOUN
cana-4362	236	11	module	module	NOUN
cana-4362	236	12	the	the	DET
cana-4362	236	13	first	first	ADJ
cana-4362	236	14	component	component	NOUN
cana-4362	236	15	of	of	ADP
cana-4362	236	16	the	the	DET
cana-4362	236	17	architecture	architecture	NOUN
cana-4362	236	18	is	be	AUX
cana-4362	236	19	the	the	DET
cana-4362	236	20	data	data	NOUN
cana-4362	236	21	collection	collection	NOUN
cana-4362	236	22	module	module	NOUN
cana-4362	236	23	,	,	PUNCT
cana-4362	236	24	which	which	PRON
cana-4362	236	25	is	be	AUX
cana-4362	236	26	responsible	responsible	ADJ
cana-4362	236	27	for	for	ADP
cana-4362	236	28	gathering	gather	VERB
cana-4362	236	29	social	social	ADJ
cana-4362	236	30	media	medium	NOUN
cana-4362	236	31	data	datum	NOUN
cana-4362	236	32	from	from	ADP
cana-4362	236	33	platforms	platform	NOUN
cana-4362	236	34	like	like	ADP
cana-4362	236	35	twitter	twitter	NOUN
cana-4362	236	36	,	,	PUNCT
cana-4362	236	37	facebook	facebook	NOUN
cana-4362	236	38	,	,	PUNCT
cana-4362	236	39	or	or	CCONJ
cana-4362	236	40	reddit	reddit	NOUN
cana-4362	236	41	.	.	PUNCT
cana-4362	237	1	this	this	DET
cana-4362	237	2	module	module	NOUN
cana-4362	237	3	employs	employ	VERB
cana-4362	237	4	application	application	NOUN
cana-4362	237	5	programming	programming	NOUN
cana-4362	237	6	interfaces	interface	NOUN
cana-4362	237	7	(	(	PUNCT
cana-4362	237	8	apis	apis	ADJ
cana-4362	237	9	)	)	PUNCT
cana-4362	237	10	or	or	CCONJ
cana-4362	237	11	web	web	NOUN
cana-4362	237	12	scraping	scrape	VERB
cana-4362	237	13	techniques	technique	NOUN
cana-4362	237	14	to	to	PART
cana-4362	237	15	extract	extract	VERB
cana-4362	237	16	relevant	relevant	ADJ
cana-4362	237	17	posts	post	NOUN
cana-4362	237	18	,	,	PUNCT
cana-4362	237	19	comments	comment	NOUN
cana-4362	237	20	,	,	PUNCT
cana-4362	237	21	or	or	CCONJ
cana-4362	237	22	tweets	tweet	NOUN
cana-4362	237	23	containing	contain	VERB
cana-4362	237	24	user	user	NOUN
cana-4362	237	25	sentiments	sentiment	NOUN
cana-4362	237	26	.	.	PUNCT
cana-4362	238	1	the	the	DET
cana-4362	238	2	collected	collect	VERB
cana-4362	238	3	data	datum	NOUN
cana-4362	238	4	typically	typically	ADV
cana-4362	238	5	includes	include	VERB
cana-4362	238	6	text	text	NOUN
cana-4362	238	7	content	content	NOUN
cana-4362	238	8	,	,	PUNCT
cana-4362	238	9	metadata	metadata	NOUN
cana-4362	238	10	such	such	ADJ
cana-4362	238	11	as	as	ADP
cana-4362	238	12	user	user	NOUN
cana-4362	238	13	information	information	NOUN
cana-4362	238	14	,	,	PUNCT
cana-4362	238	15	timestamps	timestamp	NOUN
cana-4362	238	16	,	,	PUNCT
cana-4362	238	17	and	and	CCONJ
cana-4362	238	18	any	any	DET
cana-4362	238	19	other	other	ADJ
cana-4362	238	20	relevant	relevant	ADJ
cana-4362	238	21	attributes	attribute	NOUN
cana-4362	238	22	.	.	PUNCT
cana-4362	239	1	the	the	DET
cana-4362	239	2	data	datum	NOUN
cana-4362	239	3	can	can	AUX
cana-4362	239	4	be	be	AUX
cana-4362	239	5	collected	collect	VERB
cana-4362	239	6	in	in	ADP
cana-4362	239	7	real	real	ADJ
cana-4362	239	8	-	-	PUNCT
cana-4362	239	9	time	time	NOUN
cana-4362	239	10	or	or	CCONJ
cana-4362	239	11	as	as	ADP
cana-4362	239	12	a	a	DET
cana-4362	239	13	batch	batch	NOUN
cana-4362	239	14	process	process	NOUN
cana-4362	239	15	,	,	PUNCT
cana-4362	239	16	depending	depend	VERB
cana-4362	239	17	on	on	ADP
cana-4362	239	18	the	the	DET
cana-4362	239	19	application	application	NOUN
cana-4362	239	20	requirements	requirement	NOUN
cana-4362	239	21	.	.	PUNCT
cana-4362	240	1	2	2	X
cana-4362	240	2	.	.	X
cana-4362	240	3	data	datum	NOUN
cana-4362	240	4	preprocessing	preprocessing	NOUN
cana-4362	240	5	module	module	NOUN
cana-4362	240	6	once	once	SCONJ
cana-4362	240	7	the	the	DET
cana-4362	240	8	data	data	NOUN
cana-4362	240	9	is	be	AUX
cana-4362	240	10	collected	collect	VERB
cana-4362	240	11	,	,	PUNCT
cana-4362	240	12	the	the	DET
cana-4362	240	13	preprocessing	preprocessing	NOUN
cana-4362	240	14	module	module	NOUN
cana-4362	240	15	cleans	clean	VERB
cana-4362	240	16	and	and	CCONJ
cana-4362	240	17	prepares	prepare	VERB
cana-4362	240	18	the	the	DET
cana-4362	240	19	data	datum	NOUN
cana-4362	240	20	for	for	ADP
cana-4362	240	21	analysis	analysis	NOUN
cana-4362	240	22	.	.	PUNCT
cana-4362	241	1	this	this	DET
cana-4362	241	2	stage	stage	NOUN
cana-4362	241	3	includes	include	VERB
cana-4362	241	4	several	several	ADJ
cana-4362	241	5	crucial	crucial	ADJ
cana-4362	241	6	steps	step	NOUN
cana-4362	241	7	:	:	PUNCT
cana-4362	241	8	text	text	NOUN
cana-4362	241	9	normalization	normalization	NOUN
cana-4362	241	10	:	:	PUNCT
cana-4362	241	11	this	this	PRON
cana-4362	241	12	involves	involve	VERB
cana-4362	241	13	converting	convert	VERB
cana-4362	241	14	text	text	NOUN
cana-4362	241	15	to	to	ADP
cana-4362	241	16	a	a	DET
cana-4362	241	17	uniform	uniform	ADJ
cana-4362	241	18	format	format	NOUN
cana-4362	241	19	,	,	PUNCT
cana-4362	241	20	which	which	PRON
cana-4362	241	21	may	may	AUX
cana-4362	241	22	include	include	VERB
cana-4362	241	23	converting	convert	VERB
cana-4362	241	24	to	to	PART
cana-4362	241	25	lowercase	lowercase	VERB
cana-4362	241	26	,	,	PUNCT
cana-4362	241	27	removing	remove	VERB
cana-4362	241	28	special	special	ADJ
cana-4362	241	29	characters	character	NOUN
cana-4362	241	30	,	,	PUNCT
cana-4362	241	31	and	and	CCONJ
cana-4362	241	32	correcting	correct	VERB
cana-4362	241	33	misspellings	misspelling	NOUN
cana-4362	241	34	.	.	PUNCT
cana-4362	242	1	tokenization	tokenization	NOUN
cana-4362	242	2	:	:	PUNCT
cana-4362	242	3	the	the	DET
cana-4362	242	4	cleaned	clean	VERB
cana-4362	242	5	text	text	NOUN
cana-4362	242	6	is	be	AUX
cana-4362	242	7	split	split	VERB
cana-4362	242	8	into	into	ADP
cana-4362	242	9	tokens	token	NOUN
cana-4362	242	10	(	(	PUNCT
cana-4362	242	11	words	word	NOUN
cana-4362	242	12	or	or	CCONJ
cana-4362	242	13	phrases	phrase	NOUN
cana-4362	242	14	)	)	PUNCT
cana-4362	242	15	to	to	PART
cana-4362	242	16	facilitate	facilitate	VERB
cana-4362	242	17	further	further	ADJ
cana-4362	242	18	processing	processing	NOUN
cana-4362	242	19	.	.	PUNCT
cana-4362	243	1	stop	stop	VERB
cana-4362	243	2	word	word	NOUN
cana-4362	243	3	removal	removal	NOUN
cana-4362	243	4	:	:	PUNCT
cana-4362	243	5	commonly	commonly	ADV
cana-4362	243	6	used	use	VERB
cana-4362	243	7	words	word	NOUN
cana-4362	243	8	(	(	PUNCT
cana-4362	243	9	stop	stop	VERB
cana-4362	243	10	words	word	NOUN
cana-4362	243	11	)	)	PUNCT
cana-4362	243	12	that	that	PRON
cana-4362	243	13	do	do	AUX
cana-4362	243	14	not	not	PART
cana-4362	243	15	contribute	contribute	VERB
cana-4362	243	16	to	to	ADP
cana-4362	243	17	sentiment	sentiment	NOUN
cana-4362	243	18	analysis	analysis	NOUN
cana-4362	243	19	(	(	PUNCT
cana-4362	243	20	e.g.	e.g.	ADV
cana-4362	243	21	,	,	PUNCT
cana-4362	243	22	“	"	PUNCT
cana-4362	243	23	and	and	CCONJ
cana-4362	243	24	,	,	PUNCT
cana-4362	243	25	”	"	PUNCT
cana-4362	243	26	“	"	PUNCT
cana-4362	243	27	the	the	PRON
cana-4362	243	28	,	,	PUNCT
cana-4362	243	29	”	"	PUNCT
cana-4362	243	30	“	"	PUNCT
cana-4362	243	31	is	be	AUX
cana-4362	243	32	”	"	PUNCT
cana-4362	243	33	)	)	PUNCT
cana-4362	243	34	are	be	AUX
cana-4362	243	35	removed	remove	VERB
cana-4362	243	36	from	from	ADP
cana-4362	243	37	the	the	DET
cana-4362	243	38	text	text	NOUN
cana-4362	243	39	.	.	PUNCT
cana-4362	244	1	stemming	stem	VERB
cana-4362	244	2	/	/	SYM
cana-4362	244	3	lemmatization	lemmatization	NOUN
cana-4362	244	4	:	:	PUNCT
cana-4362	244	5	words	word	NOUN
cana-4362	244	6	are	be	AUX
cana-4362	244	7	reduced	reduce	VERB
cana-4362	244	8	to	to	ADP
cana-4362	244	9	their	their	PRON
cana-4362	244	10	base	base	NOUN
cana-4362	244	11	or	or	CCONJ
cana-4362	244	12	root	root	NOUN
cana-4362	244	13	form	form	NOUN
cana-4362	244	14	to	to	PART
cana-4362	244	15	reduce	reduce	VERB
cana-4362	244	16	the	the	DET
cana-4362	244	17	dimensionality	dimensionality	NOUN
cana-4362	244	18	of	of	ADP
cana-4362	244	19	the	the	DET
cana-4362	244	20	dataset	dataset	NOUN
cana-4362	244	21	.	.	PUNCT
cana-4362	245	1	this	this	DET
cana-4362	245	2	preprocessing	preprocessing	NOUN
cana-4362	245	3	step	step	NOUN
cana-4362	245	4	ensures	ensure	VERB
cana-4362	245	5	that	that	SCONJ
cana-4362	245	6	the	the	DET
cana-4362	245	7	data	data	NOUN
cana-4362	245	8	is	be	AUX
cana-4362	245	9	clean	clean	ADJ
cana-4362	245	10	and	and	CCONJ
cana-4362	245	11	relevant	relevant	ADJ
cana-4362	245	12	,	,	PUNCT
cana-4362	245	13	laying	lay	VERB
cana-4362	245	14	a	a	DET
cana-4362	245	15	solid	solid	ADJ
cana-4362	245	16	foundation	foundation	NOUN
cana-4362	245	17	for	for	ADP
cana-4362	245	18	the	the	DET
cana-4362	245	19	subsequent	subsequent	ADJ
cana-4362	245	20	feature	feature	NOUN
cana-4362	245	21	selection	selection	NOUN
cana-4362	245	22	and	and	CCONJ
cana-4362	245	23	classification	classification	NOUN
cana-4362	245	24	stages	stage	NOUN
cana-4362	245	25	.	.	PUNCT
cana-4362	246	1	3	3	X
cana-4362	246	2	.	.	X
cana-4362	246	3	hybrid	hybrid	ADJ
cana-4362	246	4	feature	feature	NOUN
cana-4362	246	5	selection	selection	NOUN
cana-4362	246	6	module	module	NOUN
cana-4362	246	7	the	the	DET
cana-4362	246	8	core	core	NOUN
cana-4362	246	9	of	of	ADP
cana-4362	246	10	the	the	DET
cana-4362	246	11	system	system	NOUN
cana-4362	246	12	architecture	architecture	NOUN
cana-4362	246	13	lies	lie	VERB
cana-4362	246	14	in	in	ADP
cana-4362	246	15	the	the	DET
cana-4362	246	16	hybrid	hybrid	ADJ
cana-4362	246	17	feature	feature	NOUN
cana-4362	246	18	selection	selection	NOUN
cana-4362	246	19	module	module	NOUN
cana-4362	246	20	.	.	PUNCT
cana-4362	247	1	this	this	DET
cana-4362	247	2	component	component	NOUN
cana-4362	247	3	employs	employ	VERB
cana-4362	247	4	a	a	DET
cana-4362	247	5	combination	combination	NOUN
cana-4362	247	6	of	of	ADP
cana-4362	247	7	filter	filter	NOUN
cana-4362	247	8	-	-	PUNCT
cana-4362	247	9	based	base	VERB
cana-4362	247	10	and	and	CCONJ
cana-4362	247	11	wrapper	wrapper	NOUN
cana-4362	247	12	-	-	PUNCT
cana-4362	247	13	based	base	VERB
cana-4362	247	14	feature	feature	NOUN
cana-4362	247	15	selection	selection	NOUN
cana-4362	247	16	techniques	technique	NOUN
cana-4362	247	17	to	to	PART
cana-4362	247	18	identify	identify	VERB
cana-4362	247	19	the	the	DET
cana-4362	247	20	most	most	ADV
cana-4362	247	21	relevant	relevant	ADJ
cana-4362	247	22	features	feature	NOUN
cana-4362	247	23	for	for	ADP
cana-4362	247	24	sentiment	sentiment	NOUN
cana-4362	247	25	classification	classification	NOUN
cana-4362	247	26	.	.	PUNCT
cana-4362	248	1	filter	filter	NOUN
cana-4362	248	2	-	-	PUNCT
cana-4362	248	3	based	base	VERB
cana-4362	248	4	selection	selection	NOUN
cana-4362	248	5	:	:	PUNCT
cana-4362	248	6	initial	initial	ADJ
cana-4362	248	7	feature	feature	NOUN
cana-4362	248	8	selection	selection	NOUN
cana-4362	248	9	is	be	AUX
cana-4362	248	10	performed	perform	VERB
cana-4362	248	11	using	use	VERB
cana-4362	248	12	filter	filter	NOUN
cana-4362	248	13	methods	method	NOUN
cana-4362	248	14	such	such	ADJ
cana-4362	248	15	as	as	ADP
cana-4362	248	16	chi	chi	NOUN
cana-4362	248	17	-	-	PUNCT
cana-4362	248	18	square	square	ADJ
cana-4362	248	19	,	,	PUNCT
cana-4362	248	20	information	information	NOUN
cana-4362	248	21	gain	gain	NOUN
cana-4362	248	22	,	,	PUNCT
cana-4362	248	23	or	or	CCONJ
cana-4362	248	24	mutual	mutual	ADJ
cana-4362	248	25	information	information	NOUN
cana-4362	248	26	.	.	PUNCT
cana-4362	249	1	these	these	DET
cana-4362	249	2	methods	method	NOUN
cana-4362	249	3	evaluate	evaluate	VERB
cana-4362	249	4	the	the	DET
cana-4362	249	5	relevance	relevance	NOUN
cana-4362	249	6	of	of	ADP
cana-4362	249	7	individual	individual	ADJ
cana-4362	249	8	features	feature	NOUN
cana-4362	249	9	based	base	VERB
cana-4362	249	10	on	on	ADP
cana-4362	249	11	statistical	statistical	ADJ
cana-4362	249	12	measures	measure	NOUN
cana-4362	249	13	and	and	CCONJ
cana-4362	249	14	select	select	VERB
cana-4362	249	15	a	a	DET
cana-4362	249	16	subset	subset	NOUN
cana-4362	249	17	of	of	ADP
cana-4362	249	18	informative	informative	ADJ
cana-4362	249	19	features	feature	NOUN
cana-4362	249	20	while	while	SCONJ
cana-4362	249	21	discarding	discard	VERB
cana-4362	249	22	irrelevant	irrelevant	ADJ
cana-4362	249	23	ones	one	NOUN
cana-4362	249	24	.	.	PUNCT
cana-4362	250	1	wrapper	wrapper	NOUN
cana-4362	250	2	-	-	PUNCT
cana-4362	250	3	based	base	VERB
cana-4362	250	4	selection	selection	NOUN
cana-4362	250	5	:	:	PUNCT
cana-4362	250	6	following	follow	VERB
cana-4362	250	7	the	the	DET
cana-4362	250	8	filter	filter	NOUN
cana-4362	250	9	stage	stage	NOUN
cana-4362	250	10	,	,	PUNCT
cana-4362	250	11	wrapper	wrapper	NOUN
cana-4362	250	12	methods	method	NOUN
cana-4362	250	13	(	(	PUNCT
cana-4362	250	14	such	such	ADJ
cana-4362	250	15	as	as	ADP
cana-4362	250	16	recursive	recursive	ADJ
cana-4362	250	17	feature	feature	NOUN
cana-4362	250	18	elimination	elimination	NOUN
cana-4362	250	19	or	or	CCONJ
cana-4362	250	20	genetic	genetic	ADJ
cana-4362	250	21	algorithms	algorithm	NOUN
cana-4362	250	22	)	)	PUNCT
cana-4362	250	23	are	be	AUX
cana-4362	250	24	applied	apply	VERB
cana-4362	250	25	to	to	PART
cana-4362	250	26	refine	refine	VERB
cana-4362	250	27	the	the	DET
cana-4362	250	28	feature	feature	NOUN
cana-4362	250	29	subset	subset	VERB
cana-4362	250	30	.	.	PUNCT
cana-4362	251	1	these	these	DET
cana-4362	251	2	methods	method	NOUN
cana-4362	251	3	evaluate	evaluate	VERB
cana-4362	251	4	subsets	subset	NOUN
cana-4362	251	5	of	of	ADP
cana-4362	251	6	features	feature	NOUN
cana-4362	251	7	by	by	ADP
cana-4362	251	8	training	train	VERB
cana-4362	251	9	a	a	DET
cana-4362	251	10	model	model	NOUN
cana-4362	251	11	and	and	CCONJ
cana-4362	251	12	assessing	assess	VERB
cana-4362	251	13	its	its	PRON
cana-4362	251	14	performance	performance	NOUN
cana-4362	251	15	,	,	PUNCT
cana-4362	251	16	iteratively	iteratively	ADV
cana-4362	251	17	selecting	select	VERB
cana-4362	251	18	the	the	DET
cana-4362	251	19	best	good	ADJ
cana-4362	251	20	combination	combination	NOUN
cana-4362	251	21	of	of	ADP
cana-4362	251	22	features	feature	NOUN
cana-4362	251	23	that	that	PRON
cana-4362	251	24	enhances	enhance	VERB
cana-4362	251	25	model	model	NOUN
cana-4362	251	26	accuracy	accuracy	NOUN
cana-4362	251	27	.	.	PUNCT
cana-4362	252	1	deep	deep	ADJ
cana-4362	252	2	learning	learn	VERB
cana-4362	252	3	feature	feature	NOUN
cana-4362	252	4	extraction	extraction	NOUN
cana-4362	252	5	:	:	PUNCT
cana-4362	252	6	in	in	ADP
cana-4362	252	7	addition	addition	NOUN
cana-4362	252	8	to	to	ADP
cana-4362	252	9	traditional	traditional	ADJ
cana-4362	252	10	methods	method	NOUN
cana-4362	252	11	,	,	PUNCT
cana-4362	252	12	the	the	DET
cana-4362	252	13	architecture	architecture	NOUN
cana-4362	252	14	incorporates	incorporate	VERB
cana-4362	252	15	deep	deep	ADJ
cana-4362	252	16	learning	learning	NOUN
cana-4362	252	17	models	model	NOUN
cana-4362	252	18	such	such	ADJ
cana-4362	252	19	as	as	ADP
cana-4362	252	20	convolutional	convolutional	ADJ
cana-4362	252	21	neural	neural	ADJ
cana-4362	252	22	networks	network	NOUN
cana-4362	252	23	(	(	PUNCT
cana-4362	252	24	cnns	cnns	PROPN
cana-4362	252	25	)	)	PUNCT
cana-4362	252	26	or	or	CCONJ
cana-4362	252	27	long	long	ADJ
cana-4362	252	28	short	short	ADJ
cana-4362	252	29	-	-	PUNCT
cana-4362	252	30	term	term	NOUN
cana-4362	252	31	memory	memory	NOUN
cana-4362	252	32	(	(	PUNCT
cana-4362	252	33	lstm	lstm	NOUN
cana-4362	252	34	)	)	PUNCT
cana-4362	252	35	networks	network	NOUN
cana-4362	252	36	.	.	PUNCT
cana-4362	253	1	these	these	DET
cana-4362	253	2	models	model	NOUN
cana-4362	253	3	automatically	automatically	ADV
cana-4362	253	4	learn	learn	VERB
cana-4362	253	5	feature	feature	NOUN
cana-4362	253	6	representations	representation	NOUN
cana-4362	253	7	from	from	ADP
cana-4362	253	8	the	the	DET
cana-4362	253	9	data	datum	NOUN
cana-4362	253	10	,	,	PUNCT
cana-4362	253	11	identifying	identify	VERB
cana-4362	253	12	complex	complex	ADJ
cana-4362	253	13	patterns	pattern	NOUN
cana-4362	253	14	and	and	CCONJ
cana-4362	253	15	relationships	relationship	NOUN
cana-4362	253	16	that	that	SCONJ
cana-4362	253	17	traditional	traditional	ADJ
cana-4362	253	18	methods	method	NOUN
cana-4362	253	19	may	may	AUX
cana-4362	253	20	miss	miss	VERB
cana-4362	253	21	.	.	PUNCT
cana-4362	254	1	the	the	DET
cana-4362	254	2	hybrid	hybrid	ADJ
cana-4362	254	3	feature	feature	NOUN
cana-4362	254	4	selection	selection	NOUN
cana-4362	254	5	combines	combine	VERB
cana-4362	254	6	insights	insight	NOUN
cana-4362	254	7	from	from	ADP
cana-4362	254	8	all	all	DET
cana-4362	254	9	these	these	DET
cana-4362	254	10	approaches	approach	NOUN
cana-4362	254	11	,	,	PUNCT
cana-4362	254	12	ensuring	ensure	VERB
cana-4362	254	13	a	a	DET
cana-4362	254	14	comprehensive	comprehensive	ADJ
cana-4362	254	15	selection	selection	NOUN
cana-4362	254	16	of	of	ADP
cana-4362	254	17	relevant	relevant	ADJ
cana-4362	254	18	features	feature	NOUN
cana-4362	254	19	.	.	PUNCT
cana-4362	255	1	communications	communication	NOUN
cana-4362	255	2	on	on	ADP
cana-4362	255	3	applied	apply	VERB
cana-4362	255	4	nonlinear	nonlinear	ADJ
cana-4362	255	5	analysis	analysis	NOUN
cana-4362	255	6	issn	issn	NOUN
cana-4362	255	7	:	:	PUNCT
cana-4362	255	8	1074	1074	NUM
cana-4362	255	9	-	-	PUNCT
cana-4362	255	10	133x	133x	NUM
cana-4362	255	11	vol	vol	NOUN
cana-4362	255	12	32	32	NUM
cana-4362	255	13	no	no	NOUN
cana-4362	255	14	.	.	PUNCT
cana-4362	256	1	9s	9s	NUM
cana-4362	256	2	(	(	PUNCT
cana-4362	256	3	2025	2025	NUM
cana-4362	256	4	)	)	PUNCT
cana-4362	256	5	1909	1909	NUM
cana-4362	256	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-4362	256	7	4	4	X
cana-4362	256	8	.	.	NOUN
cana-4362	256	9	sentiment	sentiment	NOUN
cana-4362	256	10	classification	classification	NOUN
cana-4362	256	11	module	module	NOUN
cana-4362	256	12	after	after	ADP
cana-4362	256	13	feature	feature	NOUN
cana-4362	256	14	selection	selection	NOUN
cana-4362	256	15	,	,	PUNCT
cana-4362	256	16	the	the	DET
cana-4362	256	17	refined	refined	ADJ
cana-4362	256	18	feature	feature	NOUN
cana-4362	256	19	set	set	NOUN
cana-4362	256	20	is	be	AUX
cana-4362	256	21	fed	feed	VERB
cana-4362	256	22	into	into	ADP
cana-4362	256	23	the	the	DET
cana-4362	256	24	sentiment	sentiment	NOUN
cana-4362	256	25	classification	classification	NOUN
cana-4362	256	26	module	module	NOUN
cana-4362	256	27	.	.	PUNCT
cana-4362	257	1	this	this	DET
cana-4362	257	2	module	module	NOUN
cana-4362	257	3	utilizes	utilize	VERB
cana-4362	257	4	deep	deep	ADJ
cana-4362	257	5	learning	learning	NOUN
cana-4362	257	6	techniques	technique	NOUN
cana-4362	257	7	,	,	PUNCT
cana-4362	257	8	particularly	particularly	ADV
cana-4362	257	9	cnns	cnns	ADJ
cana-4362	257	10	and	and	CCONJ
cana-4362	257	11	lstms	lstms	ADJ
cana-4362	257	12	,	,	PUNCT
cana-4362	257	13	to	to	PART
cana-4362	257	14	classify	classify	VERB
cana-4362	257	15	sentiments	sentiment	NOUN
cana-4362	257	16	expressed	express	VERB
cana-4362	257	17	in	in	ADP
cana-4362	257	18	the	the	DET
cana-4362	257	19	social	social	ADJ
cana-4362	257	20	media	medium	NOUN
cana-4362	257	21	text	text	NOUN
cana-4362	257	22	as	as	ADP
cana-4362	257	23	positive	positive	ADJ
cana-4362	257	24	,	,	PUNCT
cana-4362	257	25	negative	negative	ADJ
cana-4362	257	26	,	,	PUNCT
cana-4362	257	27	or	or	CCONJ
cana-4362	257	28	neutral	neutral	ADJ
cana-4362	257	29	.	.	PUNCT
cana-4362	258	1	the	the	DET
cana-4362	258	2	architecture	architecture	NOUN
cana-4362	258	3	typically	typically	ADV
cana-4362	258	4	involves	involve	VERB
cana-4362	258	5	:	:	PUNCT
cana-4362	258	6	embedding	embed	VERB
cana-4362	258	7	layer	layer	NOUN
cana-4362	258	8	:	:	PUNCT
cana-4362	258	9	converts	convert	VERB
cana-4362	258	10	selected	select	VERB
cana-4362	258	11	features	feature	NOUN
cana-4362	258	12	(	(	PUNCT
cana-4362	258	13	words	word	NOUN
cana-4362	258	14	or	or	CCONJ
cana-4362	258	15	phrases	phrase	NOUN
cana-4362	258	16	)	)	PUNCT
cana-4362	258	17	into	into	ADP
cana-4362	258	18	dense	dense	ADJ
cana-4362	258	19	vectors	vector	NOUN
cana-4362	258	20	that	that	PRON
cana-4362	258	21	capture	capture	VERB
cana-4362	258	22	semantic	semantic	ADJ
cana-4362	258	23	information	information	NOUN
cana-4362	258	24	.	.	PUNCT
cana-4362	259	1	convolutional	convolutional	ADJ
cana-4362	259	2	and	and	CCONJ
cana-4362	259	3	pooling	pool	VERB
cana-4362	259	4	layers	layer	NOUN
cana-4362	259	5	(	(	PUNCT
cana-4362	259	6	for	for	ADP
cana-4362	259	7	cnns	cnn	NOUN
cana-4362	259	8	):	):	PUNCT
cana-4362	259	9	these	these	DET
cana-4362	259	10	layers	layer	NOUN
cana-4362	259	11	are	be	AUX
cana-4362	259	12	employed	employ	VERB
cana-4362	259	13	to	to	PART
cana-4362	259	14	capture	capture	VERB
cana-4362	259	15	local	local	ADJ
cana-4362	259	16	patterns	pattern	NOUN
cana-4362	259	17	and	and	CCONJ
cana-4362	259	18	relationships	relationship	NOUN
cana-4362	259	19	in	in	ADP
cana-4362	259	20	the	the	DET
cana-4362	259	21	text	text	NOUN
cana-4362	259	22	data	datum	NOUN
cana-4362	259	23	.	.	PUNCT
cana-4362	260	1	recurrent	recurrent	ADJ
cana-4362	260	2	layers	layer	NOUN
cana-4362	260	3	(	(	PUNCT
cana-4362	260	4	for	for	ADP
cana-4362	260	5	lstms	lstms	ADJ
cana-4362	260	6	):	):	PUNCT
cana-4362	260	7	capture	capture	VERB
cana-4362	260	8	sequential	sequential	ADJ
cana-4362	260	9	dependencies	dependency	NOUN
cana-4362	260	10	and	and	CCONJ
cana-4362	260	11	long	long	ADJ
cana-4362	260	12	-	-	PUNCT
cana-4362	260	13	term	term	NOUN
cana-4362	260	14	relationships	relationship	NOUN
cana-4362	260	15	within	within	ADP
cana-4362	260	16	the	the	DET
cana-4362	260	17	text	text	NOUN
cana-4362	260	18	.	.	PUNCT
cana-4362	261	1	fully	fully	ADV
cana-4362	261	2	connected	connect	VERB
cana-4362	261	3	layer	layer	NOUN
cana-4362	261	4	:	:	PUNCT
cana-4362	261	5	combines	combine	VERB
cana-4362	261	6	the	the	DET
cana-4362	261	7	outputs	output	NOUN
cana-4362	261	8	from	from	ADP
cana-4362	261	9	the	the	DET
cana-4362	261	10	previous	previous	ADJ
cana-4362	261	11	layers	layer	NOUN
cana-4362	261	12	and	and	CCONJ
cana-4362	261	13	generates	generate	VERB
cana-4362	261	14	the	the	DET
cana-4362	261	15	final	final	ADJ
cana-4362	261	16	sentiment	sentiment	NOUN
cana-4362	261	17	predictions	prediction	NOUN
cana-4362	261	18	.	.	PUNCT
cana-4362	262	1	5	5	X
cana-4362	262	2	.	.	X
cana-4362	262	3	evaluation	evaluation	NOUN
cana-4362	262	4	module	module	NOUN
cana-4362	262	5	the	the	DET
cana-4362	262	6	evaluation	evaluation	NOUN
cana-4362	262	7	module	module	NOUN
cana-4362	262	8	assesses	assess	VERB
cana-4362	262	9	the	the	DET
cana-4362	262	10	performance	performance	NOUN
cana-4362	262	11	of	of	ADP
cana-4362	262	12	the	the	DET
cana-4362	262	13	sentiment	sentiment	NOUN
cana-4362	262	14	classification	classification	NOUN
cana-4362	262	15	model	model	NOUN
cana-4362	262	16	using	use	VERB
cana-4362	262	17	metrics	metric	NOUN
cana-4362	262	18	such	such	ADJ
cana-4362	262	19	as	as	ADP
cana-4362	262	20	accuracy	accuracy	NOUN
cana-4362	262	21	,	,	PUNCT
cana-4362	262	22	precision	precision	NOUN
cana-4362	262	23	,	,	PUNCT
cana-4362	262	24	recall	recall	NOUN
cana-4362	262	25	,	,	PUNCT
cana-4362	262	26	and	and	CCONJ
cana-4362	262	27	f1	f1	NOUN
cana-4362	262	28	-	-	PUNCT
cana-4362	262	29	score	score	NOUN
cana-4362	262	30	.	.	PUNCT
cana-4362	263	1	this	this	DET
cana-4362	263	2	feedback	feedback	NOUN
cana-4362	263	3	loop	loop	NOUN
cana-4362	263	4	helps	help	VERB
cana-4362	263	5	refine	refine	VERB
cana-4362	263	6	the	the	DET
cana-4362	263	7	feature	feature	NOUN
cana-4362	263	8	selection	selection	NOUN
cana-4362	263	9	process	process	NOUN
cana-4362	263	10	and	and	CCONJ
cana-4362	263	11	model	model	NOUN
cana-4362	263	12	training	training	NOUN
cana-4362	263	13	,	,	PUNCT
cana-4362	263	14	ensuring	ensure	VERB
cana-4362	263	15	continuous	continuous	ADJ
cana-4362	263	16	improvement	improvement	NOUN
cana-4362	263	17	in	in	ADP
cana-4362	263	18	classification	classification	NOUN
cana-4362	263	19	performance	performance	NOUN
cana-4362	263	20	.	.	PUNCT
cana-4362	264	1	the	the	DET
cana-4362	264	2	final	final	ADJ
cana-4362	264	3	component	component	NOUN
cana-4362	264	4	of	of	ADP
cana-4362	264	5	the	the	DET
cana-4362	264	6	architecture	architecture	NOUN
cana-4362	264	7	is	be	AUX
cana-4362	264	8	the	the	DET
cana-4362	264	9	user	user	NOUN
cana-4362	264	10	interface	interface	NOUN
cana-4362	264	11	,	,	PUNCT
cana-4362	264	12	which	which	PRON
cana-4362	264	13	presents	present	VERB
cana-4362	264	14	the	the	DET
cana-4362	264	15	results	result	NOUN
cana-4362	264	16	of	of	ADP
cana-4362	264	17	sentiment	sentiment	NOUN
cana-4362	264	18	analysis	analysis	NOUN
cana-4362	264	19	in	in	ADP
cana-4362	264	20	a	a	DET
cana-4362	264	21	user	user	NOUN
cana-4362	264	22	-	-	PUNCT
cana-4362	264	23	friendly	friendly	ADJ
cana-4362	264	24	manner	manner	NOUN
cana-4362	264	25	.	.	PUNCT
cana-4362	265	1	this	this	PRON
cana-4362	265	2	may	may	AUX
cana-4362	265	3	include	include	VERB
cana-4362	265	4	visualizations	visualization	NOUN
cana-4362	265	5	such	such	ADJ
cana-4362	265	6	as	as	ADP
cana-4362	265	7	sentiment	sentiment	NOUN
cana-4362	265	8	trends	trend	NOUN
cana-4362	265	9	,	,	PUNCT
cana-4362	265	10	word	word	NOUN
cana-4362	265	11	clouds	cloud	NOUN
cana-4362	265	12	,	,	PUNCT
cana-4362	265	13	and	and	CCONJ
cana-4362	265	14	dashboards	dashboard	NOUN
cana-4362	265	15	that	that	PRON
cana-4362	265	16	allow	allow	VERB
cana-4362	265	17	users	user	NOUN
cana-4362	265	18	to	to	PART
cana-4362	265	19	gain	gain	VERB
cana-4362	265	20	insights	insight	NOUN
cana-4362	265	21	from	from	ADP
cana-4362	265	22	the	the	DET
cana-4362	265	23	sentiment	sentiment	NOUN
cana-4362	265	24	analysis	analysis	NOUN
cana-4362	265	25	conducted	conduct	VERB
cana-4362	265	26	on	on	ADP
cana-4362	265	27	social	social	ADJ
cana-4362	265	28	media	medium	NOUN
cana-4362	265	29	data	datum	NOUN
cana-4362	265	30	.	.	PUNCT
cana-4362	266	1	the	the	DET
cana-4362	266	2	system	system	NOUN
cana-4362	266	3	architecture	architecture	NOUN
cana-4362	266	4	for	for	ADP
cana-4362	266	5	hybrid	hybrid	ADJ
cana-4362	266	6	feature	feature	NOUN
cana-4362	266	7	selection	selection	NOUN
cana-4362	266	8	on	on	ADP
cana-4362	266	9	social	social	ADJ
cana-4362	266	10	media	medium	NOUN
cana-4362	266	11	datasets	dataset	NOUN
cana-4362	266	12	for	for	ADP
cana-4362	266	13	sentiment	sentiment	NOUN
cana-4362	266	14	classification	classification	NOUN
cana-4362	266	15	using	use	VERB
cana-4362	266	16	deep	deep	ADJ
cana-4362	266	17	learning	learning	NOUN
cana-4362	266	18	techniques	technique	NOUN
cana-4362	266	19	consists	consist	VERB
cana-4362	266	20	of	of	ADP
cana-4362	266	21	several	several	ADJ
cana-4362	266	22	interconnected	interconnected	ADJ
cana-4362	266	23	modules	module	NOUN
cana-4362	266	24	.	.	PUNCT
cana-4362	267	1	by	by	ADP
cana-4362	267	2	integrating	integrate	VERB
cana-4362	267	3	data	datum	NOUN
cana-4362	267	4	collection	collection	NOUN
cana-4362	267	5	,	,	PUNCT
cana-4362	267	6	preprocessing	preprocessing	NOUN
cana-4362	267	7	,	,	PUNCT
cana-4362	267	8	hybrid	hybrid	ADJ
cana-4362	267	9	feature	feature	NOUN
cana-4362	267	10	selection	selection	NOUN
cana-4362	267	11	,	,	PUNCT
cana-4362	267	12	deep	deep	ADJ
cana-4362	267	13	learning	learn	VERB
cana-4362	267	14	classification	classification	NOUN
cana-4362	267	15	,	,	PUNCT
cana-4362	267	16	evaluation	evaluation	NOUN
cana-4362	267	17	,	,	PUNCT
cana-4362	267	18	and	and	CCONJ
cana-4362	267	19	user	user	NOUN
cana-4362	267	20	interface	interface	NOUN
cana-4362	267	21	components	component	NOUN
cana-4362	267	22	,	,	PUNCT
cana-4362	267	23	this	this	DET
cana-4362	267	24	architecture	architecture	NOUN
cana-4362	267	25	provides	provide	VERB
cana-4362	267	26	an	an	DET
cana-4362	267	27	efficient	efficient	ADJ
cana-4362	267	28	and	and	CCONJ
cana-4362	267	29	effective	effective	ADJ
cana-4362	267	30	framework	framework	NOUN
cana-4362	267	31	for	for	ADP
cana-4362	267	32	analyzing	analyze	VERB
cana-4362	267	33	and	and	CCONJ
cana-4362	267	34	interpreting	interpret	VERB
cana-4362	267	35	sentiments	sentiment	NOUN
cana-4362	267	36	expressed	express	VERB
cana-4362	267	37	in	in	ADP
cana-4362	267	38	vast	vast	ADJ
cana-4362	267	39	amounts	amount	NOUN
cana-4362	267	40	of	of	ADP
cana-4362	267	41	social	social	ADJ
cana-4362	267	42	media	medium	NOUN
cana-4362	267	43	data	datum	NOUN
cana-4362	267	44	.	.	PUNCT
cana-4362	268	1	this	this	DET
cana-4362	268	2	comprehensive	comprehensive	ADJ
cana-4362	268	3	approach	approach	NOUN
cana-4362	268	4	ensures	ensure	VERB
cana-4362	268	5	accurate	accurate	ADJ
cana-4362	268	6	sentiment	sentiment	NOUN
cana-4362	268	7	classification	classification	NOUN
cana-4362	268	8	while	while	SCONJ
cana-4362	268	9	managing	manage	VERB
cana-4362	268	10	the	the	DET
cana-4362	268	11	complexities	complexity	NOUN
cana-4362	268	12	inherent	inherent	ADJ
cana-4362	268	13	in	in	ADP
cana-4362	268	14	social	social	ADJ
cana-4362	268	15	media	medium	NOUN
cana-4362	268	16	datasets	dataset	NOUN
cana-4362	268	17	.	.	PUNCT
cana-4362	269	1	algorithm	algorithm	PROPN
cana-4362	269	2	design	design	NOUN
cana-4362	269	3	results	result	NOUN
cana-4362	269	4	and	and	CCONJ
cana-4362	269	5	discussion	discussion	VERB
cana-4362	269	6	the	the	DET
cana-4362	269	7	effectiveness	effectiveness	NOUN
cana-4362	269	8	of	of	ADP
cana-4362	269	9	hybrid	hybrid	ADJ
cana-4362	269	10	feature	feature	NOUN
cana-4362	269	11	selection	selection	NOUN
cana-4362	269	12	methods	method	NOUN
cana-4362	269	13	integrated	integrate	VERB
cana-4362	269	14	with	with	ADP
cana-4362	269	15	deep	deep	ADJ
cana-4362	269	16	learning	learning	NOUN
cana-4362	269	17	techniques	technique	NOUN
cana-4362	269	18	was	be	AUX
cana-4362	269	19	assessed	assess	VERB
cana-4362	269	20	through	through	ADP
cana-4362	269	21	metrics	metric	NOUN
cana-4362	269	22	like	like	ADP
cana-4362	269	23	accuracy	accuracy	NOUN
cana-4362	269	24	,	,	PUNCT
cana-4362	269	25	precision	precision	NOUN
cana-4362	269	26	,	,	PUNCT
cana-4362	269	27	recall	recall	NOUN
cana-4362	269	28	,	,	PUNCT
cana-4362	269	29	and	and	CCONJ
cana-4362	269	30	f1	f1	NOUN
cana-4362	269	31	-	-	PUNCT
cana-4362	269	32	score	score	NOUN
cana-4362	269	33	.	.	PUNCT
cana-4362	270	1	to	to	PART
cana-4362	270	2	validate	validate	VERB
cana-4362	270	3	our	our	PRON
cana-4362	270	4	proposed	propose	VERB
cana-4362	270	5	bug	bug	NOUN
cana-4362	270	6	forecasting	forecasting	NOUN
cana-4362	270	7	approach	approach	NOUN
cana-4362	270	8	,	,	PUNCT
cana-4362	270	9	we	we	PRON
cana-4362	270	10	selected	select	VERB
cana-4362	270	11	rnn	rnn	NOUN
cana-4362	270	12	classification	classification	NOUN
cana-4362	270	13	algorithms	algorithm	NOUN
cana-4362	270	14	for	for	ADP
cana-4362	270	15	fault	fault	NOUN
cana-4362	270	16	prediction	prediction	NOUN
cana-4362	270	17	,	,	PUNCT
cana-4362	270	18	even	even	ADV
cana-4362	270	19	incorporating	incorporate	VERB
cana-4362	270	20	unlabelled	unlabelled	ADJ
cana-4362	270	21	datasets	dataset	NOUN
cana-4362	270	22	.	.	PUNCT
cana-4362	271	1	as	as	SCONJ
cana-4362	271	2	shown	show	VERB
cana-4362	271	3	in	in	ADP
cana-4362	271	4	table	table	NOUN
cana-4362	271	5	1	1	NUM
cana-4362	271	6	,	,	PUNCT
cana-4362	271	7	the	the	DET
cana-4362	271	8	data	data	NOUN
cana-4362	271	9	distribution	distribution	NOUN
cana-4362	271	10	is	be	AUX
cana-4362	271	11	used	use	VERB
cana-4362	271	12	to	to	PART
cana-4362	271	13	evaluate	evaluate	VERB
cana-4362	271	14	the	the	DET
cana-4362	271	15	capabilities	capability	NOUN
cana-4362	271	16	of	of	ADP
cana-4362	271	17	different	different	ADJ
cana-4362	271	18	machine	machine	NOUN
cana-4362	271	19	learning	learning	NOUN
cana-4362	271	20	models	model	NOUN
cana-4362	271	21	,	,	PUNCT
cana-4362	271	22	presenting	present	VERB
cana-4362	271	23	measurements	measurement	NOUN
cana-4362	271	24	for	for	ADP
cana-4362	271	25	accuracy	accuracy	NOUN
cana-4362	271	26	,	,	PUNCT
cana-4362	271	27	recall	recall	NOUN
cana-4362	271	28	,	,	PUNCT
cana-4362	271	29	and	and	CCONJ
cana-4362	271	30	f1	f1	NOUN
cana-4362	271	31	-	-	PUNCT
cana-4362	271	32	score	score	NOUN
cana-4362	271	33	.	.	PUNCT
cana-4362	272	1	precision	precision	NOUN
cana-4362	272	2	=	=	SYM
cana-4362	272	3	communications	communication	NOUN
cana-4362	272	4	on	on	ADP
cana-4362	272	5	applied	apply	VERB
cana-4362	272	6	nonlinear	nonlinear	ADJ
cana-4362	272	7	analysis	analysis	NOUN
cana-4362	272	8	issn	issn	NOUN
cana-4362	272	9	:	:	PUNCT
cana-4362	272	10	1074	1074	NUM
cana-4362	272	11	-	-	PUNCT
cana-4362	272	12	133x	133x	NUM
cana-4362	272	13	vol	vol	NOUN
cana-4362	272	14	32	32	NUM
cana-4362	272	15	no	no	NOUN
cana-4362	272	16	.	.	PUNCT
cana-4362	273	1	9s	9s	NUM
cana-4362	273	2	(	(	PUNCT
cana-4362	273	3	2025	2025	NUM
cana-4362	273	4	)	)	PUNCT
cana-4362	273	5	1910	1910	NUM
cana-4362	273	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-4362	273	7	recall	recall	NOUN
cana-4362	273	8	=	=	SYM
cana-4362	273	9	f	f	PROPN
cana-4362	273	10	–	–	PUNCT
cana-4362	273	11	measure	measure	NOUN
cana-4362	273	12	=	=	NOUN
cana-4362	273	13	the	the	DET
cana-4362	273	14	system	system	NOUN
cana-4362	273	15	's	's	PART
cana-4362	273	16	performance	performance	NOUN
cana-4362	273	17	evaluation	evaluation	NOUN
cana-4362	273	18	was	be	AUX
cana-4362	273	19	conducted	conduct	VERB
cana-4362	273	20	using	use	VERB
cana-4362	273	21	real	real	ADJ
cana-4362	273	22	-	-	PUNCT
cana-4362	273	23	life	life	NOUN
cana-4362	273	24	social	social	ADJ
cana-4362	273	25	media	medium	NOUN
cana-4362	273	26	datasets	dataset	NOUN
cana-4362	273	27	and	and	CCONJ
cana-4362	273	28	the	the	DET
cana-4362	273	29	proposed	propose	VERB
cana-4362	273	30	classification	classification	NOUN
cana-4362	273	31	techniques	technique	NOUN
cana-4362	273	32	.	.	PUNCT
cana-4362	274	1	table	table	NOUN
cana-4362	274	2	1	1	NUM
cana-4362	274	3	below	below	ADV
cana-4362	274	4	illustrates	illustrate	VERB
cana-4362	274	5	the	the	DET
cana-4362	274	6	number	number	NOUN
cana-4362	274	7	of	of	ADP
cana-4362	274	8	bug	bug	NOUN
cana-4362	274	9	instances	instance	NOUN
cana-4362	274	10	identified	identify	VERB
cana-4362	274	11	within	within	ADP
cana-4362	274	12	each	each	DET
cana-4362	274	13	dataset	dataset	NOUN
cana-4362	274	14	,	,	PUNCT
cana-4362	274	15	along	along	ADP
cana-4362	274	16	with	with	ADP
cana-4362	274	17	the	the	DET
cana-4362	274	18	percentage	percentage	NOUN
cana-4362	274	19	of	of	ADP
cana-4362	274	20	bug	bug	NOUN
cana-4362	274	21	records	record	NOUN
cana-4362	274	22	for	for	ADP
cana-4362	274	23	each	each	DET
cana-4362	274	24	one	one	NUM
cana-4362	274	25	.	.	PUNCT
cana-4362	275	1	table	table	NOUN
cana-4362	275	2	1	1	NUM
cana-4362	275	3	:	:	PUNCT
cana-4362	275	4	performance	performance	NOUN
cana-4362	275	5	evaluation	evaluation	NOUN
cana-4362	275	6	using	use	VERB
cana-4362	275	7	various	various	ADJ
cana-4362	275	8	learning	learn	VERB
cana-4362	275	9	algorithm	algorithm	NOUN
cana-4362	275	10	for	for	ADP
cana-4362	275	11	proposed	propose	VERB
cana-4362	275	12	model	model	NOUN
cana-4362	275	13	dataset	dataset	VERB
cana-4362	275	14	epoch	epoch	PROPN
cana-4362	275	15	accuracy	accuracy	PROPN
cana-4362	275	16	precision	precision	NOUN
cana-4362	275	17	recall	recall	VERB
cana-4362	275	18	f	f	NOUN
cana-4362	275	19	-	-	PUNCT
cana-4362	275	20	score	score	NOUN
cana-4362	275	21	rnn	rnn	VERB
cana-4362	275	22	25	25	NUM
cana-4362	275	23	0.91	0.91	NUM
cana-4362	275	24	0.93	0.93	NUM
cana-4362	275	25	0.95	0.95	NUM
cana-4362	275	26	0.95	0.95	NUM
cana-4362	275	27	50	50	NUM
cana-4362	275	28	0.92	0.92	NUM
cana-4362	275	29	0.93	0.93	NUM
cana-4362	275	30	0.95	0.95	NUM
cana-4362	275	31	0.95	0.95	NUM
cana-4362	275	32	75	75	NUM
cana-4362	275	33	0.93	0.93	NUM
cana-4362	275	34	0.94	0.94	NUM
cana-4362	275	35	0.96	0.96	NUM
cana-4362	275	36	0.96	0.96	NUM
cana-4362	275	37	100	100	NUM
cana-4362	275	38	0.95	0.95	NUM
cana-4362	275	39	0.94	0.94	NUM
cana-4362	275	40	0.95	0.95	NUM
cana-4362	275	41	0.96	0.96	NUM
cana-4362	275	42	lstm	lstm	NOUN
cana-4362	275	43	25	25	NUM
cana-4362	276	1	0.90	0.90	NUM
cana-4362	276	2	0.91	0.91	NUM
cana-4362	276	3	0.92	0.92	NUM
cana-4362	276	4	0.94	0.94	NUM
cana-4362	276	5	50	50	NUM
cana-4362	276	6	0.90	0.90	NUM
cana-4362	276	7	0.92	0.92	NUM
cana-4362	276	8	0.93	0.93	NUM
cana-4362	276	9	0.95	0.95	NUM
cana-4362	276	10	75	75	NUM
cana-4362	276	11	0.94	0.94	NUM
cana-4362	276	12	0.93	0.93	NUM
cana-4362	276	13	0.94	0.94	NUM
cana-4362	276	14	0.96	0.96	NUM
cana-4362	276	15	100	100	NUM
cana-4362	276	16	0.96	0.96	NUM
cana-4362	276	17	0.95	0.95	NUM
cana-4362	276	18	0.96	0.96	NUM
cana-4362	276	19	0.97	0.97	NUM
cana-4362	276	20	rnn	rnn	NOUN
cana-4362	276	21	-	-	PUNCT
cana-4362	276	22	lstm	lstm	ADJ
cana-4362	276	23	25	25	NUM
cana-4362	276	24	0.97	0.97	NUM
cana-4362	276	25	0.98	0.98	NUM
cana-4362	276	26	0.98	0.98	NUM
cana-4362	276	27	0.99	0.99	NUM
cana-4362	276	28	50	50	NUM
cana-4362	276	29	0.92	0.92	NUM
cana-4362	276	30	0.91	0.91	NUM
cana-4362	276	31	0.89	0.89	NUM
cana-4362	276	32	0.90	0.90	NUM
cana-4362	276	33	75	75	NUM
cana-4362	276	34	0.89	0.89	NUM
cana-4362	276	35	0.88	0.88	NUM
cana-4362	276	36	0.93	0.93	NUM
cana-4362	276	37	0.91	0.91	NUM
cana-4362	276	38	100	100	NUM
cana-4362	276	39	0.98	0.98	NUM
cana-4362	276	40	0.96	0.96	NUM
cana-4362	276	41	0.96	0.96	NUM
cana-4362	276	42	0.97	0.97	NUM
cana-4362	276	43	figure	figure	NOUN
cana-4362	276	44	2	2	NUM
cana-4362	276	45	:	:	PUNCT
cana-4362	276	46	classification	classification	NOUN
cana-4362	276	47	accuracy	accuracy	NOUN
cana-4362	276	48	for	for	ADP
cana-4362	276	49	proposed	propose	VERB
cana-4362	276	50	model	model	NOUN
cana-4362	276	51	using	use	VERB
cana-4362	276	52	rnn	rnn	VERB
cana-4362	276	53	the	the	DET
cana-4362	276	54	above	above	ADJ
cana-4362	276	55	figure	figure	NOUN
cana-4362	276	56	2	2	NUM
cana-4362	276	57	describes	describe	VERB
cana-4362	276	58	classification	classification	NOUN
cana-4362	276	59	accuracy	accuracy	NOUN
cana-4362	276	60	using	use	VERB
cana-4362	276	61	rnn	rnn	NOUN
cana-4362	276	62	algorithm	algorithm	NOUN
cana-4362	276	63	.	.	PUNCT
cana-4362	277	1	probably	probably	ADV
cana-4362	277	2	the	the	DET
cana-4362	277	3	100	100	NUM
cana-4362	277	4	epoch	epoch	NOUN
cana-4362	277	5	demonstrates	demonstrate	VERB
cana-4362	277	6	higher	high	ADJ
cana-4362	277	7	accuracy	accuracy	NOUN
cana-4362	277	8	on	on	ADP
cana-4362	277	9	real	real	ADJ
cana-4362	277	10	time	time	NOUN
cana-4362	277	11	text	text	NOUN
cana-4362	277	12	data	datum	NOUN
cana-4362	277	13	.	.	PUNCT
cana-4362	278	1	the	the	DET
cana-4362	278	2	higher	high	ADJ
cana-4362	278	3	accuracy	accuracy	NOUN
cana-4362	278	4	0.95	0.95	NUM
cana-4362	278	5	%	%	NOUN
cana-4362	278	6	on	on	ADP
cana-4362	278	7	real	real	ADJ
cana-4362	278	8	time	time	NOUN
cana-4362	278	9	social	social	ADJ
cana-4362	278	10	media	medium	NOUN
cana-4362	278	11	dataset	dataset	NOUN
cana-4362	278	12	.	.	PUNCT
cana-4362	279	1	0.88	0.88	NUM
cana-4362	279	2	0.89	0.89	NUM
cana-4362	279	3	0.9	0.9	NUM
cana-4362	279	4	0.91	0.91	NUM
cana-4362	279	5	0.92	0.92	NUM
cana-4362	279	6	0.93	0.93	NUM
cana-4362	279	7	0.94	0.94	NUM
cana-4362	279	8	0.95	0.95	NUM
cana-4362	279	9	0.96	0.96	NUM
cana-4362	279	10	0.97	0.97	NUM
cana-4362	279	11	25	25	NUM
cana-4362	279	12	50	50	NUM
cana-4362	279	13	75	75	NUM
cana-4362	279	14	100	100	NUM
cana-4362	279	15	va	va	NOUN
cana-4362	279	16	lu	lu	PROPN
cana-4362	279	17	e	e	PROPN
cana-4362	279	18	epoch	epoch	NOUN
cana-4362	279	19	size	size	NOUN
cana-4362	279	20	accuracy	accuracy	NOUN
cana-4362	279	21	precision	precision	NOUN
cana-4362	279	22	recall	recall	VERB
cana-4362	279	23	f	f	NOUN
cana-4362	279	24	-	-	PUNCT
cana-4362	279	25	score	score	NOUN
cana-4362	279	26	communications	communication	NOUN
cana-4362	279	27	on	on	ADP
cana-4362	279	28	applied	apply	VERB
cana-4362	279	29	nonlinear	nonlinear	ADJ
cana-4362	279	30	analysis	analysis	NOUN
cana-4362	279	31	issn	issn	NOUN
cana-4362	279	32	:	:	PUNCT
cana-4362	279	33	1074	1074	NUM
cana-4362	279	34	-	-	PUNCT
cana-4362	279	35	133x	133x	NUM
cana-4362	279	36	vol	vol	NOUN
cana-4362	279	37	32	32	NUM
cana-4362	279	38	no	no	NOUN
cana-4362	279	39	.	.	PUNCT
cana-4362	280	1	9s	9s	NUM
cana-4362	280	2	(	(	PUNCT
cana-4362	280	3	2025	2025	NUM
cana-4362	280	4	)	)	PUNCT
cana-4362	280	5	1911	1911	NUM
cana-4362	281	1	https://internationalpubls.com	https://internationalpubls.com	X
cana-4362	281	2	figure	figure	NOUN
cana-4362	281	3	3	3	NUM
cana-4362	281	4	:	:	PUNCT
cana-4362	281	5	classification	classification	NOUN
cana-4362	281	6	accuracy	accuracy	NOUN
cana-4362	281	7	for	for	ADP
cana-4362	281	8	proposed	propose	VERB
cana-4362	281	9	model	model	NOUN
cana-4362	281	10	using	use	VERB
cana-4362	281	11	lstm	lstm	ADJ
cana-4362	281	12	figure	figure	NOUN
cana-4362	281	13	3	3	NUM
cana-4362	281	14	illustrates	illustrate	VERB
cana-4362	281	15	the	the	DET
cana-4362	281	16	classification	classification	NOUN
cana-4362	281	17	accuracy	accuracy	NOUN
cana-4362	281	18	achieved	achieve	VERB
cana-4362	281	19	with	with	ADP
cana-4362	281	20	the	the	DET
cana-4362	281	21	lstm	lstm	PROPN
cana-4362	281	22	algorithm	algorithm	NOUN
cana-4362	281	23	.	.	PUNCT
cana-4362	282	1	notably	notably	ADV
cana-4362	282	2	,	,	PUNCT
cana-4362	282	3	at	at	ADP
cana-4362	282	4	100	100	NUM
cana-4362	282	5	epochs	epoch	NOUN
cana-4362	282	6	,	,	PUNCT
cana-4362	282	7	the	the	DET
cana-4362	282	8	model	model	NOUN
cana-4362	282	9	demonstrates	demonstrate	VERB
cana-4362	282	10	impressive	impressive	ADJ
cana-4362	282	11	accuracy	accuracy	NOUN
cana-4362	282	12	when	when	SCONJ
cana-4362	282	13	applied	apply	VERB
cana-4362	282	14	to	to	ADP
cana-4362	282	15	real	real	ADJ
cana-4362	282	16	-	-	PUNCT
cana-4362	282	17	time	time	NOUN
cana-4362	282	18	text	text	NOUN
cana-4362	282	19	data	datum	NOUN
cana-4362	282	20	.	.	PUNCT
cana-4362	283	1	this	this	DET
cana-4362	283	2	model	model	NOUN
cana-4362	283	3	achieves	achieve	VERB
cana-4362	283	4	a	a	DET
cana-4362	283	5	remarkable	remarkable	ADJ
cana-4362	283	6	accuracy	accuracy	NOUN
cana-4362	283	7	rate	rate	NOUN
cana-4362	283	8	of	of	ADP
cana-4362	283	9	0.96	0.96	NUM
cana-4362	283	10	%	%	NOUN
cana-4362	283	11	on	on	ADP
cana-4362	283	12	the	the	DET
cana-4362	283	13	social	social	ADJ
cana-4362	283	14	media	medium	NOUN
cana-4362	283	15	dataset	dataset	NOUN
cana-4362	283	16	,	,	PUNCT
cana-4362	283	17	highlighting	highlight	VERB
cana-4362	283	18	its	its	PRON
cana-4362	283	19	effectiveness	effectiveness	NOUN
cana-4362	283	20	in	in	ADP
cana-4362	283	21	sentiment	sentiment	NOUN
cana-4362	283	22	classification	classification	NOUN
cana-4362	283	23	tasks	task	NOUN
cana-4362	283	24	.	.	PUNCT
cana-4362	284	1	figure	figure	VERB
cana-4362	284	2	4	4	NUM
cana-4362	284	3	:	:	PUNCT
cana-4362	284	4	classification	classification	NOUN
cana-4362	284	5	accuracy	accuracy	NOUN
cana-4362	284	6	for	for	ADP
cana-4362	284	7	proposed	propose	VERB
cana-4362	284	8	model	model	NOUN
cana-4362	284	9	using	use	VERB
cana-4362	284	10	rnn	rnn	PROPN
cana-4362	284	11	-	-	PUNCT
cana-4362	284	12	lstm	lstm	NOUN
cana-4362	284	13	the	the	DET
cana-4362	284	14	figure	figure	NOUN
cana-4362	284	15	4	4	NUM
cana-4362	284	16	demonstrates	demonstrate	VERB
cana-4362	284	17	the	the	DET
cana-4362	284	18	classification	classification	NOUN
cana-4362	284	19	accuracy	accuracy	NOUN
cana-4362	284	20	achieved	achieve	VERB
cana-4362	284	21	using	use	VERB
cana-4362	284	22	the	the	DET
cana-4362	284	23	rnn	rnn	PROPN
cana-4362	284	24	-	-	PUNCT
cana-4362	284	25	lstm	lstm	ADJ
cana-4362	284	26	algorithm	algorithm	NOUN
cana-4362	284	27	.	.	PUNCT
cana-4362	285	1	notably	notably	ADV
cana-4362	285	2	,	,	PUNCT
cana-4362	285	3	the	the	DET
cana-4362	285	4	model	model	NOUN
cana-4362	285	5	exhibits	exhibit	VERB
cana-4362	285	6	optimal	optimal	ADJ
cana-4362	285	7	performance	performance	NOUN
cana-4362	285	8	at	at	ADP
cana-4362	285	9	100	100	NUM
cana-4362	285	10	epochs	epoch	NOUN
cana-4362	285	11	,	,	PUNCT
cana-4362	285	12	where	where	SCONJ
cana-4362	285	13	it	it	PRON
cana-4362	285	14	achieves	achieve	VERB
cana-4362	285	15	an	an	DET
cana-4362	285	16	impressive	impressive	ADJ
cana-4362	285	17	accuracy	accuracy	NOUN
cana-4362	285	18	rate	rate	NOUN
cana-4362	285	19	of	of	ADP
cana-4362	285	20	0.98	0.98	NUM
cana-4362	285	21	%	%	NOUN
cana-4362	285	22	on	on	ADP
cana-4362	285	23	real	real	ADJ
cana-4362	285	24	-	-	PUNCT
cana-4362	285	25	time	time	NOUN
cana-4362	285	26	text	text	NOUN
cana-4362	285	27	data	datum	NOUN
cana-4362	285	28	from	from	ADP
cana-4362	285	29	social	social	ADJ
cana-4362	285	30	media	medium	NOUN
cana-4362	285	31	platforms	platform	NOUN
cana-4362	285	32	.	.	PUNCT
cana-4362	286	1	this	this	PRON
cana-4362	286	2	indicates	indicate	VERB
cana-4362	286	3	that	that	SCONJ
cana-4362	286	4	the	the	DET
cana-4362	286	5	rnn	rnn	NOUN
cana-4362	286	6	algorithm	algorithm	NOUN
cana-4362	286	7	is	be	AUX
cana-4362	286	8	highly	highly	ADV
cana-4362	286	9	effective	effective	ADJ
cana-4362	286	10	in	in	ADP
cana-4362	286	11	accurately	accurately	ADV
cana-4362	286	12	classifying	classify	VERB
cana-4362	286	13	sentiments	sentiment	NOUN
cana-4362	286	14	within	within	ADP
cana-4362	286	15	dynamic	dynamic	ADJ
cana-4362	286	16	social	social	ADJ
cana-4362	286	17	media	medium	NOUN
cana-4362	286	18	datasets	dataset	NOUN
cana-4362	286	19	.	.	PUNCT
cana-4362	287	1	table	table	NOUN
cana-4362	287	2	1	1	NUM
cana-4362	287	3	:	:	PUNCT
cana-4362	287	4	performance	performance	NOUN
cana-4362	287	5	matrix	matrix	NOUN
cana-4362	287	6	for	for	ADP
cana-4362	287	7	training	training	NOUN
cana-4362	287	8	of	of	ADP
cana-4362	287	9	proposed	propose	VERB
cana-4362	287	10	rnn	rnn	NOUN
cana-4362	288	1	no	no	INTJ
cana-4362	288	2	.	.	PUNCT
cana-4362	289	1	of	of	ADP
cana-4362	289	2	epoch	epoch	PROPN
cana-4362	289	3	dropout	dropout	NOUN
cana-4362	289	4	batch	batch	NOUN
cana-4362	289	5	size	size	NOUN
cana-4362	289	6	loss	loss	NOUN
cana-4362	289	7	accuracy	accuracy	NOUN
cana-4362	289	8	precision	precision	NOUN
cana-4362	289	9	recall	recall	VERB
cana-4362	289	10	auc	auc	NOUN
cana-4362	289	11	f1	f1	PROPN
cana-4362	289	12	specificity	specificity	NOUN
cana-4362	289	13	10	10	NUM
cana-4362	289	14	without	without	ADP
cana-4362	289	15	16	16	NUM
cana-4362	289	16	0.04	0.04	NUM
cana-4362	289	17	99.32	99.32	NUM
cana-4362	289	18	99.1	99.1	NUM
cana-4362	289	19	99.2	99.2	NUM
cana-4362	289	20	99.87	99.87	NUM
cana-4362	289	21	99.1	99.1	NUM
cana-4362	289	22	98.72	98.72	NUM
cana-4362	289	23	32	32	NUM
cana-4362	289	24	0.02	0.02	NUM
cana-4362	289	25	99.82	99.82	NUM
cana-4362	289	26	99.8.7	99.8.7	NUM
cana-4362	289	27	99.7	99.7	NUM
cana-4362	289	28	1	1	NUM
cana-4362	289	29	99.8	99.8	NUM
cana-4362	289	30	99.6.8	99.6.8	NUM
cana-4362	289	31	64	64	NUM
cana-4362	289	32	0.03	0.03	NUM
cana-4362	289	33	99.88	99.88	NUM
cana-4362	289	34	99.7.9	99.7.9	NOUN
cana-4362	290	1	99.9	99.9	NUM
cana-4362	290	2	1	1	NUM
cana-4362	290	3	99.7	99.7	NUM
cana-4362	290	4	98.37	98.37	NUM
cana-4362	290	5	128	128	NUM
cana-4362	290	6	0.01	0.01	NUM
cana-4362	290	7	97.90	97.90	NUM
cana-4362	290	8	.	.	PUNCT
cana-4362	291	1	98.02	98.02	NUM
cana-4362	291	2	98.80	98.80	NUM
cana-4362	291	3	99.87	99.87	NUM
cana-4362	291	4	97.84	97.84	NUM
cana-4362	291	5	94.54	94.54	NUM
cana-4362	291	6	2	2	NUM
cana-4362	291	7	16	16	NUM
cana-4362	291	8	0.01	0.01	NUM
cana-4362	291	9	99.87	99.87	NUM
cana-4362	291	10	99.1	99.1	NUM
cana-4362	291	11	99.6.1	99.6.1	NOUN
cana-4362	291	12	1	1	NUM
cana-4362	291	13	99.6.7	99.6.7	NUM
cana-4362	291	14	99.5.7	99.5.7	NOUN
cana-4362	291	15	0.86	0.86	NUM
cana-4362	291	16	0.88	0.88	NUM
cana-4362	291	17	0.9	0.9	NUM
cana-4362	291	18	0.92	0.92	NUM
cana-4362	291	19	0.94	0.94	NUM
cana-4362	291	20	0.96	0.96	NUM
cana-4362	291	21	0.98	0.98	NUM
cana-4362	291	22	25	25	NUM
cana-4362	291	23	50	50	NUM
cana-4362	291	24	75	75	NUM
cana-4362	291	25	100	100	NUM
cana-4362	291	26	va	va	NOUN
cana-4362	291	27	lu	lu	PROPN
cana-4362	291	28	e	e	PROPN
cana-4362	291	29	epoch	epoch	NOUN
cana-4362	291	30	size	size	NOUN
cana-4362	291	31	accuracy	accuracy	NOUN
cana-4362	291	32	precision	precision	NOUN
cana-4362	291	33	recall	recall	VERB
cana-4362	291	34	f	f	X
cana-4362	291	35	-	-	PUNCT
cana-4362	291	36	score	score	VERB
cana-4362	291	37	0.8	0.8	NUM
cana-4362	291	38	0.85	0.85	NUM
cana-4362	291	39	0.9	0.9	NUM
cana-4362	291	40	0.95	0.95	NUM
cana-4362	291	41	1	1	NUM
cana-4362	291	42	25	25	NUM
cana-4362	291	43	50	50	NUM
cana-4362	291	44	75	75	NUM
cana-4362	291	45	100	100	NUM
cana-4362	291	46	va	va	NOUN
cana-4362	291	47	lu	lu	PROPN
cana-4362	291	48	e	e	PROPN
cana-4362	291	49	epoch	epoch	NOUN
cana-4362	291	50	size	size	NOUN
cana-4362	291	51	accuracy	accuracy	NOUN
cana-4362	291	52	precision	precision	NOUN
cana-4362	291	53	recall	recall	VERB
cana-4362	291	54	f	f	NOUN
cana-4362	291	55	-	-	PUNCT
cana-4362	291	56	score	score	NOUN
cana-4362	291	57	communications	communication	NOUN
cana-4362	291	58	on	on	ADP
cana-4362	291	59	applied	apply	VERB
cana-4362	291	60	nonlinear	nonlinear	ADJ
cana-4362	291	61	analysis	analysis	NOUN
cana-4362	291	62	issn	issn	NOUN
cana-4362	291	63	:	:	PUNCT
cana-4362	291	64	1074	1074	NUM
cana-4362	291	65	-	-	PUNCT
cana-4362	291	66	133x	133x	NUM
cana-4362	291	67	vol	vol	NOUN
cana-4362	291	68	32	32	NUM
cana-4362	291	69	no	no	NOUN
cana-4362	291	70	.	.	PUNCT
cana-4362	292	1	9s	9s	NUM
cana-4362	292	2	(	(	PUNCT
cana-4362	292	3	2025	2025	NUM
cana-4362	292	4	)	)	PUNCT
cana-4362	292	5	1912	1912	NUM
cana-4362	292	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-4362	292	7	32	32	NUM
cana-4362	292	8	0.05	0.05	NUM
cana-4362	292	9	99.40	99.40	NUM
cana-4362	292	10	99.6	99.6	NUM
cana-4362	292	11	.	.	PUNCT
cana-4362	293	1	99.5.2	99.5.2	NOUN
cana-4362	293	2	99.5	99.5	NUM
cana-4362	293	3	99.7.6	99.7.6	NOUN
cana-4362	293	4	98.43	98.43	NUM
cana-4362	293	5	64	64	NUM
cana-4362	293	6	0.06	0.06	NUM
cana-4362	293	7	97.80	97.80	NUM
cana-4362	293	8	98.9	98.9	NUM
cana-4362	293	9	97.68	97.68	NUM
cana-4362	293	10	99.3	99.3	NUM
cana-4362	293	11	97.78	97.78	NUM
cana-4362	293	12	96.22	96.22	NUM
cana-4362	293	13	128	128	NUM
cana-4362	293	14	0.04	0.04	NUM
cana-4362	293	15	96.45	96.45	NUM
cana-4362	293	16	97.01	97.01	NUM
cana-4362	293	17	96.31	96.31	NUM
cana-4362	293	18	99.19	99.19	NUM
cana-4362	293	19	96.39	96.39	NUM
cana-4362	293	20	92.44	92.44	NUM
cana-4362	293	21	5	5	NUM
cana-4362	293	22	16	16	NUM
cana-4362	293	23	0.04	0.04	NUM
cana-4362	293	24	97.88	97.88	NUM
cana-4362	293	25	98.22	98.22	NUM
cana-4362	293	26	98.83	98.83	NUM
cana-4362	293	27	99	99	NUM
cana-4362	293	28	..	..	SYM
cana-4362	293	29	4	4	NUM
cana-4362	293	30	97.81	97.81	NUM
cana-4362	293	31	95.72	95.72	NUM
cana-4362	293	32	32	32	NUM
cana-4362	293	33	0.02	0.02	NUM
cana-4362	293	34	97.23	97.23	NUM
cana-4362	293	35	96.22	96.22	NUM
cana-4362	293	36	97.50	97.50	NUM
cana-4362	293	37	99.72	99.72	NUM
cana-4362	293	38	97.25	97.25	NUM
cana-4362	293	39	95.34	95.34	NUM
cana-4362	293	40	64	64	NUM
cana-4362	293	41	0.03	0.03	NUM
cana-4362	293	42	96.24	96.24	NUM
cana-4362	293	43	97.10	97.10	NUM
cana-4362	293	44	96.14	96.14	NUM
cana-4362	293	45	98.3	98.3	NUM
cana-4362	293	46	96.27	96.27	NUM
cana-4362	293	47	93.63	93.63	NUM
cana-4362	293	48	128	128	NUM
cana-4362	293	49	0.01	0.01	NUM
cana-4362	293	50	93.88	93.88	NUM
cana-4362	293	51	94.0	94.0	NUM
cana-4362	293	52	93.78	93.78	NUM
cana-4362	293	53	97.91	97.91	NUM
cana-4362	293	54	93.26	93.26	NUM
cana-4362	293	55	8760	8760	NUM
cana-4362	293	56	15	15	NUM
cana-4362	293	57	without	without	ADP
cana-4362	293	58	16	16	NUM
cana-4362	293	59	0.01	0.01	NUM
cana-4362	293	60	99.30	99.30	NUM
cana-4362	293	61	1	1	NUM
cana-4362	293	62	1	1	NUM
cana-4362	293	63	1	1	NUM
cana-4362	293	64	1	1	NUM
cana-4362	293	65	99.99	99.99	NUM
cana-4362	293	66	32	32	NUM
cana-4362	293	67	0.05	0.05	NUM
cana-4362	293	68	100	100	NUM
cana-4362	293	69	1	1	NUM
cana-4362	293	70	1	1	NUM
cana-4362	293	71	1	1	NUM
cana-4362	293	72	1	1	NUM
cana-4362	293	73	99.96	99.96	NUM
cana-4362	293	74	64	64	NUM
cana-4362	293	75	0.06	0.06	NUM
cana-4362	293	76	100	100	NUM
cana-4362	293	77	1	1	NUM
cana-4362	293	78	1	1	NUM
cana-4362	293	79	1	1	NUM
cana-4362	293	80	1	1	NUM
cana-4362	293	81	99.8.2	99.8.2	NOUN
cana-4362	293	82	128	128	NUM
cana-4362	293	83	0.04	0.04	NUM
cana-4362	293	84	97.5	97.5	NUM
cana-4362	293	85	97.49	97.49	NUM
cana-4362	293	86	97.51	97.51	NUM
cana-4362	293	87	99.72	99.72	NUM
cana-4362	293	88	97.50	97.50	NUM
cana-4362	293	89	93.84	93.84	NUM
cana-4362	293	90	2	2	NUM
cana-4362	293	91	16	16	NUM
cana-4362	293	92	0.04	0.04	NUM
cana-4362	293	93	100	100	NUM
cana-4362	293	94	1	1	NUM
cana-4362	293	95	1	1	NUM
cana-4362	293	96	1	1	NUM
cana-4362	293	97	1	1	NUM
cana-4362	293	98	99.97	99.97	NUM
cana-4362	293	99	32	32	NUM
cana-4362	293	100	0.04	0.04	NUM
cana-4362	293	101	99.89	99.89	NUM
cana-4362	293	102	99.86	99.86	NUM
cana-4362	293	103	99.89	99.89	NUM
cana-4362	293	104	1	1	NUM
cana-4362	293	105	99.8	99.8	NUM
cana-4362	293	106	99.6.3	99.6.3	NUM
cana-4362	293	107	64	64	NUM
cana-4362	293	108	0.04	0.04	NUM
cana-4362	293	109	99.75	99.75	NUM
cana-4362	293	110	99.69	99.69	NUM
cana-4362	293	111	99.75	99.75	NUM
cana-4362	293	112	1	1	NUM
cana-4362	293	113	99.7	99.7	NUM
cana-4362	293	114	99.0	99.0	NUM
cana-4362	293	115	128	128	NUM
cana-4362	293	116	0.04	0.04	NUM
cana-4362	293	117	99.83	99.83	NUM
cana-4362	293	118	99.80	99.80	NUM
cana-4362	293	119	99.83	99.83	NUM
cana-4362	293	120	1	1	NUM
cana-4362	293	121	99.8	99.8	NUM
cana-4362	293	122	98.83	98.83	NUM
cana-4362	293	123	5	5	NUM
cana-4362	293	124	16	16	NUM
cana-4362	293	125	0.02	0.02	NUM
cana-4362	293	126	98.59	98.59	NUM
cana-4362	293	127	98.61	98.61	NUM
cana-4362	293	128	98.71	98.71	NUM
cana-4362	293	129	99.80	99.80	NUM
cana-4362	293	130	98.66	98.66	NUM
cana-4362	293	131	97.78	97.78	NUM
cana-4362	293	132	32	32	NUM
cana-4362	293	133	0.03	0.03	NUM
cana-4362	293	134	99.58	99.58	NUM
cana-4362	293	135	99.5.3	99.5.3	NUM
cana-4362	293	136	99.55	99.55	NUM
cana-4362	293	137	99.99	99.99	NUM
cana-4362	293	138	99.54	99.54	NUM
cana-4362	293	139	99.1.2	99.1.2	NOUN
cana-4362	293	140	64	64	NUM
cana-4362	293	141	0.01	0.01	NUM
cana-4362	293	142	99.77	99.77	NUM
cana-4362	293	143	99.72	99.72	NUM
cana-4362	293	144	99.78	99.78	NUM
cana-4362	293	145	99.99	99.99	NUM
cana-4362	293	146	.	.	PUNCT
cana-4362	294	1	99.75	99.75	NUM
cana-4362	294	2	98.85	98.85	NUM
cana-4362	294	3	128	128	NUM
cana-4362	294	4	0.01	0.01	NUM
cana-4362	294	5	98.79	98.79	NUM
cana-4362	294	6	98.86	98.86	NUM
cana-4362	294	7	98.74	98.74	NUM
cana-4362	294	8	99.87	99.87	NUM
cana-4362	294	9	98.80	98.80	NUM
cana-4362	294	10	96.54	96.54	NUM
cana-4362	294	11	20	20	NUM
cana-4362	294	12	without	without	ADP
cana-4362	294	13	16	16	NUM
cana-4362	294	14	0.05	0.05	NUM
cana-4362	294	15	100	100	NUM
cana-4362	294	16	1	1	NUM
cana-4362	294	17	1	1	NUM
cana-4362	294	18	1	1	NUM
cana-4362	294	19	1	1	NUM
cana-4362	294	20	99.99	99.99	NUM
cana-4362	294	21	.	.	PUNCT
cana-4362	295	1	32	32	NUM
cana-4362	295	2	0.06	0.06	NUM
cana-4362	295	3	100	100	NUM
cana-4362	295	4	1	1	NUM
cana-4362	295	5	1	1	NUM
cana-4362	295	6	1	1	NUM
cana-4362	295	7	1	1	NUM
cana-4362	295	8	99.98	99.98	NUM
cana-4362	295	9	.	.	PUNCT
cana-4362	296	1	64	64	NUM
cana-4362	296	2	0.04	0.04	NUM
cana-4362	296	3	100	100	NUM
cana-4362	296	4	1	1	NUM
cana-4362	296	5	1	1	NUM
cana-4362	296	6	1	1	NUM
cana-4362	296	7	1	1	NUM
cana-4362	296	8	99.98	99.98	NUM
cana-4362	296	9	.	.	PUNCT
cana-4362	297	1	128	128	NUM
cana-4362	297	2	0.04	0.04	NUM
cana-4362	297	3	100	100	NUM
cana-4362	297	4	1	1	NUM
cana-4362	297	5	1	1	NUM
cana-4362	297	6	1	1	NUM
cana-4362	297	7	1	1	NUM
cana-4362	297	8	99.5	99.5	NUM
cana-4362	297	9	2	2	NUM
cana-4362	297	10	16	16	NUM
cana-4362	297	11	0.04	0.04	NUM
cana-4362	297	12	100	100	NUM
cana-4362	297	13	1	1	NUM
cana-4362	297	14	1	1	NUM
cana-4362	297	15	1	1	NUM
cana-4362	297	16	1	1	NUM
cana-4362	297	17	99.7	99.7	NUM
cana-4362	297	18	32	32	NUM
cana-4362	297	19	0.04	0.04	NUM
cana-4362	297	20	100	100	NUM
cana-4362	297	21	1	1	NUM
cana-4362	297	22	1	1	NUM
cana-4362	297	23	1	1	NUM
cana-4362	297	24	1	1	NUM
cana-4362	297	25	99.95	99.95	NUM
cana-4362	297	26	64	64	NUM
cana-4362	297	27	0.03	0.03	NUM
cana-4362	297	28	100	100	NUM
cana-4362	297	29	1	1	NUM
cana-4362	297	30	1	1	NUM
cana-4362	297	31	1	1	NUM
cana-4362	297	32	1	1	NUM
cana-4362	297	33	99.81	99.81	NUM
cana-4362	297	34	128	128	NUM
cana-4362	297	35	0.03	0.03	NUM
cana-4362	297	36	99.86	99.86	NUM
cana-4362	297	37	99.86	99.86	NUM
cana-4362	297	38	99.86	99.86	NUM
cana-4362	297	39	1	1	NUM
cana-4362	297	40	99.8.6	99.8.6	NUM
cana-4362	297	41	99.20	99.20	NUM
cana-4362	297	42	5	5	NUM
cana-4362	297	43	16	16	NUM
cana-4362	297	44	0.02	0.02	NUM
cana-4362	297	45	99.18	99.18	NUM
cana-4362	297	46	99.16	99.16	NUM
cana-4362	297	47	99.16	99.16	NUM
cana-4362	297	48	99.89	99.89	NUM
cana-4362	297	49	99.16	99.16	NUM
cana-4362	297	50	98.71	98.71	NUM
cana-4362	297	51	32	32	NUM
cana-4362	297	52	0.03	0.03	NUM
cana-4362	297	53	99.89	99.89	NUM
cana-4362	297	54	99.89	99.89	NUM
cana-4362	297	55	99.89	99.89	NUM
cana-4362	297	56	1	1	NUM
cana-4362	297	57	99.89	99.89	NUM
cana-4362	297	58	99.7.0	99.7.0	NUM
cana-4362	297	59	64	64	NUM
cana-4362	297	60	0.036	0.036	NUM
cana-4362	297	61	99.07	99.07	NUM
cana-4362	297	62	99.03	99.03	NUM
cana-4362	297	63	98.94	98.94	NUM
cana-4362	297	64	99.87	99.87	NUM
cana-4362	297	65	98.98	98.98	NUM
cana-4362	297	66	97.73	97.73	NUM
cana-4362	297	67	128	128	NUM
cana-4362	297	68	0.035	0.035	NUM
cana-4362	297	69	98.99	98.99	NUM
cana-4362	297	70	.	.	PUNCT
cana-4362	298	1	99.11	99.11	NUM
cana-4362	298	2	99.08	99.08	NUM
cana-4362	298	3	99.93	99.93	NUM
cana-4362	298	4	99.09	99.09	NUM
cana-4362	298	5	97.53	97.53	NUM
cana-4362	298	6	table	table	NOUN
cana-4362	298	7	2	2	NUM
cana-4362	298	8	:	:	PUNCT
cana-4362	298	9	performance	performance	NOUN
cana-4362	298	10	matrix	matrix	NOUN
cana-4362	298	11	for	for	ADP
cana-4362	298	12	testing	testing	NOUN
cana-4362	298	13	of	of	ADP
cana-4362	298	14	proposed	propose	VERB
cana-4362	298	15	rnn	rnn	NOUN
cana-4362	298	16	no	no	INTJ
cana-4362	298	17	.	.	PUNCT
cana-4362	299	1	of	of	ADP
cana-4362	299	2	epoch	epoch	PROPN
cana-4362	299	3	dropout	dropout	NOUN
cana-4362	299	4	batch	batch	NOUN
cana-4362	299	5	size	size	NOUN
cana-4362	299	6	loss	loss	NOUN
cana-4362	299	7	accuracy	accuracy	NOUN
cana-4362	299	8	precision	precision	NOUN
cana-4362	299	9	recall	recall	VERB
cana-4362	299	10	auc	auc	NOUN
cana-4362	299	11	f1	f1	PROPN
cana-4362	299	12	specificity	specificity	NOUN
cana-4362	299	13	10	10	NUM
cana-4362	299	14	without	without	ADP
cana-4362	299	15	16	16	NUM
cana-4362	299	16	0.03	0.03	NUM
cana-4362	299	17	99.31	99.31	NUM
cana-4362	299	18	99.2	99.2	NUM
cana-4362	299	19	98.35	98.35	NUM
cana-4362	299	20	99.8.7	99.8.7	NOUN
cana-4362	299	21	992.3	992.3	NUM
cana-4362	299	22	98.73	98.73	NUM
cana-4362	299	23	32	32	NUM
cana-4362	299	24	0.04	0.04	NUM
cana-4362	299	25	99.98	99.98	NUM
cana-4362	299	26	.	.	PUNCT
cana-4362	300	1	99.20	99.20	NUM
cana-4362	300	2	99.6	99.6	NUM
cana-4362	300	3	1	1	NUM
cana-4362	300	4	99.5	99.5	NUM
cana-4362	300	5	99.7.6	99.7.6	NUM
cana-4362	300	6	64	64	NUM
cana-4362	300	7	0.02	0.02	NUM
cana-4362	300	8	99.98	99.98	NUM
cana-4362	300	9	.	.	PUNCT
cana-4362	301	1	99.98	99.98	NUM
cana-4362	301	2	.	.	PUNCT
cana-4362	302	1	99.98	99.98	NUM
cana-4362	302	2	1	1	NUM
cana-4362	302	3	99.98	99.98	NUM
cana-4362	302	4	98.46	98.46	NUM
cana-4362	302	5	128	128	NUM
cana-4362	302	6	0.03	0.03	NUM
cana-4362	302	7	97.97	97.97	NUM
cana-4362	302	8	.	.	PUNCT
cana-4362	303	1	97.9	97.9	NUM
cana-4362	303	2	97.80	97.80	NUM
cana-4362	303	3	99.6	99.6	NUM
cana-4362	303	4	97.85	97.85	NUM
cana-4362	303	5	94.45	94.45	NUM
cana-4362	303	6	2	2	NUM
cana-4362	303	7	16	16	NUM
cana-4362	303	8	0.04	0.04	NUM
cana-4362	303	9	99.85	99.85	NUM
cana-4362	303	10	99.8	99.8	NUM
cana-4362	303	11	99.64	99.64	NUM
cana-4362	303	12	1	1	NUM
cana-4362	303	13	99.6	99.6	NUM
cana-4362	303	14	99.5	99.5	NUM
cana-4362	303	15	communications	communication	NOUN
cana-4362	303	16	on	on	ADP
cana-4362	303	17	applied	apply	VERB
cana-4362	303	18	nonlinear	nonlinear	ADJ
cana-4362	303	19	analysis	analysis	NOUN
cana-4362	303	20	issn	issn	NOUN
cana-4362	303	21	:	:	PUNCT
cana-4362	303	22	1074	1074	NUM
cana-4362	303	23	-	-	PUNCT
cana-4362	303	24	133x	133x	NUM
cana-4362	303	25	vol	vol	NOUN
cana-4362	303	26	32	32	NUM
cana-4362	303	27	no	no	NOUN
cana-4362	303	28	.	.	PUNCT
cana-4362	304	1	9s	9s	NUM
cana-4362	304	2	(	(	PUNCT
cana-4362	304	3	2025	2025	NUM
cana-4362	304	4	)	)	PUNCT
cana-4362	304	5	1913	1913	NUM
cana-4362	304	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-4362	305	1	32	32	NUM
cana-4362	305	2	0.02	0.02	NUM
cana-4362	305	3	99.48	99.48	NUM
cana-4362	305	4	99.6	99.6	NUM
cana-4362	305	5	99.62	99.62	NUM
cana-4362	305	6	99.8	99.8	NUM
cana-4362	305	7	99.5	99.5	NUM
cana-4362	305	8	98.56	98.56	NUM
cana-4362	305	9	64	64	NUM
cana-4362	305	10	0.03	0.03	NUM
cana-4362	305	11	97.85	97.85	NUM
cana-4362	305	12	97.78	97.78	NUM
cana-4362	305	13	97.88	97.88	NUM
cana-4362	305	14	99.97	99.97	NUM
cana-4362	305	15	97.78	97.78	NUM
cana-4362	305	16	96.42	96.42	NUM
cana-4362	305	17	128	128	NUM
cana-4362	305	18	0.01	0.01	NUM
cana-4362	305	19	96.45	96.45	NUM
cana-4362	305	20	96.40	96.40	NUM
cana-4362	305	21	96.45	96.45	NUM
cana-4362	305	22	99.8.7	99.8.7	NUM
cana-4362	305	23	96.28	96.28	NUM
cana-4362	305	24	91.65	91.65	NUM
cana-4362	305	25	5	5	NUM
cana-4362	305	26	16	16	NUM
cana-4362	305	27	0.01	0.01	NUM
cana-4362	305	28	97.81	97.81	NUM
cana-4362	305	29	98.81	98.81	NUM
cana-4362	305	30	97.85	97.85	NUM
cana-4362	305	31	99.7.3	99.7.3	NOUN
cana-4362	305	32	97.81	97.81	NUM
cana-4362	305	33	97.62	97.62	NUM
cana-4362	305	34	32	32	NUM
cana-4362	305	35	0.05	0.05	NUM
cana-4362	305	36	97.30	97.30	NUM
cana-4362	305	37	97.31	97.31	NUM
cana-4362	305	38	97.20	97.20	NUM
cana-4362	305	39	99.7.1	99.7.1	NOUN
cana-4362	305	40	97.24	97.24	NUM
cana-4362	305	41	95.23	95.23	NUM
cana-4362	305	42	64	64	NUM
cana-4362	305	43	0.03	0.03	NUM
cana-4362	305	44	96.22	96.22	NUM
cana-4362	305	45	96.12	96.12	NUM
cana-4362	305	46	96.14	96.14	NUM
cana-4362	305	47	993.2	993.2	NUM
cana-4362	305	48	96.76	96.76	NUM
cana-4362	305	49	93.53	93.53	NUM
cana-4362	305	50	128	128	NUM
cana-4362	305	51	0.04	0.04	NUM
cana-4362	305	52	93.88	93.88	NUM
cana-4362	305	53	93.73	93.73	NUM
cana-4362	305	54	94.59	94.59	NUM
cana-4362	305	55	97.71	97.71	NUM
cana-4362	305	56	93.75	93.75	NUM
cana-4362	305	57	8750	8750	NUM
cana-4362	305	58	15	15	NUM
cana-4362	305	59	without	without	ADP
cana-4362	305	60	16	16	NUM
cana-4362	305	61	0.02	0.02	NUM
cana-4362	305	62	100	100	NUM
cana-4362	305	63	1	1	NUM
cana-4362	305	64	1	1	NUM
cana-4362	305	65	1	1	NUM
cana-4362	305	66	1	1	NUM
cana-4362	305	67	99.99	99.99	NUM
cana-4362	305	68	.	.	PUNCT
cana-4362	306	1	32	32	NUM
cana-4362	306	2	0.03	0.03	NUM
cana-4362	306	3	100	100	NUM
cana-4362	306	4	1	1	NUM
cana-4362	306	5	1	1	NUM
cana-4362	306	6	1	1	NUM
cana-4362	306	7	1	1	NUM
cana-4362	306	8	99.96	99.96	NUM
cana-4362	306	9	.	.	PUNCT
cana-4362	307	1	64	64	NUM
cana-4362	307	2	0.01	0.01	NUM
cana-4362	307	3	100	100	NUM
cana-4362	307	4	1	1	NUM
cana-4362	307	5	1	1	NUM
cana-4362	307	6	1	1	NUM
cana-4362	307	7	1	1	NUM
cana-4362	307	8	99.8.2	99.8.2	NOUN
cana-4362	307	9	128	128	NUM
cana-4362	307	10	0.01	0.01	NUM
cana-4362	307	11	97.5	97.5	NUM
cana-4362	307	12	97.49	97.49	NUM
cana-4362	307	13	97.51	97.51	NUM
cana-4362	307	14	99.7.2	99.7.2	NOUN
cana-4362	307	15	97.50	97.50	NUM
cana-4362	307	16	93.84	93.84	NUM
cana-4362	307	17	2	2	NUM
cana-4362	307	18	16	16	NUM
cana-4362	307	19	0.05	0.05	NUM
cana-4362	307	20	100	100	NUM
cana-4362	307	21	1	1	NUM
cana-4362	307	22	1	1	NUM
cana-4362	307	23	1	1	NUM
cana-4362	307	24	1	1	NUM
cana-4362	307	25	99.97	99.97	NUM
cana-4362	307	26	.	.	PUNCT
cana-4362	307	27	32	32	NUM
cana-4362	307	28	0.06	0.06	NUM
cana-4362	307	29	99.89	99.89	NUM
cana-4362	307	30	99.8.6	99.8.6	NUM
cana-4362	307	31	99.8.9	99.8.9	SYM
cana-4362	307	32	1	1	NUM
cana-4362	307	33	99.8.7	99.8.7	NOUN
cana-4362	307	34	99.6.3	99.6.3	NUM
cana-4362	307	35	64	64	NUM
cana-4362	307	36	0.04	0.04	NUM
cana-4362	307	37	99.75	99.75	NUM
cana-4362	307	38	99.6.9	99.6.9	NUM
cana-4362	307	39	99.7.5	99.7.5	NUM
cana-4362	307	40	1	1	NUM
cana-4362	307	41	99.7.2	99.7.2	NUM
cana-4362	307	42	99.0.5	99.0.5	NOUN
cana-4362	307	43	128	128	NUM
cana-4362	307	44	0.04	0.04	NUM
cana-4362	307	45	99.83	99.83	NUM
cana-4362	307	46	99.8.0	99.8.0	NOUN
cana-4362	308	1	99.8.3	99.8.3	CCONJ
cana-4362	308	2	1	1	NUM
cana-4362	308	3	99.8.1	99.8.1	NOUN
cana-4362	309	1	98.83	98.83	NUM
cana-4362	309	2	5	5	NUM
cana-4362	309	3	16	16	NUM
cana-4362	309	4	0.03	0.03	NUM
cana-4362	309	5	98.59	98.59	NUM
cana-4362	309	6	98.61	98.61	NUM
cana-4362	309	7	98.71	98.71	NUM
cana-4362	309	8	99.8.0	99.8.0	X
cana-4362	310	1	98.66	98.66	NUM
cana-4362	310	2	97.78	97.78	NUM
cana-4362	310	3	32	32	NUM
cana-4362	310	4	0.04	0.04	NUM
cana-4362	310	5	99.58	99.58	NUM
cana-4362	310	6	99.5.3	99.5.3	NUM
cana-4362	310	7	99.5.5	99.5.5	NOUN
cana-4362	310	8	99.99	99.99	NUM
cana-4362	310	9	.	.	PUNCT
cana-4362	311	1	99.5.4	99.5.4	NUM
cana-4362	311	2	99.1.2	99.1.2	NOUN
cana-4362	311	3	64	64	NUM
cana-4362	311	4	0.02	0.02	NUM
cana-4362	311	5	99.77	99.77	NUM
cana-4362	311	6	99.7.2	99.7.2	NUM
cana-4362	311	7	99.7.8	99.7.8	NOUN
cana-4362	311	8	99.99	99.99	NUM
cana-4362	311	9	.	.	PUNCT
cana-4362	312	1	99.7.5	99.7.5	NOUN
cana-4362	312	2	98.85	98.85	NUM
cana-4362	313	1	128	128	NUM
cana-4362	313	2	0.03	0.03	NUM
cana-4362	313	3	98.79	98.79	NUM
cana-4362	313	4	98.86	98.86	NUM
cana-4362	313	5	98.74	98.74	NUM
cana-4362	313	6	99.8.7	99.8.7	NUM
cana-4362	313	7	98.80	98.80	NUM
cana-4362	313	8	96.54	96.54	NUM
cana-4362	313	9	20	20	NUM
cana-4362	313	10	without	without	ADP
cana-4362	313	11	16	16	NUM
cana-4362	313	12	0.01	0.01	NUM
cana-4362	313	13	100	100	NUM
cana-4362	313	14	1	1	NUM
cana-4362	313	15	1	1	NUM
cana-4362	313	16	1	1	NUM
cana-4362	313	17	1	1	NUM
cana-4362	313	18	99.99	99.99	NUM
cana-4362	313	19	.	.	PUNCT
cana-4362	314	1	32	32	NUM
cana-4362	314	2	0.01	0.01	NUM
cana-4362	314	3	100	100	NUM
cana-4362	314	4	1	1	NUM
cana-4362	314	5	1	1	NUM
cana-4362	314	6	1	1	NUM
cana-4362	314	7	1	1	NUM
cana-4362	314	8	99.98	99.98	NUM
cana-4362	314	9	.	.	PUNCT
cana-4362	315	1	64	64	NUM
cana-4362	315	2	0.05	0.05	NUM
cana-4362	315	3	100	100	NUM
cana-4362	315	4	1	1	NUM
cana-4362	315	5	1	1	NUM
cana-4362	315	6	1	1	NUM
cana-4362	315	7	1	1	NUM
cana-4362	315	8	99.98	99.98	NUM
cana-4362	315	9	.	.	PUNCT
cana-4362	316	1	128	128	NUM
cana-4362	316	2	0.06	0.06	NUM
cana-4362	316	3	100	100	NUM
cana-4362	316	4	1	1	NUM
cana-4362	316	5	1	1	NUM
cana-4362	316	6	1	1	NUM
cana-4362	316	7	1	1	NUM
cana-4362	316	8	992.5	992.5	NUM
cana-4362	316	9	2	2	NUM
cana-4362	316	10	16	16	NUM
cana-4362	316	11	0.04	0.04	NUM
cana-4362	316	12	100	100	NUM
cana-4362	316	13	1	1	NUM
cana-4362	316	14	1	1	NUM
cana-4362	316	15	1	1	NUM
cana-4362	316	16	1	1	NUM
cana-4362	316	17	99.99	99.99	NUM
cana-4362	316	18	.	.	PUNCT
cana-4362	317	1	32	32	NUM
cana-4362	317	2	0.04	0.04	NUM
cana-4362	317	3	100	100	NUM
cana-4362	317	4	1	1	NUM
cana-4362	317	5	1	1	NUM
cana-4362	317	6	1	1	NUM
cana-4362	317	7	1	1	NUM
cana-4362	317	8	99.95	99.95	NUM
cana-4362	317	9	.	.	PUNCT
cana-4362	318	1	64	64	NUM
cana-4362	318	2	0.03	0.03	NUM
cana-4362	318	3	100	100	NUM
cana-4362	318	4	1	1	NUM
cana-4362	318	5	1	1	NUM
cana-4362	318	6	1	1	NUM
cana-4362	318	7	1	1	NUM
cana-4362	318	8	99.8.1	99.8.1	NOUN
cana-4362	318	9	128	128	NUM
cana-4362	318	10	0.04	0.04	NUM
cana-4362	318	11	99.86	99.86	NUM
cana-4362	318	12	99.86	99.86	NUM
cana-4362	318	13	99.86	99.86	NUM
cana-4362	318	14	1	1	NUM
cana-4362	318	15	99.8.6	99.8.6	NUM
cana-4362	318	16	992.0	992.0	NUM
cana-4362	318	17	5	5	NUM
cana-4362	318	18	16	16	NUM
cana-4362	318	19	0.02	0.02	NUM
cana-4362	318	20	99.18	99.18	NUM
cana-4362	318	21	99.16	99.16	NUM
cana-4362	318	22	99.16	99.16	NUM
cana-4362	318	23	99.8.9	99.8.9	NUM
cana-4362	318	24	99.1.6	99.1.6	NOUN
cana-4362	318	25	98.71	98.71	NUM
cana-4362	318	26	32	32	NUM
cana-4362	318	27	0.03	0.03	NUM
cana-4362	318	28	99.89	99.89	NUM
cana-4362	318	29	99.89	99.89	NUM
cana-4362	318	30	99.89	99.89	NUM
cana-4362	318	31	1	1	NUM
cana-4362	318	32	99.8.9	99.8.9	NUM
cana-4362	318	33	99.7.0	99.7.0	NUM
cana-4362	318	34	64	64	NUM
cana-4362	318	35	0.01	0.01	NUM
cana-4362	318	36	99.07	99.07	NUM
cana-4362	318	37	99.03	99.03	NUM
cana-4362	318	38	98.94	98.94	NUM
cana-4362	318	39	99.87	99.87	NUM
cana-4362	318	40	98.98	98.98	NUM
cana-4362	318	41	.	.	PUNCT
cana-4362	319	1	97.73	97.73	NUM
cana-4362	319	2	128	128	NUM
cana-4362	319	3	0.01	0.01	NUM
cana-4362	319	4	98.99	98.99	NUM
cana-4362	319	5	99.11	99.11	NUM
cana-4362	319	6	99.08	99.08	NUM
cana-4362	319	7	99.93	99.93	NUM
cana-4362	319	8	99.0.9	99.0.9	NUM
cana-4362	319	9	97.53	97.53	NUM
cana-4362	319	10	table	table	NOUN
cana-4362	319	11	3	3	NUM
cana-4362	319	12	:	:	PUNCT
cana-4362	319	13	testing	testing	NOUN
cana-4362	319	14	results	result	NOUN
cana-4362	319	15	of	of	ADP
cana-4362	319	16	proposed	propose	VERB
cana-4362	319	17	model	model	NOUN
cana-4362	319	18	with	with	ADP
cana-4362	319	19	various	various	ADJ
cana-4362	319	20	hyper	hyper	ADJ
cana-4362	319	21	parameters	parameter	NOUN
cana-4362	319	22	changing	change	VERB
cana-4362	319	23	optimizer	optimizer	NOUN
cana-4362	319	24	no	no	PROPN
cana-4362	319	25	.	.	PUNCT
cana-4362	320	1	epochs	epoch	NOUN
cana-4362	320	2	batch	batch	VERB
cana-4362	320	3	size	size	NOUN
cana-4362	320	4	loss	loss	NOUN
cana-4362	320	5	accuracy	accuracy	NOUN
cana-4362	320	6	auc	auc	NOUN
cana-4362	320	7	precision	precision	NOUN
cana-4362	320	8	recall	recall	NOUN
cana-4362	320	9	f1	f1	PROPN
cana-4362	320	10	adam	adam	PROPN
cana-4362	320	11	50	50	NUM
cana-4362	320	12	32	32	NUM
cana-4362	320	13	0.03	0.03	NUM
cana-4362	320	14	85.35	85.35	NUM
cana-4362	320	15	93.81	93.81	NUM
cana-4362	320	16	84.32	84.32	NUM
cana-4362	320	17	84.45	84.45	NUM
cana-4362	320	18	84.38	84.38	NUM
cana-4362	320	19	64	64	NUM
cana-4362	320	20	0.03	0.03	NUM
cana-4362	320	21	85.27	85.27	NUM
cana-4362	320	22	94.53	94.53	NUM
cana-4362	320	23	85.37	85.37	NUM
cana-4362	320	24	85.31	85.31	NUM
cana-4362	320	25	85.34	85.34	NUM
cana-4362	320	26	250	250	NUM
cana-4362	320	27	32	32	NUM
cana-4362	320	28	0.01	0.01	NUM
cana-4362	320	29	95.66	95.66	NUM
cana-4362	320	30	98.76	98.76	NUM
cana-4362	320	31	94.81	94.81	NUM
cana-4362	320	32	95.56	95.56	NUM
cana-4362	320	33	95.68	95.68	NUM
cana-4362	320	34	communications	communication	NOUN
cana-4362	320	35	on	on	ADP
cana-4362	320	36	applied	apply	VERB
cana-4362	320	37	nonlinear	nonlinear	ADJ
cana-4362	320	38	analysis	analysis	NOUN
cana-4362	320	39	issn	issn	NOUN
cana-4362	320	40	:	:	PUNCT
cana-4362	320	41	1074	1074	NUM
cana-4362	320	42	-	-	PUNCT
cana-4362	320	43	133x	133x	NUM
cana-4362	320	44	vol	vol	NOUN
cana-4362	320	45	32	32	NUM
cana-4362	320	46	no	no	NOUN
cana-4362	320	47	.	.	PUNCT
cana-4362	321	1	9s	9s	NUM
cana-4362	321	2	(	(	PUNCT
cana-4362	321	3	2025	2025	NUM
cana-4362	321	4	)	)	PUNCT
cana-4362	321	5	1914	1914	NUM
cana-4362	321	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-4362	321	7	64	64	NUM
cana-4362	321	8	0.01	0.01	NUM
cana-4362	321	9	95.25	95.25	NUM
cana-4362	321	10	98.89	98.89	NUM
cana-4362	321	11	95.27	95.27	NUM
cana-4362	321	12	95.32	95.32	NUM
cana-4362	321	13	95.02	95.02	NUM
cana-4362	321	14	512	512	NUM
cana-4362	321	15	32	32	NUM
cana-4362	321	16	0.01	0.01	NUM
cana-4362	321	17	95.93	95.93	NUM
cana-4362	321	18	.	.	PUNCT
cana-4362	322	1	99.4.3	99.4.3	NOUN
cana-4362	323	1	96.10	96.10	NUM
cana-4362	323	2	96.83	96.83	NUM
cana-4362	323	3	96.05	96.05	NUM
cana-4362	323	4	64	64	NUM
cana-4362	323	5	0.00	0.00	NUM
cana-4362	323	6	97.40	97.40	NUM
cana-4362	323	7	99.7	99.7	NUM
cana-4362	323	8	97.55	97.55	NUM
cana-4362	323	9	97.58	97.58	NUM
cana-4362	323	10	97.57	97.57	NUM
cana-4362	323	11	sgd	sgd	NOUN
cana-4362	323	12	50	50	NUM
cana-4362	323	13	32	32	NUM
cana-4362	323	14	0.03	0.03	NUM
cana-4362	323	15	85.9	85.9	NUM
cana-4362	323	16	89.33	89.33	NUM
cana-4362	323	17	80.54	80.54	NUM
cana-4362	323	18	81.51	81.51	NUM
cana-4362	323	19	80.42	80.42	NUM
cana-4362	323	20	64	64	NUM
cana-4362	323	21	0.04	0.04	NUM
cana-4362	323	22	78.83	78.83	NUM
cana-4362	323	23	87.30	87.30	NUM
cana-4362	323	24	79.75	79.75	NUM
cana-4362	323	25	79.89	79.89	NUM
cana-4362	323	26	79.81	79.81	NUM
cana-4362	323	27	250	250	NUM
cana-4362	323	28	32	32	NUM
cana-4362	323	29	0.02	0.02	NUM
cana-4362	323	30	88.94	88.94	NUM
cana-4362	323	31	96.32	96.32	NUM
cana-4362	323	32	89.72	89.72	NUM
cana-4362	323	33	89.04	89.04	NUM
cana-4362	323	34	89.83	89.83	NUM
cana-4362	323	35	64	64	NUM
cana-4362	323	36	0.03	0.03	NUM
cana-4362	323	37	87.18	87.18	NUM
cana-4362	323	38	94.47	94.47	NUM
cana-4362	323	39	87.28	87.28	NUM
cana-4362	323	40	87.36	87.36	NUM
cana-4362	323	41	87.32	87.32	NUM
cana-4362	323	42	512	512	NUM
cana-4362	323	43	32	32	NUM
cana-4362	323	44	0.01	0.01	NUM
cana-4362	323	45	95.20	95.20	NUM
cana-4362	323	46	98.78	98.78	NUM
cana-4362	323	47	94.21	94.21	NUM
cana-4362	323	48	94.12	94.12	NUM
cana-4362	323	49	94.16	94.16	NUM
cana-4362	323	50	64	64	NUM
cana-4362	323	51	0.01	0.01	NUM
cana-4362	323	52	92.80	92.80	NUM
cana-4362	323	53	98.22	98.22	NUM
cana-4362	323	54	93.73	93.73	NUM
cana-4362	323	55	93.65	93.65	NUM
cana-4362	323	56	93.76	93.76	NUM
cana-4362	323	57	adadelta	adadelta	NOUN
cana-4362	323	58	50	50	NUM
cana-4362	323	59	32	32	NUM
cana-4362	323	60	0.05	0.05	NUM
cana-4362	323	61	71.22	71.22	NUM
cana-4362	323	62	77.23	77.23	NUM
cana-4362	323	63	70.45	70.45	NUM
cana-4362	323	64	71.53	71.53	NUM
cana-4362	323	65	70.51	70.51	NUM
cana-4362	323	66	64	64	NUM
cana-4362	323	67	0.06	0.06	NUM
cana-4362	323	68	68.82	68.82	NUM
cana-4362	323	69	74.54	74.54	NUM
cana-4362	323	70	67.12	67.12	NUM
cana-4362	323	71	67.54	67.54	NUM
cana-4362	323	72	67.21	67.21	NUM
cana-4362	323	73	250	250	NUM
cana-4362	323	74	32	32	NUM
cana-4362	323	75	0.04	0.04	NUM
cana-4362	323	76	77.88	77.88	NUM
cana-4362	323	77	85.40	85.40	NUM
cana-4362	323	78	78.76	78.76	NUM
cana-4362	323	79	78.83	78.83	NUM
cana-4362	323	80	78.72	78.72	NUM
cana-4362	323	81	64	64	NUM
cana-4362	323	82	0.04	0.04	NUM
cana-4362	323	83	76.79	76.79	NUM
cana-4362	323	84	84.53	84.53	NUM
cana-4362	323	85	76.93	76.93	NUM
cana-4362	323	86	76.70	76.70	NUM
cana-4362	323	87	77.60	77.60	NUM
cana-4362	323	88	512	512	NUM
cana-4362	323	89	32	32	NUM
cana-4362	323	90	0.04	0.04	NUM
cana-4362	323	91	83.1	83.1	NUM
cana-4362	323	92	88.32	88.32	NUM
cana-4362	323	93	80.28	80.28	NUM
cana-4362	323	94	79.81	79.81	NUM
cana-4362	323	95	80.12	80.12	NUM
cana-4362	323	96	64	64	NUM
cana-4362	323	97	0.04	0.04	NUM
cana-4362	323	98	79.13	79.13	NUM
cana-4362	323	99	86.64	86.64	NUM
cana-4362	323	100	78.29	78.29	NUM
cana-4362	323	101	79.83	79.83	NUM
cana-4362	323	102	79.84	79.84	NUM
cana-4362	323	103	rmsprop	rmsprop	NOUN
cana-4362	323	104	50	50	NUM
cana-4362	323	105	32	32	NUM
cana-4362	323	106	0.03	0.03	NUM
cana-4362	323	107	83.96	83.96	NUM
cana-4362	323	108	84.76	84.76	NUM
cana-4362	323	109	84.34	84.34	NUM
cana-4362	323	110	84.21	84.21	NUM
cana-4362	323	111	84.25	84.25	NUM
cana-4362	323	112	64	64	NUM
cana-4362	323	113	0.03	0.03	NUM
cana-4362	323	114	83.97	83.97	NUM
cana-4362	323	115	92.78	92.78	NUM
cana-4362	323	116	84.76	84.76	NUM
cana-4362	323	117	84.64	84.64	NUM
cana-4362	323	118	85.21	85.21	NUM
cana-4362	323	119	250	250	NUM
cana-4362	323	120	32	32	NUM
cana-4362	323	121	0.02	0.02	NUM
cana-4362	323	122	99.0	99.0	NUM
cana-4362	323	123	.	.	PUNCT
cana-4362	324	1	97.51	97.51	NUM
cana-4362	324	2	90.84	90.84	NUM
cana-4362	324	3	91.73	91.73	NUM
cana-4362	324	4	90.87	90.87	NUM
cana-4362	324	5	64	64	NUM
cana-4362	324	6	0.02	0.02	NUM
cana-4362	324	7	97.9	97.9	NUM
cana-4362	324	8	97.51	97.51	NUM
cana-4362	324	9	90.84	90.84	NUM
cana-4362	324	10	91.76	91.76	NUM
cana-4362	324	11	90.70	90.70	NUM
cana-4362	324	12	512	512	NUM
cana-4362	324	13	32	32	NUM
cana-4362	324	14	0.02	0.02	NUM
cana-4362	324	15	92.40	92.40	NUM
cana-4362	324	16	97.43	97.43	NUM
cana-4362	324	17	92.44	92.44	NUM
cana-4362	324	18	92.30	92.30	NUM
cana-4362	324	19	92.38	92.38	NUM
cana-4362	324	20	64	64	NUM
cana-4362	324	21	0.01	0.01	NUM
cana-4362	324	22	98.7	98.7	NUM
cana-4362	324	23	98.1	98.1	NUM
cana-4362	324	24	96.21	96.21	NUM
cana-4362	324	25	97.91	97.91	NUM
cana-4362	324	26	97.92	97.92	NUM
cana-4362	324	27	adagrad	adagrad	ADJ
cana-4362	324	28	50	50	NUM
cana-4362	324	29	32	32	NUM
cana-4362	324	30	0.04	0.04	NUM
cana-4362	324	31	79.29	79.29	NUM
cana-4362	324	32	8762	8762	NUM
cana-4362	324	33	79.17	79.17	NUM
cana-4362	324	34	78.94	78.94	NUM
cana-4362	324	35	78.84	78.84	NUM
cana-4362	324	36	64	64	NUM
cana-4362	324	37	0.04	0.04	NUM
cana-4362	324	38	78.14	78.14	NUM
cana-4362	324	39	8543	8543	NUM
cana-4362	324	40	78.87	78.87	NUM
cana-4362	325	1	77.72	77.72	NUM
cana-4362	325	2	78.79	78.79	NUM
cana-4362	325	3	250	250	NUM
cana-4362	325	4	32	32	NUM
cana-4362	325	5	0.03	0.03	NUM
cana-4362	325	6	85.78	85.78	NUM
cana-4362	325	7	93.36	93.36	NUM
cana-4362	325	8	86.62	86.62	NUM
cana-4362	325	9	85.28	85.28	NUM
cana-4362	325	10	85.53	85.53	NUM
cana-4362	325	11	64	64	NUM
cana-4362	325	12	0.03	0.03	NUM
cana-4362	325	13	81.96	81.96	NUM
cana-4362	325	14	90.16	90.16	NUM
cana-4362	325	15	82.72	82.72	NUM
cana-4362	325	16	82.18	82.18	NUM
cana-4362	325	17	81.84	81.84	NUM
cana-4362	325	18	512	512	NUM
cana-4362	325	19	32	32	NUM
cana-4362	325	20	0.02	0.02	NUM
cana-4362	325	21	95.8	95.8	NUM
cana-4362	325	22	97.14	97.14	NUM
cana-4362	325	23	90.78	90.78	NUM
cana-4362	325	24	91.68	91.68	NUM
cana-4362	325	25	90.77	90.77	NUM
cana-4362	325	26	64	64	NUM
cana-4362	325	27	0.03	0.03	NUM
cana-4362	325	28	87.86	87.86	NUM
cana-4362	325	29	94.50	94.50	NUM
cana-4362	325	30	87.90	87.90	NUM
cana-4362	325	31	88.73	88.73	NUM
cana-4362	325	32	88.7	88.7	NUM
cana-4362	325	33	communications	communication	NOUN
cana-4362	325	34	on	on	ADP
cana-4362	325	35	applied	apply	VERB
cana-4362	325	36	nonlinear	nonlinear	ADJ
cana-4362	325	37	analysis	analysis	NOUN
cana-4362	325	38	issn	issn	NOUN
cana-4362	325	39	:	:	PUNCT
cana-4362	325	40	1074	1074	NUM
cana-4362	325	41	-	-	PUNCT
cana-4362	325	42	133x	133x	NUM
cana-4362	325	43	vol	vol	NOUN
cana-4362	325	44	32	32	NUM
cana-4362	325	45	no	no	NOUN
cana-4362	325	46	.	.	PUNCT
cana-4362	326	1	9s	9s	NUM
cana-4362	326	2	(	(	PUNCT
cana-4362	326	3	2025	2025	NUM
cana-4362	326	4	)	)	PUNCT
cana-4362	326	5	1915	1915	NUM
cana-4362	326	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-4362	327	1	the	the	DET
cana-4362	327	2	above	above	ADJ
cana-4362	327	3	results	result	NOUN
cana-4362	327	4	indicate	indicate	VERB
cana-4362	327	5	that	that	SCONJ
cana-4362	327	6	hybrid	hybrid	ADJ
cana-4362	327	7	feature	feature	NOUN
cana-4362	327	8	selection	selection	NOUN
cana-4362	327	9	techniques	technique	NOUN
cana-4362	327	10	significantly	significantly	ADV
cana-4362	327	11	enhance	enhance	VERB
cana-4362	327	12	the	the	DET
cana-4362	327	13	performance	performance	NOUN
cana-4362	327	14	of	of	ADP
cana-4362	327	15	sentiment	sentiment	NOUN
cana-4362	327	16	classification	classification	NOUN
cana-4362	327	17	models	model	NOUN
cana-4362	327	18	.	.	PUNCT
cana-4362	328	1	key	key	ADJ
cana-4362	328	2	insights	insight	NOUN
cana-4362	328	3	from	from	ADP
cana-4362	328	4	the	the	DET
cana-4362	328	5	study	study	NOUN
cana-4362	328	6	include	include	VERB
cana-4362	328	7	:	:	PUNCT
cana-4362	328	8	feature	feature	NOUN
cana-4362	328	9	importance	importance	NOUN
cana-4362	328	10	:	:	PUNCT
cana-4362	328	11	the	the	DET
cana-4362	328	12	combination	combination	NOUN
cana-4362	328	13	of	of	ADP
cana-4362	328	14	statistical	statistical	ADJ
cana-4362	328	15	and	and	CCONJ
cana-4362	328	16	model	model	NOUN
cana-4362	328	17	-	-	PUNCT
cana-4362	328	18	based	base	VERB
cana-4362	328	19	feature	feature	NOUN
cana-4362	328	20	selection	selection	NOUN
cana-4362	328	21	provided	provide	VERB
cana-4362	328	22	a	a	DET
cana-4362	328	23	balanced	balanced	ADJ
cana-4362	328	24	approach	approach	NOUN
cana-4362	328	25	,	,	PUNCT
cana-4362	328	26	allowing	allow	VERB
cana-4362	328	27	models	model	NOUN
cana-4362	328	28	to	to	PART
cana-4362	328	29	focus	focus	VERB
cana-4362	328	30	on	on	ADP
cana-4362	328	31	the	the	DET
cana-4362	328	32	most	most	ADV
cana-4362	328	33	relevant	relevant	ADJ
cana-4362	328	34	features	feature	NOUN
cana-4362	328	35	while	while	SCONJ
cana-4362	328	36	reducing	reduce	VERB
cana-4362	328	37	noise	noise	NOUN
cana-4362	328	38	from	from	ADP
cana-4362	328	39	irrelevant	irrelevant	ADJ
cana-4362	328	40	ones	one	NOUN
cana-4362	328	41	.	.	PUNCT
cana-4362	329	1	deep	deep	ADJ
cana-4362	329	2	learning	learning	NOUN
cana-4362	329	3	model	model	NOUN
cana-4362	329	4	efficacy	efficacy	NOUN
cana-4362	329	5	:	:	PUNCT
cana-4362	329	6	models	model	NOUN
cana-4362	329	7	like	like	ADP
cana-4362	329	8	rnn	rnn	PROPN
cana-4362	329	9	-	-	PUNCT
cana-4362	329	10	lstm	lstm	NOUN
cana-4362	329	11	outperformed	outperform	VERB
cana-4362	329	12	traditional	traditional	ADJ
cana-4362	329	13	deep	deep	ADJ
cana-4362	329	14	learning	learning	NOUN
cana-4362	329	15	methods	method	NOUN
cana-4362	329	16	,	,	PUNCT
cana-4362	329	17	confirming	confirm	VERB
cana-4362	329	18	the	the	DET
cana-4362	329	19	advantage	advantage	NOUN
cana-4362	329	20	of	of	ADP
cana-4362	329	21	deep	deep	ADJ
cana-4362	329	22	learning	learning	NOUN
cana-4362	329	23	in	in	ADP
cana-4362	329	24	understanding	understand	VERB
cana-4362	329	25	the	the	DET
cana-4362	329	26	nuances	nuance	NOUN
cana-4362	329	27	of	of	ADP
cana-4362	329	28	natural	natural	ADJ
cana-4362	329	29	language	language	NOUN
cana-4362	329	30	.	.	PUNCT
cana-4362	330	1	real	real	ADJ
cana-4362	330	2	-	-	PUNCT
cana-4362	330	3	world	world	NOUN
cana-4362	330	4	implications	implication	NOUN
cana-4362	330	5	:	:	PUNCT
cana-4362	330	6	this	this	DET
cana-4362	330	7	study	study	NOUN
cana-4362	330	8	's	's	PART
cana-4362	330	9	findings	finding	NOUN
cana-4362	330	10	highlight	highlight	VERB
cana-4362	330	11	the	the	DET
cana-4362	330	12	potential	potential	NOUN
cana-4362	330	13	for	for	ADP
cana-4362	330	14	using	use	VERB
cana-4362	330	15	deep	deep	ADJ
cana-4362	330	16	learning	learning	NOUN
cana-4362	330	17	and	and	CCONJ
cana-4362	330	18	hybrid	hybrid	ADJ
cana-4362	330	19	feature	feature	NOUN
cana-4362	330	20	selection	selection	NOUN
cana-4362	330	21	techniques	technique	NOUN
cana-4362	330	22	in	in	ADP
cana-4362	330	23	applications	application	NOUN
cana-4362	330	24	such	such	ADJ
cana-4362	330	25	as	as	ADP
cana-4362	330	26	brand	brand	NOUN
cana-4362	330	27	monitoring	monitoring	NOUN
cana-4362	330	28	,	,	PUNCT
cana-4362	330	29	customer	customer	NOUN
cana-4362	330	30	feedback	feedback	NOUN
cana-4362	330	31	analysis	analysis	NOUN
cana-4362	330	32	,	,	PUNCT
cana-4362	330	33	and	and	CCONJ
cana-4362	330	34	social	social	ADJ
cana-4362	330	35	media	medium	NOUN
cana-4362	330	36	sentiment	sentiment	NOUN
cana-4362	330	37	tracking	track	VERB
cana-4362	330	38	.	.	PUNCT
cana-4362	331	1	the	the	DET
cana-4362	331	2	integration	integration	NOUN
cana-4362	331	3	of	of	ADP
cana-4362	331	4	hybrid	hybrid	ADJ
cana-4362	331	5	feature	feature	NOUN
cana-4362	331	6	selection	selection	NOUN
cana-4362	331	7	methods	method	NOUN
cana-4362	331	8	with	with	ADP
cana-4362	331	9	advanced	advanced	ADJ
cana-4362	331	10	deep	deep	ADJ
cana-4362	331	11	learning	learning	NOUN
cana-4362	331	12	models	model	NOUN
cana-4362	331	13	significantly	significantly	ADV
cana-4362	331	14	improves	improve	VERB
cana-4362	331	15	the	the	DET
cana-4362	331	16	accuracy	accuracy	NOUN
cana-4362	331	17	and	and	CCONJ
cana-4362	331	18	reliability	reliability	NOUN
cana-4362	331	19	of	of	ADP
cana-4362	331	20	sentiment	sentiment	NOUN
cana-4362	331	21	classification	classification	NOUN
cana-4362	331	22	on	on	ADP
cana-4362	331	23	social	social	ADJ
cana-4362	331	24	media	medium	NOUN
cana-4362	331	25	datasets	dataset	NOUN
cana-4362	331	26	.	.	PUNCT
cana-4362	332	1	future	future	ADJ
cana-4362	332	2	work	work	NOUN
cana-4362	332	3	could	could	AUX
cana-4362	332	4	explore	explore	VERB
cana-4362	332	5	the	the	DET
cana-4362	332	6	impact	impact	NOUN
cana-4362	332	7	of	of	ADP
cana-4362	332	8	more	more	ADV
cana-4362	332	9	diverse	diverse	ADJ
cana-4362	332	10	datasets	dataset	NOUN
cana-4362	332	11	and	and	CCONJ
cana-4362	332	12	additional	additional	ADJ
cana-4362	332	13	feature	feature	NOUN
cana-4362	332	14	selection	selection	NOUN
cana-4362	332	15	techniques	technique	NOUN
cana-4362	332	16	to	to	PART
cana-4362	332	17	further	far	ADV
cana-4362	332	18	enhance	enhance	VERB
cana-4362	332	19	model	model	NOUN
cana-4362	332	20	performance	performance	NOUN
cana-4362	332	21	.	.	PUNCT
cana-4362	333	1	conclusion	conclusion	NOUN
cana-4362	333	2	and	and	CCONJ
cana-4362	333	3	future	future	ADJ
cana-4362	333	4	score	score	NOUN
cana-4362	333	5	in	in	ADP
cana-4362	333	6	this	this	DET
cana-4362	333	7	study	study	NOUN
cana-4362	333	8	,	,	PUNCT
cana-4362	333	9	we	we	PRON
cana-4362	333	10	explored	explore	VERB
cana-4362	333	11	the	the	DET
cana-4362	333	12	application	application	NOUN
cana-4362	333	13	of	of	ADP
cana-4362	333	14	deep	deep	ADJ
cana-4362	333	15	learning	learning	NOUN
cana-4362	333	16	techniques	technique	NOUN
cana-4362	333	17	for	for	ADP
cana-4362	333	18	sentiment	sentiment	NOUN
cana-4362	333	19	classification	classification	NOUN
cana-4362	333	20	on	on	ADP
cana-4362	333	21	social	social	ADJ
cana-4362	333	22	media	medium	NOUN
cana-4362	333	23	datasets	dataset	NOUN
cana-4362	333	24	,	,	PUNCT
cana-4362	333	25	highlighting	highlight	VERB
cana-4362	333	26	their	their	PRON
cana-4362	333	27	effectiveness	effectiveness	NOUN
cana-4362	333	28	in	in	ADP
cana-4362	333	29	understanding	understand	VERB
cana-4362	333	30	user	user	NOUN
cana-4362	333	31	sentiments	sentiment	NOUN
cana-4362	333	32	expressed	express	VERB
cana-4362	333	33	in	in	ADP
cana-4362	333	34	online	online	ADJ
cana-4362	333	35	platforms	platform	NOUN
cana-4362	333	36	.	.	PUNCT
cana-4362	334	1	our	our	PRON
cana-4362	334	2	experiments	experiment	NOUN
cana-4362	334	3	demonstrated	demonstrate	VERB
cana-4362	334	4	that	that	SCONJ
cana-4362	334	5	advanced	advanced	ADJ
cana-4362	334	6	models	model	NOUN
cana-4362	334	7	,	,	PUNCT
cana-4362	334	8	such	such	ADJ
cana-4362	334	9	as	as	ADP
cana-4362	334	10	rnn	rnn	NOUN
cana-4362	334	11	-	-	PUNCT
cana-4362	334	12	lstm	lstm	ADJ
cana-4362	334	13	networks	network	NOUN
cana-4362	334	14	,	,	PUNCT
cana-4362	334	15	significantly	significantly	ADV
cana-4362	334	16	outperformed	outperform	VERB
cana-4362	334	17	traditional	traditional	ADJ
cana-4362	334	18	machine	machine	NOUN
cana-4362	334	19	learning	learning	NOUN
cana-4362	334	20	approaches	approach	NOUN
cana-4362	334	21	,	,	PUNCT
cana-4362	334	22	achieving	achieve	VERB
cana-4362	334	23	higher	high	ADJ
cana-4362	334	24	accuracy	accuracy	NOUN
cana-4362	334	25	and	and	CCONJ
cana-4362	334	26	f1	f1	NOUN
cana-4362	334	27	scores	score	NOUN
cana-4362	334	28	.	.	PUNCT
cana-4362	335	1	the	the	DET
cana-4362	335	2	use	use	NOUN
cana-4362	335	3	of	of	ADP
cana-4362	335	4	hybrid	hybrid	ADJ
cana-4362	335	5	feature	feature	NOUN
cana-4362	335	6	selection	selection	NOUN
cana-4362	335	7	methods	method	NOUN
cana-4362	335	8	further	far	ADV
cana-4362	335	9	enhanced	enhance	VERB
cana-4362	335	10	model	model	NOUN
cana-4362	335	11	performance	performance	NOUN
cana-4362	335	12	by	by	ADP
cana-4362	335	13	reducing	reduce	VERB
cana-4362	335	14	dimensionality	dimensionality	NOUN
cana-4362	335	15	and	and	CCONJ
cana-4362	335	16	retaining	retain	VERB
cana-4362	335	17	only	only	ADV
cana-4362	335	18	the	the	DET
cana-4362	335	19	most	most	ADV
cana-4362	335	20	informative	informative	ADJ
cana-4362	335	21	features	feature	NOUN
cana-4362	335	22	.	.	PUNCT
cana-4362	336	1	these	these	DET
cana-4362	336	2	findings	finding	NOUN
cana-4362	336	3	underscore	underscore	VERB
cana-4362	336	4	the	the	DET
cana-4362	336	5	potential	potential	NOUN
cana-4362	336	6	of	of	ADP
cana-4362	336	7	deep	deep	ADJ
cana-4362	336	8	learning	learning	NOUN
cana-4362	336	9	in	in	ADP
cana-4362	336	10	sentiment	sentiment	NOUN
cana-4362	336	11	analysis	analysis	NOUN
cana-4362	336	12	,	,	PUNCT
cana-4362	336	13	particularly	particularly	ADV
cana-4362	336	14	in	in	ADP
cana-4362	336	15	capturing	capture	VERB
cana-4362	336	16	the	the	DET
cana-4362	336	17	nuanced	nuanced	ADJ
cana-4362	336	18	emotional	emotional	ADJ
cana-4362	336	19	tones	tone	NOUN
cana-4362	336	20	embedded	embed	VERB
cana-4362	336	21	in	in	ADP
cana-4362	336	22	user	user	NOUN
cana-4362	336	23	-	-	PUNCT
cana-4362	336	24	generated	generate	VERB
cana-4362	336	25	content	content	NOUN
cana-4362	336	26	.	.	PUNCT
cana-4362	337	1	the	the	DET
cana-4362	337	2	results	result	NOUN
cana-4362	337	3	indicate	indicate	VERB
cana-4362	337	4	that	that	DET
cana-4362	337	5	sentiment	sentiment	NOUN
cana-4362	337	6	classification	classification	NOUN
cana-4362	337	7	can	can	AUX
cana-4362	337	8	be	be	AUX
cana-4362	337	9	reliably	reliably	ADV
cana-4362	337	10	performed	perform	VERB
cana-4362	337	11	using	use	VERB
cana-4362	337	12	deep	deep	ADJ
cana-4362	337	13	learning	learning	NOUN
cana-4362	337	14	techniques	technique	NOUN
cana-4362	337	15	,	,	PUNCT
cana-4362	337	16	providing	provide	VERB
cana-4362	337	17	valuable	valuable	ADJ
cana-4362	337	18	insights	insight	NOUN
cana-4362	337	19	for	for	ADP
cana-4362	337	20	businesses	business	NOUN
cana-4362	337	21	,	,	PUNCT
cana-4362	337	22	researchers	researcher	NOUN
cana-4362	337	23	,	,	PUNCT
cana-4362	337	24	and	and	CCONJ
cana-4362	337	25	policymakers	policymaker	NOUN
cana-4362	337	26	.	.	PUNCT
cana-4362	338	1	the	the	DET
cana-4362	338	2	ability	ability	NOUN
cana-4362	338	3	to	to	PART
cana-4362	338	4	automate	automate	VERB
cana-4362	338	5	sentiment	sentiment	NOUN
cana-4362	338	6	analysis	analysis	NOUN
cana-4362	338	7	facilitates	facilitate	VERB
cana-4362	338	8	real	real	ADJ
cana-4362	338	9	-	-	PUNCT
cana-4362	338	10	time	time	NOUN
cana-4362	338	11	monitoring	monitoring	NOUN
cana-4362	338	12	of	of	ADP
cana-4362	338	13	public	public	ADJ
cana-4362	338	14	opinion	opinion	NOUN
cana-4362	338	15	,	,	PUNCT
cana-4362	338	16	enabling	enable	VERB
cana-4362	338	17	swift	swift	ADJ
cana-4362	338	18	responses	response	NOUN
cana-4362	338	19	to	to	ADP
cana-4362	338	20	emerging	emerge	VERB
cana-4362	338	21	trends	trend	NOUN
cana-4362	338	22	or	or	CCONJ
cana-4362	338	23	crises	crisis	NOUN
cana-4362	338	24	.	.	PUNCT
cana-4362	339	1	future	future	ADJ
cana-4362	339	2	research	research	NOUN
cana-4362	339	3	should	should	AUX
cana-4362	339	4	focus	focus	VERB
cana-4362	339	5	on	on	ADP
cana-4362	339	6	several	several	ADJ
cana-4362	339	7	key	key	ADJ
cana-4362	339	8	areas	area	NOUN
cana-4362	339	9	to	to	PART
cana-4362	339	10	enhance	enhance	VERB
cana-4362	339	11	sentiment	sentiment	NOUN
cana-4362	339	12	classification	classification	NOUN
cana-4362	339	13	models	model	NOUN
cana-4362	339	14	further	far	ADV
cana-4362	339	15	.	.	PUNCT
cana-4362	340	1	firstly	firstly	ADV
cana-4362	340	2	,	,	PUNCT
cana-4362	340	3	expanding	expand	VERB
cana-4362	340	4	the	the	DET
cana-4362	340	5	diversity	diversity	NOUN
cana-4362	340	6	and	and	CCONJ
cana-4362	340	7	volume	volume	NOUN
cana-4362	340	8	of	of	ADP
cana-4362	340	9	training	training	NOUN
cana-4362	340	10	datasets	dataset	NOUN
cana-4362	340	11	will	will	AUX
cana-4362	340	12	improve	improve	VERB
cana-4362	340	13	model	model	NOUN
cana-4362	340	14	robustness	robustness	NOUN
cana-4362	340	15	and	and	CCONJ
cana-4362	340	16	generalizability	generalizability	NOUN
cana-4362	340	17	.	.	PUNCT
cana-4362	341	1	incorporating	incorporate	VERB
cana-4362	341	2	multilingual	multilingual	ADJ
cana-4362	341	3	capabilities	capability	NOUN
cana-4362	341	4	can	can	AUX
cana-4362	341	5	also	also	ADV
cana-4362	341	6	broaden	broaden	VERB
cana-4362	341	7	the	the	DET
cana-4362	341	8	applicability	applicability	NOUN
cana-4362	341	9	of	of	ADP
cana-4362	341	10	sentiment	sentiment	NOUN
cana-4362	341	11	analysis	analysis	NOUN
cana-4362	341	12	across	across	ADP
cana-4362	341	13	different	different	ADJ
cana-4362	341	14	languages	language	NOUN
cana-4362	341	15	and	and	CCONJ
cana-4362	341	16	cultures	culture	NOUN
cana-4362	341	17	,	,	PUNCT
cana-4362	341	18	making	make	VERB
cana-4362	341	19	the	the	DET
cana-4362	341	20	models	model	NOUN
cana-4362	341	21	more	more	ADV
cana-4362	341	22	inclusive	inclusive	ADJ
cana-4362	341	23	.	.	PUNCT
cana-4362	342	1	finally	finally	ADV
cana-4362	342	2	,	,	PUNCT
cana-4362	342	3	integrating	integrate	VERB
cana-4362	342	4	multimodal	multimodal	NOUN
cana-4362	342	5	data	datum	NOUN
cana-4362	342	6	,	,	PUNCT
cana-4362	342	7	such	such	ADJ
cana-4362	342	8	as	as	ADP
cana-4362	342	9	images	image	NOUN
cana-4362	342	10	and	and	CCONJ
cana-4362	342	11	videos	video	NOUN
cana-4362	342	12	alongside	alongside	ADP
cana-4362	342	13	text	text	NOUN
cana-4362	342	14	,	,	PUNCT
cana-4362	342	15	can	can	AUX
cana-4362	342	16	provide	provide	VERB
cana-4362	342	17	a	a	DET
cana-4362	342	18	more	more	ADV
cana-4362	342	19	holistic	holistic	ADJ
cana-4362	342	20	understanding	understanding	NOUN
cana-4362	342	21	of	of	ADP
cana-4362	342	22	sentiment	sentiment	NOUN
cana-4362	342	23	,	,	PUNCT
cana-4362	342	24	capturing	capture	VERB
cana-4362	342	25	contextual	contextual	ADJ
cana-4362	342	26	cues	cue	NOUN
cana-4362	342	27	that	that	PRON
cana-4362	342	28	text	text	NOUN
cana-4362	342	29	alone	alone	ADV
cana-4362	342	30	may	may	AUX
cana-4362	342	31	miss	miss	VERB
cana-4362	342	32	.	.	PUNCT
cana-4362	343	1	by	by	ADP
cana-4362	343	2	addressing	address	VERB
cana-4362	343	3	these	these	DET
cana-4362	343	4	areas	area	NOUN
cana-4362	343	5	,	,	PUNCT
cana-4362	343	6	future	future	ADJ
cana-4362	343	7	studies	study	NOUN
cana-4362	343	8	can	can	AUX
cana-4362	343	9	advance	advance	VERB
cana-4362	343	10	the	the	DET
cana-4362	343	11	field	field	NOUN
cana-4362	343	12	of	of	ADP
cana-4362	343	13	sentiment	sentiment	NOUN
cana-4362	343	14	classification	classification	NOUN
cana-4362	343	15	,	,	PUNCT
cana-4362	343	16	paving	pave	VERB
cana-4362	343	17	the	the	DET
cana-4362	343	18	way	way	NOUN
cana-4362	343	19	for	for	ADP
cana-4362	343	20	more	more	ADV
cana-4362	343	21	sophisticated	sophisticated	ADJ
cana-4362	343	22	and	and	CCONJ
cana-4362	343	23	accurate	accurate	ADJ
cana-4362	343	24	analyses	analysis	NOUN
cana-4362	343	25	in	in	ADP
cana-4362	343	26	the	the	DET
cana-4362	343	27	dynamic	dynamic	ADJ
cana-4362	343	28	realm	realm	NOUN
cana-4362	343	29	of	of	ADP
cana-4362	343	30	social	social	ADJ
cana-4362	343	31	media	medium	NOUN
cana-4362	343	32	.	.	PUNCT
cana-4362	344	1	references	reference	NOUN
cana-4362	344	2	1	1	NUM
cana-4362	344	3	.	.	X
cana-4362	345	1	zhang	zhang	PROPN
cana-4362	345	2	,	,	PUNCT
cana-4362	345	3	z.	z.	PROPN
cana-4362	345	4	,	,	PUNCT
cana-4362	345	5	liu	liu	PROPN
cana-4362	345	6	,	,	PUNCT
cana-4362	345	7	x.	x.	PROPN
cana-4362	345	8	,	,	PUNCT
cana-4362	345	9	sun	sun	PROPN
cana-4362	345	10	,	,	PUNCT
cana-4362	345	11	j.	j.	PROPN
cana-4362	345	12	,	,	PUNCT
cana-4362	345	13	&	&	CCONJ
cana-4362	345	14	wang	wang	PROPN
cana-4362	345	15	,	,	PUNCT
cana-4362	345	16	h.	h.	PROPN
cana-4362	345	17	(	(	PUNCT
cana-4362	345	18	2023	2023	NUM
cana-4362	345	19	)	)	PUNCT
cana-4362	345	20	.	.	PUNCT
cana-4362	346	1	a	a	DET
cana-4362	346	2	hybrid	hybrid	ADJ
cana-4362	346	3	feature	feature	NOUN
cana-4362	346	4	selection	selection	NOUN
cana-4362	346	5	method	method	NOUN
cana-4362	346	6	for	for	ADP
cana-4362	346	7	sentiment	sentiment	NOUN
cana-4362	346	8	classification	classification	NOUN
cana-4362	346	9	based	base	VERB
cana-4362	346	10	on	on	ADP
cana-4362	346	11	social	social	ADJ
cana-4362	346	12	media	medium	NOUN
cana-4362	346	13	data	datum	NOUN
cana-4362	346	14	.	.	PUNCT
cana-4362	347	1	ieee	ieee	NOUN
cana-4362	347	2	transactions	transaction	NOUN
cana-4362	347	3	on	on	ADP
cana-4362	347	4	cybernetics	cybernetic	NOUN
cana-4362	347	5	,	,	PUNCT
cana-4362	347	6	53(5	53(5	NUM
cana-4362	347	7	)	)	PUNCT
cana-4362	347	8	,	,	PUNCT
cana-4362	347	9	1827	1827	NUM
cana-4362	347	10	-	-	SYM
cana-4362	347	11	1839	1839	NUM
cana-4362	347	12	.	.	PUNCT
cana-4362	348	1	https://doi.org/10.1109/tcyb.2022.3165441	https://doi.org/10.1109/tcyb.2022.3165441	PROPN
cana-4362	348	2	communications	communication	NOUN
cana-4362	348	3	on	on	ADP
cana-4362	348	4	applied	apply	VERB
cana-4362	348	5	nonlinear	nonlinear	ADJ
cana-4362	348	6	analysis	analysis	NOUN
cana-4362	348	7	issn	issn	NOUN
cana-4362	348	8	:	:	PUNCT
cana-4362	348	9	1074	1074	NUM
cana-4362	348	10	-	-	PUNCT
cana-4362	348	11	133x	133x	NUM
cana-4362	348	12	vol	vol	NOUN
cana-4362	348	13	32	32	NUM
cana-4362	348	14	no	no	NOUN
cana-4362	348	15	.	.	PUNCT
cana-4362	349	1	9s	9s	NUM
cana-4362	349	2	(	(	PUNCT
cana-4362	349	3	2025	2025	NUM
cana-4362	349	4	)	)	PUNCT
cana-4362	349	5	1916	1916	NUM
cana-4362	349	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-4362	349	7	2	2	X
cana-4362	349	8	.	.	X
cana-4362	350	1	chen	chen	PROPN
cana-4362	350	2	,	,	PUNCT
cana-4362	350	3	y.	y.	PROPN
cana-4362	350	4	,	,	PUNCT
cana-4362	350	5	wang	wang	PROPN
cana-4362	350	6	,	,	PUNCT
cana-4362	350	7	l.	l.	PROPN
cana-4362	350	8	,	,	PUNCT
cana-4362	350	9	&	&	CCONJ
cana-4362	350	10	zhao	zhao	PROPN
cana-4362	350	11	,	,	PUNCT
cana-4362	350	12	x.	x.	NOUN
cana-4362	350	13	(	(	PUNCT
cana-4362	350	14	2023	2023	NUM
cana-4362	350	15	)	)	PUNCT
cana-4362	350	16	.	.	PUNCT
cana-4362	351	1	hybrid	hybrid	ADJ
cana-4362	351	2	deep	deep	ADJ
cana-4362	351	3	learning	learning	NOUN
cana-4362	351	4	and	and	CCONJ
cana-4362	351	5	feature	feature	NOUN
cana-4362	351	6	selection	selection	NOUN
cana-4362	351	7	for	for	ADP
cana-4362	351	8	social	social	ADJ
cana-4362	351	9	media	medium	NOUN
cana-4362	351	10	sentiment	sentiment	NOUN
cana-4362	351	11	classification	classification	NOUN
cana-4362	351	12	.	.	PUNCT
cana-4362	352	1	ieee	ieee	NOUN
cana-4362	352	2	transactions	transaction	NOUN
cana-4362	352	3	on	on	ADP
cana-4362	352	4	neural	neural	ADJ
cana-4362	352	5	networks	network	NOUN
cana-4362	352	6	and	and	CCONJ
cana-4362	352	7	learning	learn	VERB
cana-4362	352	8	systems	system	NOUN
cana-4362	352	9	,	,	PUNCT
cana-4362	352	10	34(1	34(1	NUM
cana-4362	352	11	)	)	PUNCT
cana-4362	352	12	,	,	PUNCT
cana-4362	352	13	45	45	NUM
cana-4362	352	14	-	-	SYM
cana-4362	352	15	56	56	NUM
cana-4362	352	16	.	.	PUNCT
cana-4362	353	1	https://doi.org/10.1109/tnnls.2022.3196647	https://doi.org/10.1109/tnnls.2022.3196647	PROPN
cana-4362	353	2	3	3	NUM
cana-4362	353	3	.	.	PUNCT
cana-4362	353	4	li	li	PROPN
cana-4362	353	5	,	,	PUNCT
cana-4362	353	6	h.	h.	PROPN
cana-4362	353	7	,	,	PUNCT
cana-4362	353	8	zhou	zhou	PROPN
cana-4362	353	9	,	,	PUNCT
cana-4362	353	10	z.	z.	PROPN
cana-4362	353	11	,	,	PUNCT
cana-4362	353	12	wang	wang	PROPN
cana-4362	353	13	,	,	PUNCT
cana-4362	353	14	s.	s.	PROPN
cana-4362	353	15	,	,	PUNCT
cana-4362	353	16	&	&	CCONJ
cana-4362	353	17	du	du	PROPN
cana-4362	353	18	,	,	PUNCT
cana-4362	353	19	q.	q.	PROPN
cana-4362	353	20	(	(	PUNCT
cana-4362	353	21	2022	2022	NUM
cana-4362	353	22	)	)	PUNCT
cana-4362	353	23	.	.	PUNCT
cana-4362	354	1	an	an	DET
cana-4362	354	2	efficient	efficient	ADJ
cana-4362	354	3	hybrid	hybrid	NOUN
cana-4362	354	4	approach	approach	NOUN
cana-4362	354	5	for	for	ADP
cana-4362	354	6	sentiment	sentiment	NOUN
cana-4362	354	7	analysis	analysis	NOUN
cana-4362	354	8	and	and	CCONJ
cana-4362	354	9	feature	feature	NOUN
cana-4362	354	10	selection	selection	NOUN
cana-4362	354	11	on	on	ADP
cana-4362	354	12	twitter	twitter	NOUN
cana-4362	354	13	data	data	PROPN
cana-4362	354	14	.	.	PUNCT
cana-4362	355	1	ieee	ieee	NOUN
cana-4362	355	2	transactions	transaction	NOUN
cana-4362	355	3	on	on	ADP
cana-4362	355	4	computational	computational	ADJ
cana-4362	355	5	social	social	ADJ
cana-4362	355	6	systems	system	NOUN
cana-4362	355	7	,	,	PUNCT
cana-4362	355	8	9(3	9(3	NUM
cana-4362	355	9	)	)	PUNCT
cana-4362	355	10	,	,	PUNCT
cana-4362	355	11	401	401	NUM
cana-4362	355	12	-	-	SYM
cana-4362	355	13	412	412	NUM
cana-4362	355	14	.	.	PUNCT
cana-4362	356	1	https://doi.org/10.1109/tcss.2022.3154488	https://doi.org/10.1109/tcss.2022.3154488	PROPN
cana-4362	356	2	4	4	NUM
cana-4362	356	3	.	.	PUNCT
cana-4362	357	1	kumar	kumar	PROPN
cana-4362	357	2	,	,	PUNCT
cana-4362	357	3	r.	r.	PROPN
cana-4362	357	4	,	,	PUNCT
cana-4362	357	5	patel	patel	PROPN
cana-4362	357	6	,	,	PUNCT
cana-4362	357	7	a.	a.	PROPN
cana-4362	357	8	,	,	PUNCT
cana-4362	357	9	&	&	CCONJ
cana-4362	357	10	verma	verma	PROPN
cana-4362	357	11	,	,	PUNCT
cana-4362	357	12	m.	m.	NOUN
cana-4362	357	13	(	(	PUNCT
cana-4362	357	14	2023	2023	NUM
cana-4362	357	15	)	)	PUNCT
cana-4362	357	16	.	.	PUNCT
cana-4362	358	1	feature	feature	NOUN
cana-4362	358	2	selection	selection	NOUN
cana-4362	358	3	for	for	ADP
cana-4362	358	4	sentiment	sentiment	NOUN
cana-4362	358	5	classification	classification	NOUN
cana-4362	358	6	using	use	VERB
cana-4362	358	7	deep	deep	ADJ
cana-4362	358	8	learning	learning	NOUN
cana-4362	358	9	:	:	PUNCT
cana-4362	358	10	a	a	DET
cana-4362	358	11	hybrid	hybrid	ADJ
cana-4362	358	12	approach	approach	NOUN
cana-4362	358	13	.	.	PUNCT
cana-4362	359	1	ieee	ieee	NOUN
cana-4362	359	2	transactions	transaction	NOUN
cana-4362	359	3	on	on	ADP
cana-4362	359	4	knowledge	knowledge	NOUN
cana-4362	359	5	and	and	CCONJ
cana-4362	359	6	data	datum	NOUN
cana-4362	359	7	engineering	engineering	NOUN
cana-4362	359	8	,	,	PUNCT
cana-4362	359	9	35(2	35(2	NUM
cana-4362	359	10	)	)	PUNCT
cana-4362	359	11	,	,	PUNCT
cana-4362	359	12	1234	1234	NUM
cana-4362	359	13	-	-	SYM
cana-4362	359	14	1245	1245	NUM
cana-4362	359	15	.	.	PUNCT
cana-4362	360	1	https://doi.org/10.1109/tkde.2022.3184519	https://doi.org/10.1109/tkde.2022.3184519	NOUN
cana-4362	360	2	5	5	NUM
cana-4362	360	3	.	.	X
cana-4362	360	4	jiang	jiang	PROPN
cana-4362	360	5	,	,	PUNCT
cana-4362	360	6	x.	x.	PROPN
cana-4362	360	7	,	,	PUNCT
cana-4362	360	8	luo	luo	PROPN
cana-4362	360	9	,	,	PUNCT
cana-4362	360	10	q.	q.	PROPN
cana-4362	360	11	,	,	PUNCT
cana-4362	360	12	&	&	CCONJ
cana-4362	360	13	chen	chen	PROPN
cana-4362	360	14	,	,	PUNCT
cana-4362	360	15	b.	b.	PROPN
cana-4362	360	16	(	(	PUNCT
cana-4362	360	17	2022	2022	NUM
cana-4362	360	18	)	)	PUNCT
cana-4362	360	19	.	.	PUNCT
cana-4362	361	1	hybrid	hybrid	ADJ
cana-4362	361	2	feature	feature	NOUN
cana-4362	361	3	selection	selection	NOUN
cana-4362	361	4	with	with	ADP
cana-4362	361	5	deep	deep	ADJ
cana-4362	361	6	learning	learning	NOUN
cana-4362	361	7	for	for	ADP
cana-4362	361	8	sentiment	sentiment	NOUN
cana-4362	361	9	classification	classification	NOUN
cana-4362	361	10	on	on	ADP
cana-4362	361	11	social	social	ADJ
cana-4362	361	12	media	medium	NOUN
cana-4362	361	13	platforms	platform	NOUN
cana-4362	361	14	.	.	PUNCT
cana-4362	362	1	ieee	ieee	NOUN
cana-4362	362	2	transactions	transaction	NOUN
cana-4362	362	3	on	on	ADP
cana-4362	362	4	big	big	ADJ
cana-4362	362	5	data	datum	NOUN
cana-4362	362	6	,	,	PUNCT
cana-4362	362	7	8(2	8(2	NUM
cana-4362	362	8	)	)	PUNCT
cana-4362	362	9	,	,	PUNCT
cana-4362	362	10	295	295	NUM
cana-4362	362	11	-	-	SYM
cana-4362	362	12	308	308	NUM
cana-4362	362	13	.	.	PUNCT
cana-4362	363	1	https://doi.org/10.1109/tbdata.2021.3079985	https://doi.org/10.1109/tbdata.2021.3079985	PROPN
cana-4362	363	2	6	6	NUM
cana-4362	363	3	.	.	PUNCT
cana-4362	363	4	wang	wang	PROPN
cana-4362	363	5	,	,	PUNCT
cana-4362	363	6	j.	j.	PROPN
cana-4362	363	7	,	,	PUNCT
cana-4362	363	8	hu	hu	PROPN
cana-4362	363	9	,	,	PUNCT
cana-4362	363	10	s.	s.	PROPN
cana-4362	363	11	,	,	PUNCT
cana-4362	363	12	&	&	CCONJ
cana-4362	363	13	liu	liu	PROPN
cana-4362	363	14	,	,	PUNCT
cana-4362	363	15	y.	y.	PROPN
cana-4362	363	16	(	(	PUNCT
cana-4362	363	17	2023	2023	NUM
cana-4362	363	18	)	)	PUNCT
cana-4362	363	19	.	.	PUNCT
cana-4362	364	1	sentiment	sentiment	NOUN
cana-4362	364	2	analysis	analysis	NOUN
cana-4362	364	3	on	on	ADP
cana-4362	364	4	social	social	ADJ
cana-4362	364	5	media	medium	NOUN
cana-4362	364	6	data	datum	NOUN
cana-4362	364	7	using	use	VERB
cana-4362	364	8	hybrid	hybrid	ADJ
cana-4362	364	9	feature	feature	NOUN
cana-4362	364	10	selection	selection	NOUN
cana-4362	364	11	and	and	CCONJ
cana-4362	364	12	lstm	lstm	NOUN
cana-4362	364	13	networks	network	NOUN
cana-4362	364	14	.	.	PUNCT
cana-4362	365	1	ieee	ieee	NOUN
cana-4362	365	2	transactions	transaction	NOUN
cana-4362	365	3	on	on	ADP
cana-4362	365	4	affective	affective	ADJ
cana-4362	365	5	computing	computing	NOUN
cana-4362	365	6	,	,	PUNCT
cana-4362	365	7	14(1	14(1	NUM
cana-4362	365	8	)	)	PUNCT
cana-4362	365	9	,	,	PUNCT
cana-4362	365	10	67	67	NUM
cana-4362	365	11	-	-	SYM
cana-4362	365	12	78	78	NUM
cana-4362	365	13	.	.	PUNCT
cana-4362	366	1	https://doi.org/10.1109/taffc.2022.3170543	https://doi.org/10.1109/taffc.2022.3170543	ADJ
cana-4362	366	2	7	7	PROPN
cana-4362	366	3	.	.	PUNCT
cana-4362	366	4	chen	chen	PROPN
cana-4362	366	5	,	,	PUNCT
cana-4362	366	6	x.	x.	PROPN
cana-4362	366	7	,	,	PUNCT
cana-4362	366	8	liu	liu	PROPN
cana-4362	366	9	,	,	PUNCT
cana-4362	366	10	g.	g.	PROPN
cana-4362	366	11	,	,	PUNCT
cana-4362	366	12	zhang	zhang	PROPN
cana-4362	366	13	,	,	PUNCT
cana-4362	366	14	y.	y.	PROPN
cana-4362	366	15	,	,	PUNCT
cana-4362	366	16	&	&	CCONJ
cana-4362	366	17	xu	xu	PROPN
cana-4362	366	18	,	,	PUNCT
cana-4362	366	19	w.	w.	PROPN
cana-4362	366	20	(	(	PUNCT
cana-4362	366	21	2023	2023	NUM
cana-4362	366	22	)	)	PUNCT
cana-4362	366	23	.	.	PUNCT
cana-4362	367	1	multi	multi	ADJ
cana-4362	367	2	-	-	ADJ
cana-4362	367	3	modal	modal	ADJ
cana-4362	367	4	feature	feature	NOUN
cana-4362	367	5	selection	selection	NOUN
cana-4362	367	6	for	for	ADP
cana-4362	367	7	sentiment	sentiment	NOUN
cana-4362	367	8	classification	classification	NOUN
cana-4362	367	9	in	in	ADP
cana-4362	367	10	social	social	ADJ
cana-4362	367	11	media	medium	NOUN
cana-4362	367	12	text	text	NOUN
cana-4362	367	13	.	.	PUNCT
cana-4362	368	1	ieee	ieee	NOUN
cana-4362	368	2	transactions	transaction	NOUN
cana-4362	368	3	on	on	ADP
cana-4362	368	4	multimedia	multimedia	NOUN
cana-4362	368	5	,	,	PUNCT
cana-4362	368	6	25(4	25(4	NOUN
cana-4362	368	7	)	)	PUNCT
cana-4362	368	8	,	,	PUNCT
cana-4362	368	9	1035	1035	NUM
cana-4362	368	10	-	-	SYM
cana-4362	368	11	1047	1047	NUM
cana-4362	368	12	.	.	PUNCT
cana-4362	369	1	https://doi.org/10.1109/tmm.2023.3187964	https://doi.org/10.1109/tmm.2023.3187964	NOUN
cana-4362	369	2	8	8	NUM
cana-4362	369	3	.	.	PUNCT
cana-4362	370	1	huang	huang	PROPN
cana-4362	370	2	,	,	PUNCT
cana-4362	370	3	l.	l.	PROPN
cana-4362	370	4	,	,	PUNCT
cana-4362	370	5	zhao	zhao	PROPN
cana-4362	370	6	,	,	PUNCT
cana-4362	370	7	j.	j.	PROPN
cana-4362	370	8	,	,	PUNCT
cana-4362	370	9	li	li	PROPN
cana-4362	370	10	,	,	PUNCT
cana-4362	370	11	f.	f.	PROPN
cana-4362	370	12	,	,	PUNCT
cana-4362	370	13	&	&	CCONJ
cana-4362	370	14	peng	peng	PROPN
cana-4362	370	15	,	,	PUNCT
cana-4362	370	16	s.	s.	PROPN
cana-4362	370	17	(	(	PUNCT
cana-4362	370	18	2022	2022	NUM
cana-4362	370	19	)	)	PUNCT
cana-4362	370	20	.	.	PUNCT
cana-4362	371	1	hybrid	hybrid	ADJ
cana-4362	371	2	feature	feature	NOUN
cana-4362	371	3	engineering	engineering	NOUN
cana-4362	371	4	and	and	CCONJ
cana-4362	371	5	sentiment	sentiment	NOUN
cana-4362	371	6	analysis	analysis	NOUN
cana-4362	371	7	on	on	ADP
cana-4362	371	8	social	social	ADJ
cana-4362	371	9	media	medium	NOUN
cana-4362	371	10	posts	post	NOUN
cana-4362	371	11	using	use	VERB
cana-4362	371	12	deep	deep	ADJ
cana-4362	371	13	learning	learning	NOUN
cana-4362	371	14	models	model	NOUN
cana-4362	371	15	.	.	PUNCT
cana-4362	372	1	ieee	ieee	NOUN
cana-4362	372	2	transactions	transaction	NOUN
cana-4362	372	3	on	on	ADP
cana-4362	372	4	cognitive	cognitive	ADJ
cana-4362	372	5	and	and	CCONJ
cana-4362	372	6	developmental	developmental	ADJ
cana-4362	372	7	systems	system	NOUN
cana-4362	372	8	,	,	PUNCT
cana-4362	372	9	14(3	14(3	NUM
cana-4362	372	10	)	)	PUNCT
cana-4362	372	11	,	,	PUNCT
cana-4362	372	12	689	689	NUM
cana-4362	372	13	-	-	SYM
cana-4362	372	14	701	701	NUM
cana-4362	372	15	.	.	PUNCT
cana-4362	373	1	https://doi.org/10.1109/tcds.2021.3145894	https://doi.org/10.1109/tcds.2021.3145894	PROPN
cana-4362	373	2	9	9	NUM
cana-4362	373	3	.	.	PUNCT
cana-4362	374	1	liu	liu	PROPN
cana-4362	374	2	,	,	PUNCT
cana-4362	374	3	q.	q.	PROPN
cana-4362	374	4	,	,	PUNCT
cana-4362	374	5	wang	wang	PROPN
cana-4362	374	6	,	,	PUNCT
cana-4362	374	7	p.	p.	PROPN
cana-4362	374	8	,	,	PUNCT
cana-4362	374	9	&	&	CCONJ
cana-4362	374	10	yang	yang	PROPN
cana-4362	374	11	,	,	PUNCT
cana-4362	374	12	y.	y.	PROPN
cana-4362	374	13	(	(	PUNCT
cana-4362	374	14	2023	2023	NUM
cana-4362	374	15	)	)	PUNCT
cana-4362	374	16	.	.	PUNCT
cana-4362	375	1	hybrid	hybrid	ADJ
cana-4362	375	2	feature	feature	NOUN
cana-4362	375	3	selection	selection	NOUN
cana-4362	375	4	approach	approach	NOUN
cana-4362	375	5	based	base	VERB
cana-4362	375	6	on	on	ADP
cana-4362	375	7	sentiment	sentiment	NOUN
cana-4362	375	8	lexicons	lexicon	NOUN
cana-4362	375	9	and	and	CCONJ
cana-4362	375	10	cnn	cnn	PROPN
cana-4362	375	11	for	for	ADP
cana-4362	375	12	twitter	twitter	NOUN
cana-4362	375	13	sentiment	sentiment	NOUN
cana-4362	375	14	classification	classification	NOUN
cana-4362	375	15	.	.	PUNCT
cana-4362	376	1	ieee	ieee	NOUN
cana-4362	376	2	transactions	transaction	NOUN
cana-4362	376	3	on	on	ADP
cana-4362	376	4	information	information	NOUN
cana-4362	376	5	forensics	forensic	NOUN
cana-4362	376	6	and	and	CCONJ
cana-4362	376	7	security	security	NOUN
cana-4362	376	8	,	,	PUNCT
cana-4362	376	9	18(2	18(2	NUM
cana-4362	376	10	)	)	PUNCT
cana-4362	376	11	,	,	PUNCT
cana-4362	376	12	304	304	NUM
cana-4362	376	13	-	-	SYM
cana-4362	376	14	315	315	NUM
cana-4362	376	15	.	.	PUNCT
cana-4362	377	1	https://doi.org/10.1109/tifs.2022.3207615	https://doi.org/10.1109/tifs.2022.3207615	PROPN
cana-4362	377	2	10	10	NUM
cana-4362	377	3	.	.	PUNCT
cana-4362	378	1	gupta	gupta	PROPN
cana-4362	378	2	,	,	PUNCT
cana-4362	378	3	v.	v.	PROPN
cana-4362	378	4	,	,	PUNCT
cana-4362	378	5	singh	singh	PROPN
cana-4362	378	6	,	,	PUNCT
cana-4362	378	7	r.	r.	PROPN
cana-4362	378	8	,	,	PUNCT
cana-4362	378	9	&	&	CCONJ
cana-4362	378	10	bansal	bansal	PROPN
cana-4362	378	11	,	,	PUNCT
cana-4362	378	12	s.	s.	PROPN
cana-4362	378	13	(	(	PUNCT
cana-4362	378	14	2023	2023	NUM
cana-4362	378	15	)	)	PUNCT
cana-4362	378	16	.	.	PUNCT
cana-4362	379	1	a	a	DET
cana-4362	379	2	novel	novel	ADJ
cana-4362	379	3	hybrid	hybrid	NOUN
cana-4362	379	4	approach	approach	NOUN
cana-4362	379	5	combining	combine	VERB
cana-4362	379	6	feature	feature	NOUN
cana-4362	379	7	selection	selection	NOUN
cana-4362	379	8	and	and	CCONJ
cana-4362	379	9	deep	deep	ADJ
cana-4362	379	10	learning	learning	NOUN
cana-4362	379	11	for	for	ADP
cana-4362	379	12	social	social	ADJ
cana-4362	379	13	media	medium	NOUN
cana-4362	379	14	sentiment	sentiment	NOUN
cana-4362	379	15	analysis	analysis	NOUN
cana-4362	379	16	.	.	PUNCT
cana-4362	380	1	ieee	ieee	NOUN
cana-4362	380	2	transactions	transaction	NOUN
cana-4362	380	3	on	on	ADP
cana-4362	380	4	computational	computational	ADJ
cana-4362	380	5	intelligence	intelligence	NOUN
cana-4362	380	6	and	and	CCONJ
cana-4362	380	7	ai	ai	VERB
cana-4362	380	8	in	in	ADP
cana-4362	380	9	games	game	NOUN
cana-4362	380	10	,	,	PUNCT
cana-4362	380	11	15(1	15(1	NUM
cana-4362	380	12	)	)	PUNCT
cana-4362	380	13	,	,	PUNCT
cana-4362	380	14	110	110	NUM
cana-4362	380	15	-	-	SYM
cana-4362	380	16	123	123	NUM
cana-4362	380	17	.	.	PUNCT
cana-4362	381	1	https://doi.org/10.1109/tciaig.2022.3197845	https://doi.org/10.1109/tciaig.2022.3197845	PROPN
cana-4362	381	2	11	11	NUM
cana-4362	381	3	.	.	PUNCT
cana-4362	382	1	zhang	zhang	PROPN
cana-4362	382	2	,	,	PUNCT
cana-4362	382	3	l.	l.	PROPN
cana-4362	382	4	,	,	PUNCT
cana-4362	382	5	wang	wang	PROPN
cana-4362	382	6	,	,	PUNCT
cana-4362	382	7	y.	y.	PROPN
cana-4362	382	8	,	,	PUNCT
cana-4362	382	9	&	&	CCONJ
cana-4362	382	10	chen	chen	PROPN
cana-4362	382	11	,	,	PUNCT
cana-4362	382	12	k.	k.	PROPN
cana-4362	382	13	(	(	PUNCT
cana-4362	382	14	2022	2022	NUM
cana-4362	382	15	)	)	PUNCT
cana-4362	382	16	.	.	PUNCT
cana-4362	383	1	hybrid	hybrid	ADJ
cana-4362	383	2	feature	feature	NOUN
cana-4362	383	3	selection	selection	NOUN
cana-4362	383	4	using	use	VERB
cana-4362	383	5	ga	ga	PROPN
cana-4362	383	6	and	and	CCONJ
cana-4362	383	7	cnn	cnn	PROPN
cana-4362	383	8	for	for	ADP
cana-4362	383	9	sentiment	sentiment	NOUN
cana-4362	383	10	classification	classification	NOUN
cana-4362	383	11	in	in	ADP
cana-4362	383	12	social	social	ADJ
cana-4362	383	13	media	medium	NOUN
cana-4362	383	14	content	content	NOUN
cana-4362	383	15	.	.	PUNCT
cana-4362	384	1	ieee	ieee	NOUN
cana-4362	384	2	transactions	transaction	NOUN
cana-4362	384	3	on	on	ADP
cana-4362	384	4	emerging	emerge	VERB
cana-4362	384	5	topics	topic	NOUN
cana-4362	384	6	in	in	ADP
cana-4362	384	7	computational	computational	ADJ
cana-4362	384	8	intelligence	intelligence	NOUN
cana-4362	384	9	,	,	PUNCT
cana-4362	384	10	6(3	6(3	NUM
cana-4362	384	11	)	)	PUNCT
cana-4362	384	12	,	,	PUNCT
cana-4362	384	13	227	227	NUM
cana-4362	384	14	-	-	SYM
cana-4362	384	15	239	239	NUM
cana-4362	384	16	.	.	PUNCT
cana-4362	385	1	https://doi.org/10.1109/tetci.2022.3156593	https://doi.org/10.1109/tetci.2022.3156593	PROPN
cana-4362	385	2	12	12	NUM
cana-4362	385	3	.	.	PUNCT
cana-4362	386	1	hu	hu	PROPN
cana-4362	386	2	,	,	PUNCT
cana-4362	386	3	y.	y.	PROPN
cana-4362	386	4	,	,	PUNCT
cana-4362	386	5	zhang	zhang	PROPN
cana-4362	386	6	,	,	PUNCT
cana-4362	386	7	x.	x.	PROPN
cana-4362	386	8	,	,	PUNCT
cana-4362	386	9	&	&	CCONJ
cana-4362	386	10	li	li	PROPN
cana-4362	386	11	,	,	PUNCT
cana-4362	386	12	y.	y.	PROPN
cana-4362	386	13	(	(	PUNCT
cana-4362	386	14	2023	2023	NUM
cana-4362	386	15	)	)	PUNCT
cana-4362	386	16	.	.	PUNCT
cana-4362	387	1	sentiment	sentiment	NOUN
cana-4362	387	2	classification	classification	NOUN
cana-4362	387	3	of	of	ADP
cana-4362	387	4	tweets	tweet	NOUN
cana-4362	387	5	using	use	VERB
cana-4362	387	6	hybrid	hybrid	ADJ
cana-4362	387	7	feature	feature	NOUN
cana-4362	387	8	selection	selection	NOUN
cana-4362	387	9	and	and	CCONJ
cana-4362	387	10	cnn	cnn	PROPN
cana-4362	387	11	-	-	PUNCT
cana-4362	387	12	lstm	lstm	PROPN
cana-4362	387	13	model	model	NOUN
cana-4362	387	14	.	.	PUNCT
cana-4362	388	1	ieee	ieee	NOUN
cana-4362	388	2	transactions	transaction	NOUN
cana-4362	388	3	on	on	ADP
cana-4362	388	4	services	service	NOUN
cana-4362	388	5	computing	computing	NOUN
cana-4362	388	6	,	,	PUNCT
cana-4362	388	7	16(2	16(2	NUM
cana-4362	388	8	)	)	PUNCT
cana-4362	388	9	,	,	PUNCT
cana-4362	388	10	420	420	NUM
cana-4362	388	11	-	-	SYM
cana-4362	388	12	431	431	NUM
cana-4362	388	13	.	.	PUNCT
cana-4362	389	1	https://doi.org/10.1109/tsc.2022.3159863	https://doi.org/10.1109/tsc.2022.3159863	PROPN
cana-4362	389	2	13	13	NUM
cana-4362	389	3	.	.	PUNCT
cana-4362	390	1	chen	chen	PROPN
cana-4362	390	2	,	,	PUNCT
cana-4362	390	3	t.	t.	PROPN
cana-4362	390	4	,	,	PUNCT
cana-4362	390	5	wei	wei	PROPN
cana-4362	390	6	,	,	PUNCT
cana-4362	390	7	h.	h.	PROPN
cana-4362	390	8	,	,	PUNCT
cana-4362	390	9	&	&	CCONJ
cana-4362	390	10	shen	shen	PROPN
cana-4362	390	11	,	,	PUNCT
cana-4362	390	12	w.	w.	PROPN
cana-4362	390	13	(	(	PUNCT
cana-4362	390	14	2022	2022	NUM
cana-4362	390	15	)	)	PUNCT
cana-4362	390	16	.	.	PUNCT
cana-4362	391	1	hybrid	hybrid	ADJ
cana-4362	391	2	feature	feature	NOUN
cana-4362	391	3	selection	selection	NOUN
cana-4362	391	4	for	for	ADP
cana-4362	391	5	emotion	emotion	NOUN
cana-4362	391	6	and	and	CCONJ
cana-4362	391	7	sentiment	sentiment	NOUN
cana-4362	391	8	analysis	analysis	NOUN
cana-4362	391	9	on	on	ADP
cana-4362	391	10	multilingual	multilingual	ADJ
cana-4362	391	11	social	social	ADJ
cana-4362	391	12	media	medium	NOUN
cana-4362	391	13	texts	text	NOUN
cana-4362	391	14	.	.	PUNCT
cana-4362	392	1	ieee	ieee	NOUN
cana-4362	392	2	transactions	transaction	NOUN
cana-4362	392	3	on	on	ADP
cana-4362	392	4	knowledge	knowledge	NOUN
cana-4362	392	5	and	and	CCONJ
cana-4362	392	6	data	datum	NOUN
cana-4362	392	7	engineering	engineering	NOUN
cana-4362	392	8	,	,	PUNCT
cana-4362	392	9	34(8	34(8	NUM
cana-4362	392	10	)	)	PUNCT
cana-4362	392	11	,	,	PUNCT
cana-4362	392	12	1997	1997	NUM
cana-4362	392	13	-	-	SYM
cana-4362	392	14	2010	2010	NUM
cana-4362	392	15	.	.	PUNCT
cana-4362	393	1	https://doi.org/10.1109/tkde.2021.3126792	https://doi.org/10.1109/tkde.2021.3126792	NOUN
cana-4362	393	2	communications	communication	NOUN
cana-4362	393	3	on	on	ADP
cana-4362	393	4	applied	apply	VERB
cana-4362	393	5	nonlinear	nonlinear	ADJ
cana-4362	393	6	analysis	analysis	NOUN
cana-4362	393	7	issn	issn	NOUN
cana-4362	393	8	:	:	PUNCT
cana-4362	393	9	1074	1074	NUM
cana-4362	393	10	-	-	PUNCT
cana-4362	393	11	133x	133x	NUM
cana-4362	393	12	vol	vol	NOUN
cana-4362	393	13	32	32	NUM
cana-4362	393	14	no	no	NOUN
cana-4362	393	15	.	.	PUNCT
cana-4362	394	1	9s	9s	NUM
cana-4362	394	2	(	(	PUNCT
cana-4362	394	3	2025	2025	NUM
cana-4362	394	4	)	)	PUNCT
cana-4362	394	5	1917	1917	NUM
cana-4362	394	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-4362	394	7	14	14	NUM
cana-4362	394	8	.	.	PUNCT
cana-4362	395	1	zhou	zhou	PROPN
cana-4362	395	2	,	,	PUNCT
cana-4362	395	3	x.	x.	PROPN
cana-4362	395	4	,	,	PUNCT
cana-4362	395	5	yan	yan	PROPN
cana-4362	395	6	,	,	PUNCT
cana-4362	395	7	l.	l.	PROPN
cana-4362	395	8	,	,	PUNCT
cana-4362	395	9	&	&	CCONJ
cana-4362	395	10	li	li	PROPN
cana-4362	395	11	,	,	PUNCT
cana-4362	395	12	s.	s.	PROPN
cana-4362	395	13	(	(	PUNCT
cana-4362	395	14	2023	2023	NUM
cana-4362	395	15	)	)	PUNCT
cana-4362	395	16	.	.	PUNCT
cana-4362	396	1	a	a	DET
cana-4362	396	2	hybrid	hybrid	ADJ
cana-4362	396	3	deep	deep	ADJ
cana-4362	396	4	learning	learning	NOUN
cana-4362	396	5	and	and	CCONJ
cana-4362	396	6	feature	feature	NOUN
cana-4362	396	7	selection	selection	NOUN
cana-4362	396	8	framework	framework	NOUN
cana-4362	396	9	for	for	ADP
cana-4362	396	10	sentiment	sentiment	NOUN
cana-4362	396	11	classification	classification	NOUN
cana-4362	396	12	.	.	PUNCT
cana-4362	397	1	ieee	ieee	NOUN
cana-4362	397	2	transactions	transaction	NOUN
cana-4362	397	3	on	on	ADP
cana-4362	397	4	artificial	artificial	ADJ
cana-4362	397	5	intelligence	intelligence	NOUN
cana-4362	397	6	,	,	PUNCT
cana-4362	397	7	4(2	4(2	NUM
cana-4362	397	8	)	)	PUNCT
cana-4362	397	9	,	,	PUNCT
cana-4362	397	10	78	78	NUM
cana-4362	397	11	-	-	SYM
cana-4362	397	12	90	90	NUM
cana-4362	397	13	.	.	PUNCT
cana-4362	398	1	https://doi.org/10.1109/tai.2022.3200715	https://doi.org/10.1109/tai.2022.3200715	NOUN
cana-4362	398	2	15	15	NUM
cana-4362	398	3	.	.	PUNCT
cana-4362	399	1	feng	feng	PROPN
cana-4362	399	2	,	,	PUNCT
cana-4362	399	3	r.	r.	PROPN
cana-4362	399	4	,	,	PUNCT
cana-4362	399	5	shi	shi	PROPN
cana-4362	399	6	,	,	PUNCT
cana-4362	399	7	h.	h.	PROPN
cana-4362	399	8	,	,	PUNCT
cana-4362	399	9	&	&	CCONJ
cana-4362	399	10	liu	liu	PROPN
cana-4362	399	11	,	,	PUNCT
cana-4362	399	12	c.	c.	PROPN
cana-4362	399	13	(	(	PUNCT
cana-4362	399	14	2022	2022	NUM
cana-4362	399	15	)	)	PUNCT
cana-4362	399	16	.	.	PUNCT
cana-4362	400	1	hybrid	hybrid	ADJ
cana-4362	400	2	method	method	NOUN
cana-4362	400	3	for	for	ADP
cana-4362	400	4	twitter	twitter	NOUN
cana-4362	400	5	sentiment	sentiment	NOUN
cana-4362	400	6	analysis	analysis	NOUN
cana-4362	400	7	using	use	VERB
cana-4362	400	8	feature	feature	NOUN
cana-4362	400	9	selection	selection	NOUN
cana-4362	400	10	and	and	CCONJ
cana-4362	400	11	deep	deep	ADJ
cana-4362	400	12	learning	learning	NOUN
cana-4362	400	13	.	.	PUNCT
cana-4362	401	1	ieee	ieee	NOUN
cana-4362	401	2	transactions	transaction	NOUN
cana-4362	401	3	on	on	ADP
cana-4362	401	4	signal	signal	ADJ
cana-4362	401	5	processing	processing	NOUN
cana-4362	401	6	,	,	PUNCT
cana-4362	401	7	70(5	70(5	NUM
cana-4362	401	8	)	)	PUNCT
cana-4362	401	9	,	,	PUNCT
cana-4362	401	10	2293	2293	NUM
cana-4362	401	11	-	-	SYM
cana-4362	401	12	2306	2306	NUM
cana-4362	401	13	.	.	PUNCT
cana-4362	402	1	https://doi.org/10.1109/tsp.2022.3195642	https://doi.org/10.1109/tsp.2022.3195642	ADJ
cana-4362	402	2	16	16	NUM
cana-4362	402	3	.	.	PUNCT
cana-4362	403	1	wang	wang	PROPN
cana-4362	403	2	,	,	PUNCT
cana-4362	403	3	x.	x.	PROPN
cana-4362	403	4	,	,	PUNCT
cana-4362	403	5	li	li	PROPN
cana-4362	403	6	,	,	PUNCT
cana-4362	403	7	m.	m.	NOUN
cana-4362	403	8	,	,	PUNCT
cana-4362	403	9	&	&	CCONJ
cana-4362	403	10	hu	hu	PROPN
cana-4362	403	11	,	,	PUNCT
cana-4362	403	12	z.	z.	PROPN
cana-4362	403	13	(	(	PUNCT
cana-4362	403	14	2023	2023	NUM
cana-4362	403	15	)	)	PUNCT
cana-4362	403	16	.	.	PUNCT
cana-4362	404	1	multi	multi	ADJ
cana-4362	404	2	-	-	ADJ
cana-4362	404	3	layer	layer	ADJ
cana-4362	404	4	hybrid	hybrid	ADJ
cana-4362	404	5	feature	feature	NOUN
cana-4362	404	6	selection	selection	NOUN
cana-4362	404	7	for	for	ADP
cana-4362	404	8	social	social	ADJ
cana-4362	404	9	media	medium	NOUN
cana-4362	404	10	sentiment	sentiment	NOUN
cana-4362	404	11	analysis	analysis	NOUN
cana-4362	404	12	.	.	PUNCT
cana-4362	405	1	ieee	ieee	NOUN
cana-4362	405	2	transactions	transaction	NOUN
cana-4362	405	3	on	on	ADP
cana-4362	405	4	network	network	NOUN
cana-4362	405	5	science	science	NOUN
cana-4362	405	6	and	and	CCONJ
cana-4362	405	7	engineering	engineering	NOUN
cana-4362	405	8	,	,	PUNCT
cana-4362	405	9	10(1	10(1	NUM
cana-4362	405	10	)	)	PUNCT
cana-4362	405	11	,	,	PUNCT
cana-4362	405	12	355	355	NUM
cana-4362	405	13	-	-	SYM
cana-4362	405	14	368	368	NUM
cana-4362	405	15	.	.	PUNCT
cana-4362	406	1	https://doi.org/10.1109/tnse.2023.3211964	https://doi.org/10.1109/tnse.2023.3211964	PROPN
cana-4362	406	2	17	17	NUM
cana-4362	406	3	.	.	PUNCT
cana-4362	407	1	zhao	zhao	PROPN
cana-4362	407	2	,	,	PUNCT
cana-4362	407	3	y.	y.	PROPN
cana-4362	407	4	,	,	PUNCT
cana-4362	407	5	liang	liang	PROPN
cana-4362	407	6	,	,	PUNCT
cana-4362	407	7	w.	w.	PROPN
cana-4362	407	8	,	,	PUNCT
cana-4362	407	9	&	&	CCONJ
cana-4362	407	10	zhang	zhang	PROPN
cana-4362	407	11	,	,	PUNCT
cana-4362	407	12	t.	t.	PROPN
cana-4362	407	13	(	(	PUNCT
cana-4362	407	14	2022	2022	NUM
cana-4362	407	15	)	)	PUNCT
cana-4362	407	16	.	.	PUNCT
cana-4362	408	1	a	a	DET
cana-4362	408	2	hybrid	hybrid	ADJ
cana-4362	408	3	approach	approach	NOUN
cana-4362	408	4	combining	combine	VERB
cana-4362	408	5	feature	feature	NOUN
cana-4362	408	6	selection	selection	NOUN
cana-4362	408	7	with	with	ADP
cana-4362	408	8	transfer	transfer	NOUN
cana-4362	408	9	learning	learning	NOUN
cana-4362	408	10	for	for	ADP
cana-4362	408	11	cross	cross	ADJ
cana-4362	408	12	-	-	ADJ
cana-4362	408	13	domain	domain	ADJ
cana-4362	408	14	sentiment	sentiment	NOUN
cana-4362	408	15	classification	classification	NOUN
cana-4362	408	16	.	.	PUNCT
cana-4362	409	1	ieee	ieee	NOUN
cana-4362	409	2	transactions	transaction	NOUN
cana-4362	409	3	on	on	ADP
cana-4362	409	4	artificial	artificial	ADJ
cana-4362	409	5	intelligence	intelligence	NOUN
cana-4362	409	6	,	,	PUNCT
cana-4362	409	7	3(4	3(4	NUM
cana-4362	409	8	)	)	PUNCT
cana-4362	409	9	,	,	PUNCT
cana-4362	409	10	233	233	NUM
cana-4362	409	11	-	-	SYM
cana-4362	409	12	246	246	NUM
cana-4362	409	13	.	.	PUNCT
cana-4362	410	1	https://doi.org/10.1109/tai.2022.3177439	https://doi.org/10.1109/tai.2022.3177439	ADJ
cana-4362	410	2	18	18	NUM
cana-4362	410	3	.	.	PUNCT
cana-4362	411	1	lin	lin	PROPN
cana-4362	411	2	,	,	PUNCT
cana-4362	411	3	f.	f.	PROPN
cana-4362	411	4	,	,	PUNCT
cana-4362	411	5	liu	liu	PROPN
cana-4362	411	6	,	,	PUNCT
cana-4362	411	7	y.	y.	PROPN
cana-4362	411	8	,	,	PUNCT
cana-4362	411	9	&	&	CCONJ
cana-4362	411	10	zhou	zhou	PROPN
cana-4362	411	11	,	,	PUNCT
cana-4362	411	12	q.	q.	PROPN
cana-4362	411	13	(	(	PUNCT
cana-4362	411	14	2023	2023	NUM
cana-4362	411	15	)	)	PUNCT
cana-4362	411	16	.	.	PUNCT
cana-4362	412	1	sentiment	sentiment	NOUN
cana-4362	412	2	classification	classification	NOUN
cana-4362	412	3	on	on	ADP
cana-4362	412	4	short	short	ADJ
cana-4362	412	5	social	social	ADJ
cana-4362	412	6	media	medium	NOUN
cana-4362	412	7	texts	text	NOUN
cana-4362	412	8	using	use	VERB
cana-4362	412	9	hybrid	hybrid	ADJ
cana-4362	412	10	feature	feature	NOUN
cana-4362	412	11	engineering	engineering	NOUN
cana-4362	412	12	.	.	PUNCT
cana-4362	413	1	ieee	ieee	NOUN
cana-4362	413	2	transactions	transaction	NOUN
cana-4362	413	3	on	on	ADP
cana-4362	413	4	cloud	cloud	NOUN
cana-4362	413	5	computing	computing	NOUN
cana-4362	413	6	,	,	PUNCT
cana-4362	413	7	11(2	11(2	NUM
cana-4362	413	8	)	)	PUNCT
cana-4362	413	9	,	,	PUNCT
cana-4362	413	10	584	584	NUM
cana-4362	413	11	-	-	SYM
cana-4362	413	12	597	597	NUM
cana-4362	413	13	.	.	PUNCT
cana-4362	414	1	https://doi.org/10.1109/tcc.2022.3182654	https://doi.org/10.1109/tcc.2022.3182654	ADJ
cana-4362	414	2	19	19	NUM
cana-4362	414	3	.	.	PUNCT
cana-4362	415	1	xie	xie	PROPN
cana-4362	415	2	,	,	PUNCT
cana-4362	415	3	j.	j.	PROPN
cana-4362	415	4	,	,	PUNCT
cana-4362	415	5	sun	sun	PROPN
cana-4362	415	6	,	,	PUNCT
cana-4362	415	7	h.	h.	PROPN
cana-4362	415	8	,	,	PUNCT
cana-4362	415	9	&	&	CCONJ
cana-4362	415	10	cheng	cheng	PROPN
cana-4362	415	11	,	,	PUNCT
cana-4362	415	12	x.	x.	NOUN
cana-4362	415	13	(	(	PUNCT
cana-4362	415	14	2022	2022	NUM
cana-4362	415	15	)	)	PUNCT
cana-4362	415	16	.	.	PUNCT
cana-4362	416	1	a	a	DET
cana-4362	416	2	hybrid	hybrid	ADJ
cana-4362	416	3	feature	feature	NOUN
cana-4362	416	4	selection	selection	NOUN
cana-4362	416	5	and	and	CCONJ
cana-4362	416	6	attention	attention	NOUN
cana-4362	416	7	mechanism	mechanism	NOUN
cana-4362	416	8	for	for	ADP
cana-4362	416	9	improving	improve	VERB
cana-4362	416	10	sentiment	sentiment	NOUN
cana-4362	416	11	classification	classification	NOUN
cana-4362	416	12	on	on	ADP
cana-4362	416	13	twitter	twitter	NOUN
cana-4362	416	14	data	data	PROPN
cana-4362	416	15	.	.	PUNCT
cana-4362	417	1	ieee	ieee	NOUN
cana-4362	417	2	transactions	transaction	NOUN
cana-4362	417	3	on	on	ADP
cana-4362	417	4	artificial	artificial	ADJ
cana-4362	417	5	intelligence	intelligence	NOUN
cana-4362	417	6	,	,	PUNCT
cana-4362	417	7	3(2	3(2	NUM
cana-4362	417	8	)	)	PUNCT
cana-4362	417	9	,	,	PUNCT
cana-4362	417	10	134	134	NUM
cana-4362	417	11	-	-	SYM
cana-4362	417	12	145	145	NUM
cana-4362	417	13	.	.	PUNCT
cana-4362	418	1	https://doi.org/10.1109/tai.2021.3117873	https://doi.org/10.1109/tai.2021.3117873	NOUN
cana-4362	418	2	20	20	NUM
cana-4362	418	3	.	.	PUNCT
cana-4362	419	1	peng	peng	PROPN
cana-4362	419	2	,	,	PUNCT
cana-4362	419	3	c.	c.	PROPN
cana-4362	419	4	,	,	PUNCT
cana-4362	419	5	li	li	PROPN
cana-4362	419	6	,	,	PUNCT
cana-4362	419	7	z.	z.	PROPN
cana-4362	419	8	,	,	PUNCT
cana-4362	419	9	&	&	CCONJ
cana-4362	419	10	xu	xu	PROPN
cana-4362	419	11	,	,	PUNCT
cana-4362	419	12	h.	h.	PROPN
cana-4362	419	13	(	(	PUNCT
cana-4362	419	14	2023	2023	NUM
cana-4362	419	15	)	)	PUNCT
cana-4362	419	16	.	.	PUNCT
cana-4362	420	1	hybrid	hybrid	ADJ
cana-4362	420	2	deep	deep	ADJ
cana-4362	420	3	neural	neural	ADJ
cana-4362	420	4	networks	network	NOUN
cana-4362	420	5	with	with	ADP
cana-4362	420	6	feature	feature	NOUN
cana-4362	420	7	selection	selection	NOUN
cana-4362	420	8	for	for	ADP
cana-4362	420	9	sentiment	sentiment	NOUN
cana-4362	420	10	classification	classification	NOUN
cana-4362	420	11	on	on	ADP
cana-4362	420	12	social	social	ADJ
cana-4362	420	13	media	medium	NOUN
cana-4362	420	14	.	.	PUNCT
cana-4362	421	1	ieee	ieee	NOUN
cana-4362	421	2	transactions	transaction	NOUN
cana-4362	421	3	on	on	ADP
cana-4362	421	4	multimedia	multimedia	NOUN
cana-4362	421	5	,	,	PUNCT
cana-4362	421	6	25(3	25(3	NUM
cana-4362	421	7	)	)	PUNCT
cana-4362	421	8	,	,	PUNCT
cana-4362	421	9	1018	1018	NUM
cana-4362	421	10	-	-	SYM
cana-4362	421	11	1031	1031	NUM
cana-4362	421	12	.	.	PUNCT
cana-4362	422	1	https://doi.org/10.1109/tmm.2023.3184519	https://doi.org/10.1109/tmm.2023.3184519	PROPN
cana-4362	422	2	21	21	NUM
cana-4362	422	3	.	.	PUNCT
cana-4362	423	1	luo	luo	PROPN
cana-4362	423	2	,	,	PUNCT
cana-4362	423	3	j.	j.	PROPN
cana-4362	423	4	,	,	PUNCT
cana-4362	423	5	wang	wang	PROPN
cana-4362	423	6	,	,	PUNCT
cana-4362	423	7	k.	k.	PROPN
cana-4362	423	8	,	,	PUNCT
cana-4362	423	9	&	&	CCONJ
cana-4362	423	10	tang	tang	PROPN
cana-4362	423	11	,	,	PUNCT
cana-4362	423	12	l.	l.	PROPN
cana-4362	423	13	(	(	PUNCT
cana-4362	423	14	2022	2022	NUM
cana-4362	423	15	)	)	PUNCT
cana-4362	423	16	.	.	PUNCT
cana-4362	424	1	hybrid	hybrid	ADJ
cana-4362	424	2	feature	feature	NOUN
cana-4362	424	3	selection	selection	NOUN
cana-4362	424	4	method	method	NOUN
cana-4362	424	5	for	for	ADP
cana-4362	424	6	sentiment	sentiment	NOUN
cana-4362	424	7	analysis	analysis	NOUN
cana-4362	424	8	of	of	ADP
cana-4362	424	9	reviews	review	NOUN
cana-4362	424	10	in	in	ADP
cana-4362	424	11	online	online	ADJ
cana-4362	424	12	social	social	ADJ
cana-4362	424	13	media	medium	NOUN
cana-4362	424	14	platforms	platform	NOUN
cana-4362	424	15	.	.	PUNCT
cana-4362	425	1	ieee	ieee	NOUN
cana-4362	425	2	transactions	transaction	NOUN
cana-4362	425	3	on	on	ADP
cana-4362	425	4	multimedia	multimedia	NOUN
cana-4362	425	5	,	,	PUNCT
cana-4362	425	6	24(6	24(6	NUM
cana-4362	425	7	)	)	PUNCT
cana-4362	425	8	,	,	PUNCT
cana-4362	425	9	2025	2025	NUM
cana-4362	425	10	-	-	SYM
cana-4362	425	11	2038	2038	NUM
cana-4362	425	12	.	.	PUNCT
cana-4362	426	1	https://doi.org/10.1109/tmm.2021.3137859	https://doi.org/10.1109/tmm.2021.3137859	PRON
cana-4362	427	1	22	22	NUM
cana-4362	427	2	.	.	PUNCT
cana-4362	428	1	wu	wu	PROPN
cana-4362	428	2	,	,	PUNCT
cana-4362	428	3	x.	x.	PROPN
cana-4362	428	4	,	,	PUNCT
cana-4362	428	5	qin	qin	PROPN
cana-4362	428	6	,	,	PUNCT
cana-4362	428	7	y.	y.	PROPN
cana-4362	428	8	,	,	PUNCT
cana-4362	428	9	&	&	CCONJ
cana-4362	428	10	zhang	zhang	PROPN
cana-4362	428	11	,	,	PUNCT
cana-4362	428	12	l.	l.	PROPN
cana-4362	428	13	(	(	PUNCT
cana-4362	428	14	2023	2023	NUM
cana-4362	428	15	)	)	PUNCT
cana-4362	428	16	.	.	PUNCT
cana-4362	429	1	feature	feature	NOUN
cana-4362	429	2	selection	selection	NOUN
cana-4362	429	3	for	for	ADP
cana-4362	429	4	social	social	ADJ
cana-4362	429	5	media	medium	NOUN
cana-4362	429	6	sentiment	sentiment	NOUN
cana-4362	429	7	classification	classification	NOUN
cana-4362	429	8	using	use	VERB
cana-4362	429	9	hybrid	hybrid	ADJ
cana-4362	429	10	filters	filter	NOUN
cana-4362	429	11	and	and	CCONJ
cana-4362	429	12	embedded	embed	VERB
cana-4362	429	13	methods	method	NOUN
cana-4362	429	14	.	.	PUNCT
cana-4362	430	1	ieee	ieee	NOUN
cana-4362	430	2	transactions	transaction	NOUN
cana-4362	430	3	on	on	ADP
cana-4362	430	4	pattern	pattern	NOUN
cana-4362	430	5	analysis	analysis	NOUN
cana-4362	430	6	and	and	CCONJ
cana-4362	430	7	machine	machine	NOUN
cana-4362	430	8	intelligence	intelligence	NOUN
cana-4362	430	9	,	,	PUNCT
cana-4362	430	10	45(2	45(2	NOUN
cana-4362	430	11	)	)	PUNCT
cana-4362	430	12	,	,	PUNCT
cana-4362	430	13	892	892	NUM
cana-4362	430	14	-	-	SYM
cana-4362	430	15	906	906	NUM
cana-4362	430	16	.	.	PUNCT
cana-4362	431	1	https://doi.org/10.1109/tpami.2022.3147984	https://doi.org/10.1109/tpami.2022.3147984	PROPN
cana-4362	431	2	23	23	NUM
cana-4362	431	3	.	.	PUNCT
cana-4362	432	1	liu	liu	PROPN
cana-4362	432	2	,	,	PUNCT
cana-4362	432	3	f.	f.	PROPN
cana-4362	432	4	,	,	PUNCT
cana-4362	432	5	zhang	zhang	PROPN
cana-4362	432	6	,	,	PUNCT
cana-4362	432	7	s.	s.	PROPN
cana-4362	432	8	,	,	PUNCT
cana-4362	432	9	&	&	CCONJ
cana-4362	432	10	huang	huang	PROPN
cana-4362	432	11	,	,	PUNCT
cana-4362	432	12	g.	g.	PROPN
cana-4362	432	13	(	(	PUNCT
cana-4362	432	14	2022	2022	NUM
cana-4362	432	15	)	)	PUNCT
cana-4362	432	16	.	.	PUNCT
cana-4362	433	1	feature	feature	NOUN
cana-4362	433	2	selection	selection	NOUN
cana-4362	433	3	for	for	ADP
cana-4362	433	4	twitter	twitter	NOUN
cana-4362	433	5	sentiment	sentiment	NOUN
cana-4362	433	6	analysis	analysis	NOUN
cana-4362	433	7	using	use	VERB
cana-4362	433	8	hybrid	hybrid	ADJ
cana-4362	433	9	search	search	NOUN
cana-4362	433	10	and	and	CCONJ
cana-4362	433	11	gradient	gradient	ADJ
cana-4362	433	12	descent	descent	NOUN
cana-4362	433	13	.	.	PUNCT
cana-4362	434	1	ieee	ieee	NOUN
cana-4362	434	2	transactions	transaction	NOUN
cana-4362	434	3	on	on	ADP
cana-4362	434	4	cybernetics	cybernetic	NOUN
cana-4362	434	5	,	,	PUNCT
cana-4362	434	6	53(7	53(7	NUM
cana-4362	434	7	)	)	PUNCT
cana-4362	434	8	,	,	PUNCT
cana-4362	434	9	3567	3567	NUM
cana-4362	434	10	-	-	SYM
cana-4362	434	11	3580	3580	NUM
cana-4362	434	12	.	.	PUNCT
cana-4362	435	1	https://doi.org/10.1109/tcyb.2021.3145398	https://doi.org/10.1109/tcyb.2021.3145398	PROPN
cana-4362	435	2	24	24	NUM
cana-4362	435	3	.	.	PUNCT
cana-4362	436	1	cheng	cheng	PROPN
cana-4362	436	2	,	,	PUNCT
cana-4362	436	3	y.	y.	PROPN
cana-4362	436	4	,	,	PUNCT
cana-4362	436	5	zhang	zhang	PROPN
cana-4362	436	6	,	,	PUNCT
cana-4362	436	7	q.	q.	PROPN
cana-4362	436	8	,	,	PUNCT
cana-4362	436	9	&	&	CCONJ
cana-4362	436	10	wang	wang	PROPN
cana-4362	436	11	,	,	PUNCT
cana-4362	436	12	h.	h.	PROPN
cana-4362	436	13	(	(	PUNCT
cana-4362	436	14	2023	2023	NUM
cana-4362	436	15	)	)	PUNCT
cana-4362	436	16	.	.	PUNCT
cana-4362	437	1	multi	multi	ADJ
cana-4362	437	2	-	-	ADJ
cana-4362	437	3	view	view	ADJ
cana-4362	437	4	hybrid	hybrid	ADJ
cana-4362	437	5	feature	feature	NOUN
cana-4362	437	6	selection	selection	NOUN
cana-4362	437	7	and	and	CCONJ
cana-4362	437	8	deep	deep	ADJ
cana-4362	437	9	learning	learning	NOUN
cana-4362	437	10	for	for	ADP
cana-4362	437	11	sentiment	sentiment	NOUN
cana-4362	437	12	classification	classification	NOUN
cana-4362	437	13	of	of	ADP
cana-4362	437	14	tweets	tweet	NOUN
cana-4362	437	15	.	.	PUNCT
cana-4362	438	1	ieee	ieee	NOUN
cana-4362	438	2	transactions	transaction	NOUN
cana-4362	438	3	on	on	ADP
cana-4362	438	4	multimedia	multimedia	NOUN
cana-4362	438	5	,	,	PUNCT
cana-4362	438	6	25(2	25(2	NUM
cana-4362	438	7	)	)	PUNCT
cana-4362	438	8	,	,	PUNCT
cana-4362	438	9	775	775	NUM
cana-4362	438	10	-	-	SYM
cana-4362	438	11	787	787	NUM
cana-4362	438	12	.	.	PUNCT
cana-4362	439	1	https://doi.org/10.1109/tmm.2023.3165975	https://doi.org/10.1109/tmm.2023.3165975	NOUN
cana-4362	439	2	25	25	NUM
cana-4362	439	3	.	.	PUNCT
cana-4362	439	4	fang	fang	PROPN
cana-4362	439	5	,	,	PUNCT
cana-4362	439	6	l.	l.	PROPN
cana-4362	439	7	,	,	PUNCT
cana-4362	439	8	wang	wang	PROPN
cana-4362	439	9	,	,	PUNCT
cana-4362	439	10	x.	x.	PROPN
cana-4362	439	11	,	,	PUNCT
cana-4362	439	12	&	&	CCONJ
cana-4362	439	13	xu	xu	PROPN
cana-4362	439	14	,	,	PUNCT
cana-4362	439	15	z.	z.	PROPN
cana-4362	439	16	(	(	PUNCT
cana-4362	439	17	2023	2023	NUM
cana-4362	439	18	)	)	PUNCT
cana-4362	439	19	.	.	PUNCT
cana-4362	440	1	a	a	DET
cana-4362	440	2	hybrid	hybrid	ADJ
cana-4362	440	3	sentiment	sentiment	NOUN
cana-4362	440	4	analysis	analysis	NOUN
cana-4362	440	5	approach	approach	NOUN
cana-4362	440	6	for	for	ADP
cana-4362	440	7	social	social	ADJ
cana-4362	440	8	media	medium	NOUN
cana-4362	440	9	data	datum	NOUN
cana-4362	440	10	using	use	VERB
cana-4362	440	11	feature	feature	NOUN
cana-4362	440	12	selection	selection	NOUN
cana-4362	440	13	and	and	CCONJ
cana-4362	440	14	deep	deep	ADJ
cana-4362	440	15	learning	learning	NOUN
cana-4362	440	16	.	.	PUNCT
cana-4362	441	1	ieee	ieee	NOUN
cana-4362	441	2	transactions	transaction	NOUN
cana-4362	441	3	on	on	ADP
cana-4362	441	4	affective	affective	ADJ
cana-4362	441	5	computing	computing	NOUN
cana-4362	441	6	,	,	PUNCT
cana-4362	441	7	14(3	14(3	NUM
cana-4362	441	8	)	)	PUNCT
cana-4362	441	9	,	,	PUNCT
cana-4362	441	10	1412	1412	NUM
cana-4362	441	11	-	-	SYM
cana-4362	441	12	1423	1423	NUM
cana-4362	441	13	.	.	PUNCT
cana-4362	442	1	https://doi.org/10.1109/taffc.2022.3214312	https://doi.org/10.1109/taffc.2022.3214312	PROPN
cana-4362	442	2	communications	communication	NOUN
cana-4362	442	3	on	on	ADP
cana-4362	442	4	applied	apply	VERB
cana-4362	442	5	nonlinear	nonlinear	ADJ
cana-4362	442	6	analysis	analysis	NOUN
cana-4362	442	7	issn	issn	NOUN
cana-4362	442	8	:	:	PUNCT
cana-4362	442	9	1074	1074	NUM
cana-4362	442	10	-	-	PUNCT
cana-4362	442	11	133x	133x	NUM
cana-4362	442	12	vol	vol	NOUN
cana-4362	442	13	32	32	NUM
cana-4362	442	14	no	no	NOUN
cana-4362	442	15	.	.	PUNCT
cana-4362	443	1	9s	9s	NUM
cana-4362	443	2	(	(	PUNCT
cana-4362	443	3	2025	2025	NUM
cana-4362	443	4	)	)	PUNCT
cana-4362	443	5	1918	1918	NUM
cana-4362	443	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-4362	443	7	26	26	NUM
cana-4362	443	8	.	.	PUNCT
cana-4362	444	1	zhou	zhou	PROPN
cana-4362	444	2	,	,	PUNCT
cana-4362	444	3	p.	p.	PROPN
cana-4362	444	4	,	,	PUNCT
cana-4362	444	5	yang	yang	PROPN
cana-4362	444	6	,	,	PUNCT
cana-4362	444	7	l.	l.	PROPN
cana-4362	444	8	,	,	PUNCT
cana-4362	444	9	&	&	CCONJ
cana-4362	444	10	xu	xu	PROPN
cana-4362	444	11	,	,	PUNCT
cana-4362	444	12	m.	m.	NOUN
cana-4362	444	13	(	(	PUNCT
cana-4362	444	14	2022	2022	NUM
cana-4362	444	15	)	)	PUNCT
cana-4362	444	16	.	.	PUNCT
cana-4362	445	1	hybrid	hybrid	ADJ
cana-4362	445	2	cnn	cnn	PROPN
cana-4362	445	3	and	and	CCONJ
cana-4362	445	4	lstm	lstm	NOUN
cana-4362	445	5	for	for	ADP
cana-4362	445	6	sentiment	sentiment	NOUN
cana-4362	445	7	analysis	analysis	NOUN
cana-4362	445	8	using	use	VERB
cana-4362	445	9	social	social	ADJ
cana-4362	445	10	media	medium	NOUN
cana-4362	445	11	data	datum	NOUN
cana-4362	445	12	with	with	ADP
cana-4362	445	13	feature	feature	NOUN
cana-4362	445	14	selection	selection	NOUN
cana-4362	445	15	.	.	PUNCT
cana-4362	446	1	ieee	ieee	NOUN
cana-4362	446	2	transactions	transaction	NOUN
cana-4362	446	3	on	on	ADP
cana-4362	446	4	multimedia	multimedia	NOUN
cana-4362	446	5	,	,	PUNCT
cana-4362	446	6	24(7	24(7	NUM
cana-4362	446	7	)	)	PUNCT
cana-4362	446	8	,	,	PUNCT
cana-4362	446	9	1553	1553	NUM
cana-4362	446	10	-	-	SYM
cana-4362	446	11	1567	1567	NUM
cana-4362	446	12	.	.	PUNCT
cana-4362	447	1	https://doi.org/10.1109/tmm.2021.3144567	https://doi.org/10.1109/tmm.2021.3144567	ADJ
cana-4362	447	2	27	27	NUM
cana-4362	447	3	.	.	PUNCT
cana-4362	448	1	han	han	PROPN
cana-4362	448	2	,	,	PUNCT
cana-4362	448	3	y.	y.	PROPN
cana-4362	448	4	,	,	PUNCT
cana-4362	448	5	li	li	PROPN
cana-4362	448	6	,	,	PUNCT
cana-4362	448	7	w.	w.	PROPN
cana-4362	448	8	,	,	PUNCT
cana-4362	448	9	&	&	CCONJ
cana-4362	448	10	liu	liu	PROPN
cana-4362	448	11	,	,	PUNCT
cana-4362	448	12	j.	j.	PROPN
cana-4362	448	13	(	(	PUNCT
cana-4362	448	14	2023	2023	NUM
cana-4362	448	15	)	)	PUNCT
cana-4362	448	16	.	.	PUNCT
cana-4362	449	1	hybrid	hybrid	ADJ
cana-4362	449	2	feature	feature	NOUN
cana-4362	449	3	selection	selection	NOUN
cana-4362	449	4	with	with	ADP
cana-4362	449	5	ensemble	ensemble	ADJ
cana-4362	449	6	learning	learning	NOUN
cana-4362	449	7	for	for	ADP
cana-4362	449	8	social	social	ADJ
cana-4362	449	9	media	medium	NOUN
cana-4362	449	10	sentiment	sentiment	NOUN
cana-4362	449	11	classification	classification	NOUN
cana-4362	449	12	.	.	PUNCT
cana-4362	450	1	ieee	ieee	NOUN
cana-4362	450	2	transactions	transaction	NOUN
cana-4362	450	3	on	on	ADP
cana-4362	450	4	neural	neural	ADJ
cana-4362	450	5	networks	network	NOUN
cana-4362	450	6	and	and	CCONJ
cana-4362	450	7	learning	learning	NOUN
cana-4362	450	8	systems	system	NOUN
cana-4362	450	9	,	,	PUNCT
cana-4362	450	10	34(4	34(4	NUM
cana-4362	450	11	)	)	PUNCT
cana-4362	450	12	,	,	PUNCT
cana-4362	450	13	1784	1784	NUM
cana-4362	450	14	-	-	SYM
cana-4362	450	15	1796	1796	NUM
cana-4362	450	16	.	.	PUNCT
cana-4362	451	1	https://doi.org/10.1109/tnnls.2023.3204596	https://doi.org/10.1109/tnnls.2023.3204596	PROPN
cana-4362	451	2	28	28	NUM
cana-4362	451	3	.	.	PUNCT
cana-4362	452	1	liu	liu	PROPN
cana-4362	452	2	,	,	PUNCT
cana-4362	452	3	z.	z.	PROPN
cana-4362	452	4	,	,	PUNCT
cana-4362	452	5	chen	chen	PROPN
cana-4362	452	6	,	,	PUNCT
cana-4362	452	7	y.	y.	PROPN
cana-4362	452	8	,	,	PUNCT
cana-4362	452	9	&	&	CCONJ
cana-4362	452	10	huang	huang	PROPN
cana-4362	452	11	,	,	PUNCT
cana-4362	452	12	y.	y.	PROPN
cana-4362	452	13	(	(	PUNCT
cana-4362	452	14	2022	2022	NUM
cana-4362	452	15	)	)	PUNCT
cana-4362	452	16	.	.	PUNCT
cana-4362	453	1	deep	deep	ADJ
cana-4362	453	2	learning	learning	NOUN
cana-4362	453	3	-	-	PUNCT
cana-4362	453	4	based	base	VERB
cana-4362	453	5	hybrid	hybrid	NOUN
cana-4362	453	6	29	29	NUM
cana-4362	453	7	.	.	PUNCT
cana-4362	454	1	zhang	zhang	PROPN
cana-4362	454	2	,	,	PUNCT
cana-4362	454	3	j.	j.	PROPN
cana-4362	454	4	,	,	PUNCT
cana-4362	454	5	liu	liu	PROPN
cana-4362	454	6	,	,	PUNCT
cana-4362	454	7	b.	b.	PROPN
cana-4362	454	8	,	,	PUNCT
cana-4362	454	9	&	&	CCONJ
cana-4362	454	10	zhang	zhang	PROPN
cana-4362	454	11	,	,	PUNCT
cana-4362	454	12	c.	c.	PROPN
cana-4362	454	13	(	(	PUNCT
cana-4362	454	14	2023	2023	NUM
cana-4362	454	15	)	)	PUNCT
cana-4362	454	16	.	.	PUNCT
cana-4362	455	1	hybrid	hybrid	ADJ
cana-4362	455	2	feature	feature	NOUN
cana-4362	455	3	selection	selection	NOUN
cana-4362	455	4	and	and	CCONJ
cana-4362	455	5	ensemble	ensemble	ADJ
cana-4362	455	6	learning	learning	NOUN
cana-4362	455	7	for	for	ADP
cana-4362	455	8	sentiment	sentiment	NOUN
cana-4362	455	9	classification	classification	NOUN
cana-4362	455	10	in	in	ADP
cana-4362	455	11	social	social	ADJ
cana-4362	455	12	media	medium	NOUN
cana-4362	455	13	data	datum	NOUN
cana-4362	455	14	.	.	PUNCT
cana-4362	456	1	ieee	ieee	NOUN
cana-4362	456	2	transactions	transaction	NOUN
cana-4362	456	3	on	on	ADP
cana-4362	456	4	knowledge	knowledge	NOUN
cana-4362	456	5	and	and	CCONJ
cana-4362	456	6	data	datum	NOUN
cana-4362	456	7	engineering	engineering	NOUN
cana-4362	456	8	,	,	PUNCT
cana-4362	456	9	35(7	35(7	NUM
cana-4362	456	10	)	)	PUNCT
cana-4362	456	11	,	,	PUNCT
cana-4362	456	12	3641	3641	NUM
cana-4362	456	13	-	-	SYM
cana-4362	456	14	3652	3652	NUM
cana-4362	456	15	.	.	PUNCT
cana-4362	457	1	https://doi.org/10.1109/tkde.2023.3189653	https://doi.org/10.1109/tkde.2023.3189653	PROPN
cana-4362	457	2	30	30	NUM
cana-4362	457	3	.	.	PUNCT
cana-4362	457	4	yu	yu	PROPN
cana-4362	457	5	,	,	PUNCT
cana-4362	457	6	p.	p.	PROPN
cana-4362	457	7	,	,	PUNCT
cana-4362	457	8	wang	wang	PROPN
cana-4362	457	9	,	,	PUNCT
cana-4362	457	10	s.	s.	PROPN
cana-4362	457	11	,	,	PUNCT
cana-4362	457	12	&	&	CCONJ
cana-4362	457	13	zheng	zheng	PROPN
cana-4362	457	14	,	,	PUNCT
cana-4362	457	15	x.	x.	NOUN
cana-4362	457	16	(	(	PUNCT
cana-4362	457	17	2023	2023	NUM
cana-4362	457	18	)	)	PUNCT
cana-4362	457	19	.	.	PUNCT
cana-4362	458	1	hybrid	hybrid	ADJ
cana-4362	458	2	feature	feature	NOUN
cana-4362	458	3	extraction	extraction	NOUN
cana-4362	458	4	and	and	CCONJ
cana-4362	458	5	sentiment	sentiment	NOUN
cana-4362	458	6	classification	classification	NOUN
cana-4362	458	7	for	for	ADP
cana-4362	458	8	online	online	ADJ
cana-4362	458	9	social	social	ADJ
cana-4362	458	10	media	medium	NOUN
cana-4362	458	11	posts	post	NOUN
cana-4362	458	12	.	.	PUNCT
cana-4362	459	1	ieee	ieee	NOUN
cana-4362	459	2	transactions	transaction	NOUN
cana-4362	459	3	on	on	ADP
cana-4362	459	4	computational	computational	ADJ
cana-4362	459	5	social	social	ADJ
cana-4362	459	6	systems	system	NOUN
cana-4362	459	7	,	,	PUNCT
cana-4362	459	8	10(2	10(2	NUM
cana-4362	459	9	)	)	PUNCT
cana-4362	459	10	,	,	PUNCT
cana-4362	459	11	500	500	NUM
cana-4362	459	12	-	-	SYM
cana-4362	459	13	513	513	NUM
cana-4362	459	14	.	.	PUNCT
cana-4362	460	1	https://doi.org/10.1109/tcss.2023.3194569	https://doi.org/10.1109/tcss.2023.3194569	PROPN