id	sid	tid	token	lemma	pos
cana-2345	1	1	communications	communication	NOUN
cana-2345	1	2	on	on	ADP
cana-2345	1	3	applied	apply	VERB
cana-2345	1	4	nonlinear	nonlinear	ADJ
cana-2345	1	5	analysis	analysis	NOUN
cana-2345	1	6	issn	issn	NOUN
cana-2345	1	7	:	:	PUNCT
cana-2345	1	8	1074	1074	NUM
cana-2345	1	9	-	-	PUNCT
cana-2345	1	10	133x	133x	NUM
cana-2345	1	11	vol	vol	NOUN
cana-2345	1	12	32	32	NUM
cana-2345	1	13	no	no	NOUN
cana-2345	1	14	.	.	PUNCT
cana-2345	2	1	1s	1s	NUM
cana-2345	2	2	(	(	PUNCT
cana-2345	2	3	2025	2025	NUM
cana-2345	2	4	)	)	PUNCT
cana-2345	2	5	573	573	NUM
cana-2345	2	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2345	2	7	nonlinear	nonlinear	ADJ
cana-2345	2	8	deep	deep	ADJ
cana-2345	2	9	learning	learning	NOUN
cana-2345	2	10	framework	framework	NOUN
cana-2345	2	11	for	for	ADP
cana-2345	2	12	precise	precise	ADJ
cana-2345	2	13	sentiment	sentiment	NOUN
cana-2345	2	14	classification	classification	NOUN
cana-2345	2	15	in	in	ADP
cana-2345	2	16	textual	textual	ADJ
cana-2345	2	17	data	datum	NOUN
cana-2345	2	18	1dr.attili	1dr.attili	NUM
cana-2345	2	19	venkata	venkata	PROPN
cana-2345	2	20	ramana	ramana	PROPN
cana-2345	2	21	,	,	PUNCT
cana-2345	2	22	2dr	2dr	ADJ
cana-2345	2	23	.	.	PUNCT
cana-2345	3	1	kalli	kalli	PROPN
cana-2345	3	2	srinivasa	srinivasa	PROPN
cana-2345	3	3	nageswara	nageswara	PROPN
cana-2345	3	4	prasad3	prasad3	NOUN
cana-2345	3	5	,	,	PUNCT
cana-2345	3	6	annaluri	annaluri	PROPN
cana-2345	3	7	sreenivasa	sreenivasa	PROPN
cana-2345	3	8	rao	rao	PROPN
cana-2345	3	9	1associate	1associate	NUM
cana-2345	3	10	professor	professor	NOUN
cana-2345	3	11	,	,	PUNCT
cana-2345	3	12	department	department	PROPN
cana-2345	3	13	of	of	ADP
cana-2345	3	14	cse	cse	PROPN
cana-2345	3	15	(	(	PUNCT
cana-2345	3	16	aiml	aiml	NOUN
cana-2345	3	17	)	)	PUNCT
cana-2345	3	18	,	,	PUNCT
cana-2345	3	19	geethanjali	geethanjali	PROPN
cana-2345	3	20	college	college	PROPN
cana-2345	3	21	of	of	ADP
cana-2345	3	22	engineering	engineering	NOUN
cana-2345	3	23	and	and	CCONJ
cana-2345	3	24	technology	technology	NOUN
cana-2345	3	25	,	,	PUNCT
cana-2345	3	26	cheeryala	cheeryala	NOUN
cana-2345	3	27	,	,	PUNCT
cana-2345	3	28	kesara	kesara	NOUN
cana-2345	3	29	,	,	PUNCT
cana-2345	3	30	hyderabad-501301	hyderabad-501301	NOUN
cana-2345	3	31	.	.	PUNCT
cana-2345	4	1	email	email	NOUN
cana-2345	4	2	i	i	PROPN
cana-2345	4	3	d	d	PROPN
cana-2345	4	4	:	:	PUNCT
cana-2345	5	1	avrrdg@gmail.com	avrrdg@gmail.com	X
cana-2345	6	1	2department	2department	NUM
cana-2345	6	2	of	of	ADP
cana-2345	6	3	computer	computer	NOUN
cana-2345	6	4	science	science	NOUN
cana-2345	6	5	and	and	CCONJ
cana-2345	6	6	engineering	engineering	NOUN
cana-2345	6	7	,	,	PUNCT
cana-2345	6	8	sasi	sasi	PROPN
cana-2345	6	9	institute	institute	PROPN
cana-2345	6	10	of	of	ADP
cana-2345	6	11	technology	technology	PROPN
cana-2345	6	12	and	and	CCONJ
cana-2345	6	13	engineering	engineering	NOUN
cana-2345	6	14	,	,	PUNCT
cana-2345	6	15	tadepalligudem	tadepalligudem	PROPN
cana-2345	6	16	.	.	PUNCT
cana-2345	7	1	a.p	a.p	PROPN
cana-2345	7	2	.	.	PROPN
cana-2345	7	3	orcid	orcid	PROPN
cana-2345	8	1	i	i	PRON
cana-2345	8	2	d	d	NOUN
cana-2345	8	3	:	:	PUNCT
cana-2345	8	4	0000	0000	NUM
cana-2345	8	5	-	-	PUNCT
cana-2345	8	6	0002	0002	NUM
cana-2345	8	7	-	-	PUNCT
cana-2345	8	8	8083	8083	NUM
cana-2345	8	9	-	-	SYM
cana-2345	8	10	3672	3672	NUM
cana-2345	8	11	.	.	PUNCT
cana-2345	9	1	email	email	NOUN
cana-2345	10	1	i	i	PROPN
cana-2345	10	2	d	d	PROPN
cana-2345	10	3	:	:	PUNCT
cana-2345	11	1	kallisnprasad@gmail.com	kallisnprasad@gmail.com	X
cana-2345	11	2	3department	3department	NUM
cana-2345	11	3	of	of	ADP
cana-2345	11	4	information	information	NOUN
cana-2345	11	5	technology	technology	NOUN
cana-2345	11	6	,	,	PUNCT
cana-2345	11	7	vnr	vnr	PROPN
cana-2345	11	8	vignana	vignana	PROPN
cana-2345	11	9	jyothi	jyothi	PROPN
cana-2345	11	10	institute	institute	PROPN
cana-2345	11	11	of	of	ADP
cana-2345	11	12	engineering	engineering	NOUN
cana-2345	11	13	and	and	CCONJ
cana-2345	11	14	technology	technology	NOUN
cana-2345	11	15	,	,	PUNCT
cana-2345	11	16	hyderabad	hyderabad	PROPN
cana-2345	11	17	,	,	PUNCT
cana-2345	11	18	orchid	orchid	NOUN
cana-2345	11	19	i	i	PROPN
cana-2345	11	20	d	d	PROPN
cana-2345	11	21	0000	0000	NUM
cana-2345	11	22	-	-	PUNCT
cana-2345	11	23	0003	0003	NUM
cana-2345	11	24	-	-	PUNCT
cana-2345	11	25	1618	1618	NUM
cana-2345	11	26	-	-	PUNCT
cana-2345	11	27	672x	672x	PROPN
cana-2345	11	28	.	.	PUNCT
cana-2345	12	1	email	email	NOUN
cana-2345	12	2	:	:	PUNCT
cana-2345	12	3	annaluri.rao@gmail.com	annaluri.rao@gmail.com	X
cana-2345	12	4	article	article	NOUN
cana-2345	12	5	history	history	NOUN
cana-2345	12	6	:	:	PUNCT
cana-2345	12	7	received	receive	VERB
cana-2345	12	8	:	:	PUNCT
cana-2345	12	9	01	01	NUM
cana-2345	12	10	-	-	SYM
cana-2345	12	11	09	09	NUM
cana-2345	12	12	-	-	PUNCT
cana-2345	12	13	2024	2024	NUM
cana-2345	12	14	revised	revise	VERB
cana-2345	12	15	:	:	PUNCT
cana-2345	12	16	16	16	NUM
cana-2345	12	17	-	-	SYM
cana-2345	12	18	10	10	NUM
cana-2345	12	19	-	-	PUNCT
cana-2345	12	20	2024	2024	NUM
cana-2345	12	21	accepted	accept	VERB
cana-2345	12	22	:	:	PUNCT
cana-2345	12	23	01	01	NUM
cana-2345	12	24	-	-	SYM
cana-2345	12	25	11	11	NUM
cana-2345	12	26	-	-	PUNCT
cana-2345	12	27	2024	2024	NUM
cana-2345	12	28	abstract	abstract	NOUN
cana-2345	12	29	:	:	PUNCT
cana-2345	12	30	the	the	DET
cana-2345	12	31	study	study	NOUN
cana-2345	12	32	presents	present	VERB
cana-2345	12	33	the	the	DET
cana-2345	12	34	nonlinear	nonlinear	ADJ
cana-2345	12	35	deep	deep	ADJ
cana-2345	12	36	learning	learning	NOUN
cana-2345	12	37	framework	framework	NOUN
cana-2345	12	38	for	for	ADP
cana-2345	12	39	precise	precise	ADJ
cana-2345	12	40	sentiment	sentiment	NOUN
cana-2345	12	41	classification	classification	NOUN
cana-2345	12	42	(	(	PUNCT
cana-2345	12	43	npsc	npsc	PROPN
cana-2345	12	44	)	)	PUNCT
cana-2345	12	45	,	,	PUNCT
cana-2345	12	46	focusing	focus	VERB
cana-2345	12	47	on	on	ADP
cana-2345	12	48	improving	improve	VERB
cana-2345	12	49	sentiment	sentiment	NOUN
cana-2345	12	50	analysis	analysis	NOUN
cana-2345	12	51	in	in	ADP
cana-2345	12	52	text	text	NOUN
cana-2345	12	53	data	datum	NOUN
cana-2345	12	54	.	.	PUNCT
cana-2345	13	1	the	the	DET
cana-2345	13	2	model	model	NOUN
cana-2345	13	3	addresses	address	VERB
cana-2345	13	4	the	the	DET
cana-2345	13	5	challenge	challenge	NOUN
cana-2345	13	6	of	of	ADP
cana-2345	13	7	capturing	capture	VERB
cana-2345	13	8	complex	complex	ADJ
cana-2345	13	9	and	and	CCONJ
cana-2345	13	10	hidden	hidden	ADJ
cana-2345	13	11	sentiment	sentiment	NOUN
cana-2345	13	12	patterns	pattern	NOUN
cana-2345	13	13	using	use	VERB
cana-2345	13	14	a	a	DET
cana-2345	13	15	combination	combination	NOUN
cana-2345	13	16	of	of	ADP
cana-2345	13	17	bert	bert	PROPN
cana-2345	13	18	embeddings	embedding	NOUN
cana-2345	13	19	,	,	PUNCT
cana-2345	13	20	bidirectional	bidirectional	ADJ
cana-2345	13	21	lstm	lstm	NOUN
cana-2345	13	22	(	(	PUNCT
cana-2345	13	23	bilstm	bilstm	NOUN
cana-2345	13	24	)	)	PUNCT
cana-2345	13	25	,	,	PUNCT
cana-2345	13	26	convolutional	convolutional	ADJ
cana-2345	13	27	neural	neural	ADJ
cana-2345	13	28	networks	network	NOUN
cana-2345	13	29	(	(	PUNCT
cana-2345	13	30	cnn	cnn	PROPN
cana-2345	13	31	)	)	PUNCT
cana-2345	13	32	,	,	PUNCT
cana-2345	13	33	and	and	CCONJ
cana-2345	13	34	an	an	DET
cana-2345	13	35	attention	attention	NOUN
cana-2345	13	36	mechanism	mechanism	NOUN
cana-2345	13	37	.	.	PUNCT
cana-2345	14	1	this	this	DET
cana-2345	14	2	setup	setup	NOUN
cana-2345	14	3	helps	help	VERB
cana-2345	14	4	in	in	ADP
cana-2345	14	5	identifying	identify	VERB
cana-2345	14	6	subtle	subtle	ADJ
cana-2345	14	7	relationships	relationship	NOUN
cana-2345	14	8	within	within	ADP
cana-2345	14	9	text	text	NOUN
cana-2345	14	10	,	,	PUNCT
cana-2345	14	11	enabling	enable	VERB
cana-2345	14	12	more	more	ADV
cana-2345	14	13	precise	precise	ADJ
cana-2345	14	14	sentiment	sentiment	NOUN
cana-2345	14	15	predictions	prediction	NOUN
cana-2345	14	16	.	.	PUNCT
cana-2345	15	1	npsc	npsc	PROPN
cana-2345	15	2	tackles	tackle	VERB
cana-2345	15	3	common	common	ADJ
cana-2345	15	4	issues	issue	NOUN
cana-2345	15	5	faced	face	VERB
cana-2345	15	6	by	by	ADP
cana-2345	15	7	existing	exist	VERB
cana-2345	15	8	models	model	NOUN
cana-2345	15	9	that	that	PRON
cana-2345	15	10	struggle	struggle	VERB
cana-2345	15	11	with	with	ADP
cana-2345	15	12	detailed	detailed	ADJ
cana-2345	15	13	text	text	NOUN
cana-2345	15	14	analysis	analysis	NOUN
cana-2345	15	15	.	.	PUNCT
cana-2345	16	1	using	use	VERB
cana-2345	16	2	the	the	DET
cana-2345	16	3	imdb	imdb	NOUN
cana-2345	16	4	movie	movie	NOUN
cana-2345	16	5	review	review	NOUN
cana-2345	16	6	dataset	dataset	NOUN
cana-2345	16	7	,	,	PUNCT
cana-2345	16	8	the	the	DET
cana-2345	16	9	approach	approach	NOUN
cana-2345	16	10	relies	rely	VERB
cana-2345	16	11	on	on	ADP
cana-2345	16	12	bert	bert	PROPN
cana-2345	16	13	to	to	PART
cana-2345	16	14	convert	convert	VERB
cana-2345	16	15	words	word	NOUN
cana-2345	16	16	into	into	ADP
cana-2345	16	17	dense	dense	ADJ
cana-2345	16	18	vectors	vector	NOUN
cana-2345	16	19	,	,	PUNCT
cana-2345	16	20	bilstm	bilstm	NOUN
cana-2345	16	21	for	for	ADP
cana-2345	16	22	capturing	capture	VERB
cana-2345	16	23	context	context	NOUN
cana-2345	16	24	from	from	ADP
cana-2345	16	25	both	both	DET
cana-2345	16	26	directions	direction	NOUN
cana-2345	16	27	,	,	PUNCT
cana-2345	16	28	and	and	CCONJ
cana-2345	16	29	cnn	cnn	PROPN
cana-2345	16	30	to	to	PART
cana-2345	16	31	extract	extract	VERB
cana-2345	16	32	key	key	ADJ
cana-2345	16	33	local	local	ADJ
cana-2345	16	34	features	feature	NOUN
cana-2345	16	35	.	.	PUNCT
cana-2345	17	1	an	an	DET
cana-2345	17	2	attention	attention	NOUN
cana-2345	17	3	layer	layer	NOUN
cana-2345	17	4	highlights	highlight	VERB
cana-2345	17	5	important	important	ADJ
cana-2345	17	6	words	word	NOUN
cana-2345	17	7	,	,	PUNCT
cana-2345	17	8	refining	refine	VERB
cana-2345	17	9	the	the	DET
cana-2345	17	10	sentiment	sentiment	NOUN
cana-2345	17	11	detection	detection	NOUN
cana-2345	17	12	process	process	NOUN
cana-2345	17	13	.	.	PUNCT
cana-2345	18	1	the	the	DET
cana-2345	18	2	integration	integration	NOUN
cana-2345	18	3	of	of	ADP
cana-2345	18	4	these	these	DET
cana-2345	18	5	components	component	NOUN
cana-2345	18	6	in	in	ADP
cana-2345	18	7	a	a	DET
cana-2345	18	8	nonlinear	nonlinear	ADJ
cana-2345	18	9	fusion	fusion	NOUN
cana-2345	18	10	layer	layer	NOUN
cana-2345	18	11	allows	allow	VERB
cana-2345	18	12	for	for	ADP
cana-2345	18	13	enhanced	enhanced	ADJ
cana-2345	18	14	feature	feature	NOUN
cana-2345	18	15	interaction	interaction	NOUN
cana-2345	18	16	,	,	PUNCT
cana-2345	18	17	leading	lead	VERB
cana-2345	18	18	to	to	ADP
cana-2345	18	19	accurate	accurate	ADJ
cana-2345	18	20	classification	classification	NOUN
cana-2345	18	21	through	through	ADP
cana-2345	18	22	a	a	DET
cana-2345	18	23	final	final	ADJ
cana-2345	18	24	softmax	softmax	NOUN
cana-2345	18	25	layer	layer	NOUN
cana-2345	18	26	.	.	PUNCT
cana-2345	19	1	the	the	DET
cana-2345	19	2	findings	finding	NOUN
cana-2345	19	3	show	show	VERB
cana-2345	19	4	that	that	SCONJ
cana-2345	19	5	npsc	npsc	NOUN
cana-2345	19	6	performs	perform	VERB
cana-2345	19	7	better	well	ADV
cana-2345	19	8	across	across	ADP
cana-2345	19	9	metrics	metric	NOUN
cana-2345	19	10	like	like	ADP
cana-2345	19	11	accuracy	accuracy	NOUN
cana-2345	19	12	,	,	PUNCT
cana-2345	19	13	precision	precision	NOUN
cana-2345	19	14	,	,	PUNCT
cana-2345	19	15	recall	recall	NOUN
cana-2345	19	16	,	,	PUNCT
cana-2345	19	17	and	and	CCONJ
cana-2345	19	18	f1	f1	NOUN
cana-2345	19	19	-	-	PUNCT
cana-2345	19	20	score	score	NOUN
cana-2345	19	21	.	.	PUNCT
cana-2345	20	1	ablation	ablation	NOUN
cana-2345	20	2	studies	study	NOUN
cana-2345	20	3	reveal	reveal	VERB
cana-2345	20	4	the	the	DET
cana-2345	20	5	impact	impact	NOUN
cana-2345	20	6	of	of	ADP
cana-2345	20	7	each	each	DET
cana-2345	20	8	component	component	NOUN
cana-2345	20	9	,	,	PUNCT
cana-2345	20	10	demonstrating	demonstrate	VERB
cana-2345	20	11	their	their	PRON
cana-2345	20	12	role	role	NOUN
cana-2345	20	13	in	in	ADP
cana-2345	20	14	enhancing	enhance	VERB
cana-2345	20	15	the	the	DET
cana-2345	20	16	model	model	NOUN
cana-2345	20	17	’s	’s	PART
cana-2345	20	18	effectiveness	effectiveness	NOUN
cana-2345	20	19	.	.	PUNCT
cana-2345	21	1	error	error	NOUN
cana-2345	21	2	analysis	analysis	NOUN
cana-2345	21	3	shows	show	VERB
cana-2345	21	4	how	how	SCONJ
cana-2345	21	5	npsc	npsc	ADJ
cana-2345	21	6	manages	manage	VERB
cana-2345	21	7	mixed	mixed	ADJ
cana-2345	21	8	sentiments	sentiment	NOUN
cana-2345	21	9	and	and	CCONJ
cana-2345	21	10	challenging	challenge	VERB
cana-2345	21	11	text	text	NOUN
cana-2345	21	12	patterns	pattern	NOUN
cana-2345	21	13	.	.	PUNCT
cana-2345	22	1	the	the	DET
cana-2345	22	2	study	study	NOUN
cana-2345	22	3	highlights	highlight	NOUN
cana-2345	22	4	npsc	npsc	PROPN
cana-2345	22	5	’s	’s	PART
cana-2345	22	6	potential	potential	NOUN
cana-2345	22	7	as	as	ADP
cana-2345	22	8	a	a	DET
cana-2345	22	9	reliable	reliable	ADJ
cana-2345	22	10	method	method	NOUN
cana-2345	22	11	for	for	ADP
cana-2345	22	12	sentiment	sentiment	NOUN
cana-2345	22	13	analysis	analysis	NOUN
cana-2345	22	14	,	,	PUNCT
cana-2345	22	15	contributing	contribute	VERB
cana-2345	22	16	to	to	ADP
cana-2345	22	17	better	well	ADJ
cana-2345	22	18	handling	handling	NOUN
cana-2345	22	19	of	of	ADP
cana-2345	22	20	complex	complex	ADJ
cana-2345	22	21	textual	textual	ADJ
cana-2345	22	22	data	datum	NOUN
cana-2345	22	23	.	.	PUNCT
cana-2345	23	1	keywords	keyword	NOUN
cana-2345	23	2	:	:	PUNCT
cana-2345	23	3	nonlinear	nonlinear	ADJ
cana-2345	23	4	framework	framework	NOUN
cana-2345	23	5	,	,	PUNCT
cana-2345	23	6	sentiment	sentiment	NOUN
cana-2345	23	7	classification	classification	NOUN
cana-2345	23	8	,	,	PUNCT
cana-2345	23	9	text	text	NOUN
cana-2345	23	10	analysis	analysis	NOUN
cana-2345	23	11	,	,	PUNCT
cana-2345	23	12	bert	bert	PROPN
cana-2345	23	13	,	,	PUNCT
cana-2345	23	14	bilstm	bilstm	NOUN
cana-2345	23	15	,	,	PUNCT
cana-2345	23	16	cnn	cnn	PROPN
cana-2345	23	17	,	,	PUNCT
cana-2345	23	18	attention	attention	NOUN
cana-2345	23	19	mechanism	mechanism	NOUN
cana-2345	23	20	,	,	PUNCT
cana-2345	23	21	npsc	npsc	ADJ
cana-2345	23	22	introduction	introduction	NOUN
cana-2345	23	23	sentiment	sentiment	NOUN
cana-2345	23	24	analysis	analysis	NOUN
cana-2345	23	25	of	of	ADP
cana-2345	23	26	text	text	NOUN
cana-2345	23	27	is	be	AUX
cana-2345	23	28	important	important	ADJ
cana-2345	23	29	in	in	ADP
cana-2345	23	30	understanding	understand	VERB
cana-2345	23	31	opinions	opinion	NOUN
cana-2345	23	32	and	and	CCONJ
cana-2345	23	33	emotions	emotion	NOUN
cana-2345	23	34	in	in	ADP
cana-2345	23	35	various	various	ADJ
cana-2345	23	36	fields	field	NOUN
cana-2345	23	37	like	like	ADP
cana-2345	23	38	social	social	ADJ
cana-2345	23	39	media	medium	NOUN
cana-2345	23	40	,	,	PUNCT
cana-2345	23	41	customer	customer	NOUN
cana-2345	23	42	feedback	feedback	NOUN
cana-2345	23	43	,	,	PUNCT
cana-2345	23	44	and	and	CCONJ
cana-2345	23	45	market	market	NOUN
cana-2345	23	46	research	research	NOUN
cana-2345	23	47	.	.	PUNCT
cana-2345	24	1	traditional	traditional	ADJ
cana-2345	24	2	methods	method	NOUN
cana-2345	24	3	using	use	VERB
cana-2345	24	4	rule	rule	NOUN
cana-2345	24	5	-	-	PUNCT
cana-2345	24	6	based	base	VERB
cana-2345	24	7	or	or	CCONJ
cana-2345	24	8	machine	machine	NOUN
cana-2345	24	9	learning	learning	NOUN
cana-2345	24	10	models	model	NOUN
cana-2345	24	11	often	often	ADV
cana-2345	24	12	fail	fail	VERB
cana-2345	24	13	to	to	PART
cana-2345	24	14	capture	capture	VERB
cana-2345	24	15	the	the	DET
cana-2345	24	16	deeper	deep	ADJ
cana-2345	24	17	meaning	meaning	NOUN
cana-2345	24	18	of	of	ADP
cana-2345	24	19	language	language	NOUN
cana-2345	24	20	,	,	PUNCT
cana-2345	24	21	such	such	ADJ
cana-2345	24	22	as	as	ADP
cana-2345	24	23	sarcasm	sarcasm	NOUN
cana-2345	24	24	,	,	PUNCT
cana-2345	24	25	context	context	NOUN
cana-2345	24	26	changes	change	NOUN
cana-2345	24	27	,	,	PUNCT
cana-2345	24	28	or	or	CCONJ
cana-2345	24	29	hidden	hidden	ADJ
cana-2345	24	30	sentiments	sentiment	NOUN
cana-2345	24	31	.	.	PUNCT
cana-2345	25	1	this	this	PRON
cana-2345	25	2	makes	make	VERB
cana-2345	25	3	the	the	DET
cana-2345	25	4	sentiment	sentiment	NOUN
cana-2345	25	5	analysis	analysis	NOUN
cana-2345	25	6	task	task	NOUN
cana-2345	25	7	complex	complex	NOUN
cana-2345	25	8	and	and	CCONJ
cana-2345	25	9	less	less	ADV
cana-2345	25	10	accurate	accurate	ADJ
cana-2345	25	11	.	.	PUNCT
cana-2345	26	1	advanced	advanced	ADJ
cana-2345	26	2	deep	deep	ADJ
cana-2345	26	3	learning	learning	NOUN
cana-2345	26	4	models	model	NOUN
cana-2345	26	5	,	,	PUNCT
cana-2345	26	6	such	such	ADJ
cana-2345	26	7	as	as	ADP
cana-2345	26	8	lstm	lstm	PROPN
cana-2345	26	9	,	,	PUNCT
cana-2345	26	10	gru	gru	PROPN
cana-2345	26	11	,	,	PUNCT
cana-2345	26	12	and	and	CCONJ
cana-2345	26	13	bert	bert	PROPN
cana-2345	26	14	,	,	PUNCT
cana-2345	26	15	have	have	AUX
cana-2345	26	16	been	be	AUX
cana-2345	26	17	used	use	VERB
cana-2345	26	18	to	to	PART
cana-2345	26	19	address	address	VERB
cana-2345	26	20	these	these	DET
cana-2345	26	21	issues	issue	NOUN
cana-2345	26	22	,	,	PUNCT
cana-2345	26	23	showing	show	VERB
cana-2345	26	24	better	well	ADJ
cana-2345	26	25	performance	performance	NOUN
cana-2345	26	26	in	in	ADP
cana-2345	26	27	analyzing	analyze	VERB
cana-2345	26	28	and	and	CCONJ
cana-2345	26	29	classifying	classify	VERB
cana-2345	26	30	sentiments	sentiment	NOUN
cana-2345	26	31	in	in	ADP
cana-2345	26	32	text	text	NOUN
cana-2345	27	1	[	[	X
cana-2345	27	2	1	1	NUM
cana-2345	27	3	]	]	PUNCT
cana-2345	27	4	.	.	PUNCT
cana-2345	28	1	however	however	ADV
cana-2345	28	2	,	,	PUNCT
cana-2345	28	3	these	these	DET
cana-2345	28	4	models	model	NOUN
cana-2345	28	5	still	still	ADV
cana-2345	28	6	face	face	VERB
cana-2345	28	7	challenges	challenge	NOUN
cana-2345	28	8	,	,	PUNCT
cana-2345	28	9	especially	especially	ADV
cana-2345	28	10	when	when	SCONJ
cana-2345	28	11	dealing	deal	VERB
cana-2345	28	12	with	with	ADP
cana-2345	28	13	domain	domain	NOUN
cana-2345	28	14	-	-	PUNCT
cana-2345	28	15	specific	specific	ADJ
cana-2345	28	16	language	language	NOUN
cana-2345	28	17	,	,	PUNCT
cana-2345	28	18	diverse	diverse	ADJ
cana-2345	28	19	data	datum	NOUN
cana-2345	28	20	,	,	PUNCT
cana-2345	28	21	and	and	CCONJ
cana-2345	28	22	multi	multi	ADJ
cana-2345	28	23	-	-	ADJ
cana-2345	28	24	emotion	emotion	NOUN
cana-2345	28	25	expressions	expression	NOUN
cana-2345	28	26	in	in	ADP
cana-2345	28	27	a	a	DET
cana-2345	28	28	single	single	ADJ
cana-2345	28	29	text	text	NOUN
cana-2345	28	30	[	[	X
cana-2345	28	31	2	2	NUM
cana-2345	28	32	]	]	PUNCT
cana-2345	28	33	,	,	PUNCT
cana-2345	28	34	[	[	X
cana-2345	28	35	3	3	NUM
cana-2345	28	36	]	]	PUNCT
cana-2345	28	37	.	.	PUNCT
cana-2345	29	1	mailto:avrrdg@gmail.com	mailto:avrrdg@gmail.com	X
cana-2345	29	2	mailto:kallisnprasad@gmail.com	mailto:kallisnprasad@gmail.com	PROPN
cana-2345	29	3	mailto:annaluri.rao@gmail.com	mailto:annaluri.rao@gmail.com	X
cana-2345	29	4	communications	communication	NOUN
cana-2345	29	5	on	on	ADP
cana-2345	29	6	applied	apply	VERB
cana-2345	29	7	nonlinear	nonlinear	ADJ
cana-2345	29	8	analysis	analysis	NOUN
cana-2345	29	9	issn	issn	NOUN
cana-2345	29	10	:	:	PUNCT
cana-2345	29	11	1074	1074	NUM
cana-2345	29	12	-	-	PUNCT
cana-2345	29	13	133x	133x	NUM
cana-2345	29	14	vol	vol	NOUN
cana-2345	29	15	32	32	NUM
cana-2345	29	16	no	no	NOUN
cana-2345	29	17	.	.	PUNCT
cana-2345	30	1	1s	1s	NUM
cana-2345	30	2	(	(	PUNCT
cana-2345	30	3	2025	2025	NUM
cana-2345	30	4	)	)	PUNCT
cana-2345	30	5	574	574	NUM
cana-2345	30	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2345	30	7	many	many	ADJ
cana-2345	30	8	studies	study	NOUN
cana-2345	30	9	focus	focus	VERB
cana-2345	30	10	on	on	ADP
cana-2345	30	11	neural	neural	ADJ
cana-2345	30	12	networks	network	NOUN
cana-2345	30	13	,	,	PUNCT
cana-2345	30	14	but	but	CCONJ
cana-2345	30	15	they	they	PRON
cana-2345	30	16	often	often	ADV
cana-2345	30	17	overlook	overlook	VERB
cana-2345	30	18	how	how	SCONJ
cana-2345	30	19	integrating	integrate	VERB
cana-2345	30	20	these	these	DET
cana-2345	30	21	models	model	NOUN
cana-2345	30	22	with	with	ADP
cana-2345	30	23	traditional	traditional	ADJ
cana-2345	30	24	lexical	lexical	ADJ
cana-2345	30	25	approaches	approach	NOUN
cana-2345	30	26	can	can	AUX
cana-2345	30	27	enhance	enhance	VERB
cana-2345	30	28	sentiment	sentiment	NOUN
cana-2345	30	29	analysis	analysis	NOUN
cana-2345	30	30	.	.	PUNCT
cana-2345	31	1	combining	combine	VERB
cana-2345	31	2	neural	neural	ADJ
cana-2345	31	3	and	and	CCONJ
cana-2345	31	4	lexical	lexical	ADJ
cana-2345	31	5	methods	method	NOUN
cana-2345	31	6	can	can	AUX
cana-2345	31	7	capture	capture	VERB
cana-2345	31	8	subtle	subtle	ADJ
cana-2345	31	9	expressions	expression	NOUN
cana-2345	31	10	and	and	CCONJ
cana-2345	31	11	improve	improve	VERB
cana-2345	31	12	context	context	NOUN
cana-2345	31	13	understanding	understanding	NOUN
cana-2345	31	14	,	,	PUNCT
cana-2345	31	15	yet	yet	CCONJ
cana-2345	31	16	this	this	DET
cana-2345	31	17	combination	combination	NOUN
cana-2345	31	18	is	be	AUX
cana-2345	31	19	rarely	rarely	ADV
cana-2345	31	20	explored	explore	VERB
cana-2345	31	21	deeply	deeply	ADV
cana-2345	31	22	[	[	X
cana-2345	31	23	4	4	NUM
cana-2345	31	24	]	]	X
cana-2345	31	25	.	.	PUNCT
cana-2345	32	1	dependency	dependency	NOUN
cana-2345	32	2	-	-	PUNCT
cana-2345	32	3	based	base	VERB
cana-2345	32	4	techniques	technique	NOUN
cana-2345	32	5	that	that	PRON
cana-2345	32	6	extract	extract	VERB
cana-2345	32	7	features	feature	NOUN
cana-2345	32	8	from	from	ADP
cana-2345	32	9	grammatical	grammatical	ADJ
cana-2345	32	10	structures	structure	NOUN
cana-2345	32	11	offer	offer	VERB
cana-2345	32	12	additional	additional	ADJ
cana-2345	32	13	insights	insight	NOUN
cana-2345	32	14	but	but	CCONJ
cana-2345	32	15	need	need	VERB
cana-2345	32	16	further	further	ADJ
cana-2345	32	17	refinement	refinement	NOUN
cana-2345	32	18	for	for	ADP
cana-2345	32	19	better	well	ADJ
cana-2345	32	20	accuracy	accuracy	NOUN
cana-2345	32	21	[	[	X
cana-2345	32	22	5	5	NUM
cana-2345	32	23	]	]	PUNCT
cana-2345	32	24	.	.	PUNCT
cana-2345	33	1	the	the	DET
cana-2345	33	2	lack	lack	NOUN
cana-2345	33	3	of	of	ADP
cana-2345	33	4	such	such	ADJ
cana-2345	33	5	integrated	integrated	ADJ
cana-2345	33	6	,	,	PUNCT
cana-2345	33	7	comprehensive	comprehensive	ADJ
cana-2345	33	8	frameworks	framework	NOUN
cana-2345	33	9	limits	limit	VERB
cana-2345	33	10	the	the	DET
cana-2345	33	11	effectiveness	effectiveness	NOUN
cana-2345	33	12	of	of	ADP
cana-2345	33	13	current	current	ADJ
cana-2345	33	14	sentiment	sentiment	NOUN
cana-2345	33	15	analysis	analysis	NOUN
cana-2345	33	16	models	model	NOUN
cana-2345	33	17	,	,	PUNCT
cana-2345	33	18	especially	especially	ADV
cana-2345	33	19	when	when	SCONJ
cana-2345	33	20	applied	apply	VERB
cana-2345	33	21	to	to	ADP
cana-2345	33	22	complex	complex	ADJ
cana-2345	33	23	real	real	ADJ
cana-2345	33	24	-	-	PUNCT
cana-2345	33	25	world	world	NOUN
cana-2345	33	26	data	datum	NOUN
cana-2345	33	27	.	.	PUNCT
cana-2345	34	1	current	current	ADJ
cana-2345	34	2	neural	neural	ADJ
cana-2345	34	3	network	network	NOUN
cana-2345	34	4	models	model	NOUN
cana-2345	34	5	excel	excel	VERB
cana-2345	34	6	in	in	ADP
cana-2345	34	7	processing	process	VERB
cana-2345	34	8	sequential	sequential	ADJ
cana-2345	34	9	data	datum	NOUN
cana-2345	34	10	,	,	PUNCT
cana-2345	34	11	but	but	CCONJ
cana-2345	34	12	they	they	PRON
cana-2345	34	13	struggle	struggle	VERB
cana-2345	34	14	with	with	ADP
cana-2345	34	15	capturing	capture	VERB
cana-2345	34	16	deeper	deep	ADJ
cana-2345	34	17	contextual	contextual	ADJ
cana-2345	34	18	relationships	relationship	NOUN
cana-2345	34	19	,	,	PUNCT
cana-2345	34	20	especially	especially	ADV
cana-2345	34	21	when	when	SCONJ
cana-2345	34	22	sentiment	sentiment	NOUN
cana-2345	34	23	is	be	AUX
cana-2345	34	24	subtle	subtle	ADJ
cana-2345	34	25	or	or	CCONJ
cana-2345	34	26	hidden	hide	VERB
cana-2345	34	27	in	in	ADP
cana-2345	34	28	complex	complex	ADJ
cana-2345	34	29	sentence	sentence	NOUN
cana-2345	34	30	structures	structure	NOUN
cana-2345	34	31	.	.	PUNCT
cana-2345	35	1	while	while	SCONJ
cana-2345	35	2	there	there	PRON
cana-2345	35	3	are	be	VERB
cana-2345	35	4	efforts	effort	NOUN
cana-2345	35	5	to	to	PART
cana-2345	35	6	integrate	integrate	VERB
cana-2345	35	7	neural	neural	ADJ
cana-2345	35	8	models	model	NOUN
cana-2345	35	9	with	with	ADP
cana-2345	35	10	traditional	traditional	ADJ
cana-2345	35	11	methods	method	NOUN
cana-2345	35	12	,	,	PUNCT
cana-2345	35	13	they	they	PRON
cana-2345	35	14	often	often	ADV
cana-2345	35	15	lack	lack	VERB
cana-2345	35	16	the	the	DET
cana-2345	35	17	depth	depth	NOUN
cana-2345	35	18	and	and	CCONJ
cana-2345	35	19	flexibility	flexibility	NOUN
cana-2345	35	20	needed	need	VERB
cana-2345	35	21	to	to	PART
cana-2345	35	22	adapt	adapt	VERB
cana-2345	35	23	to	to	ADP
cana-2345	35	24	various	various	ADJ
cana-2345	35	25	domains	domain	NOUN
cana-2345	35	26	.	.	PUNCT
cana-2345	36	1	emotion	emotion	NOUN
cana-2345	36	2	recognition	recognition	NOUN
cana-2345	36	3	and	and	CCONJ
cana-2345	36	4	multi	multi	ADJ
cana-2345	36	5	-	-	ADJ
cana-2345	36	6	label	label	ADJ
cana-2345	36	7	classification	classification	NOUN
cana-2345	36	8	,	,	PUNCT
cana-2345	36	9	which	which	PRON
cana-2345	36	10	require	require	VERB
cana-2345	36	11	models	model	NOUN
cana-2345	36	12	to	to	PART
cana-2345	36	13	identify	identify	VERB
cana-2345	36	14	multiple	multiple	ADJ
cana-2345	36	15	sentiments	sentiment	NOUN
cana-2345	36	16	in	in	ADP
cana-2345	36	17	a	a	DET
cana-2345	36	18	single	single	ADJ
cana-2345	36	19	text	text	NOUN
cana-2345	36	20	,	,	PUNCT
cana-2345	36	21	are	be	AUX
cana-2345	36	22	also	also	ADV
cana-2345	36	23	challenging	challenge	VERB
cana-2345	36	24	areas	area	NOUN
cana-2345	36	25	where	where	SCONJ
cana-2345	36	26	existing	exist	VERB
cana-2345	36	27	methods	method	NOUN
cana-2345	36	28	fall	fall	VERB
cana-2345	36	29	short	short	ADJ
cana-2345	36	30	[	[	X
cana-2345	36	31	6][7	6][7	NUM
cana-2345	36	32	]	]	PUNCT
cana-2345	36	33	.	.	PUNCT
cana-2345	37	1	addressing	address	VERB
cana-2345	37	2	these	these	DET
cana-2345	37	3	gaps	gap	NOUN
cana-2345	37	4	is	be	AUX
cana-2345	37	5	essential	essential	ADJ
cana-2345	37	6	for	for	ADP
cana-2345	37	7	creating	create	VERB
cana-2345	37	8	more	more	ADV
cana-2345	37	9	effective	effective	ADJ
cana-2345	37	10	sentiment	sentiment	NOUN
cana-2345	37	11	analysis	analysis	NOUN
cana-2345	37	12	tools	tool	NOUN
cana-2345	37	13	that	that	PRON
cana-2345	37	14	work	work	VERB
cana-2345	37	15	well	well	ADV
cana-2345	37	16	across	across	ADP
cana-2345	37	17	different	different	ADJ
cana-2345	37	18	applications	application	NOUN
cana-2345	37	19	.	.	PUNCT
cana-2345	38	1	the	the	DET
cana-2345	38	2	motivation	motivation	NOUN
cana-2345	38	3	for	for	ADP
cana-2345	38	4	this	this	DET
cana-2345	38	5	research	research	NOUN
cana-2345	38	6	lies	lie	VERB
cana-2345	38	7	in	in	ADP
cana-2345	38	8	developing	develop	VERB
cana-2345	38	9	a	a	DET
cana-2345	38	10	nonlinear	nonlinear	ADJ
cana-2345	38	11	deep	deep	ADJ
cana-2345	38	12	learning	learning	NOUN
cana-2345	38	13	framework	framework	NOUN
cana-2345	38	14	that	that	PRON
cana-2345	38	15	integrates	integrate	VERB
cana-2345	38	16	different	different	ADJ
cana-2345	38	17	neural	neural	ADJ
cana-2345	38	18	models	model	NOUN
cana-2345	38	19	and	and	CCONJ
cana-2345	38	20	lexical	lexical	ADJ
cana-2345	38	21	approaches	approach	NOUN
cana-2345	38	22	to	to	PART
cana-2345	38	23	improve	improve	VERB
cana-2345	38	24	sentiment	sentiment	NOUN
cana-2345	38	25	classification	classification	NOUN
cana-2345	38	26	.	.	PUNCT
cana-2345	39	1	the	the	DET
cana-2345	39	2	aim	aim	NOUN
cana-2345	39	3	is	be	AUX
cana-2345	39	4	to	to	PART
cana-2345	39	5	create	create	VERB
cana-2345	39	6	a	a	DET
cana-2345	39	7	more	more	ADV
cana-2345	39	8	adaptable	adaptable	ADJ
cana-2345	39	9	and	and	CCONJ
cana-2345	39	10	accurate	accurate	ADJ
cana-2345	39	11	framework	framework	NOUN
cana-2345	39	12	that	that	PRON
cana-2345	39	13	addresses	address	VERB
cana-2345	39	14	the	the	DET
cana-2345	39	15	current	current	ADJ
cana-2345	39	16	limitations	limitation	NOUN
cana-2345	39	17	by	by	ADP
cana-2345	39	18	enhancing	enhance	VERB
cana-2345	39	19	how	how	SCONJ
cana-2345	39	20	sentiment	sentiment	NOUN
cana-2345	39	21	and	and	CCONJ
cana-2345	39	22	emotions	emotion	NOUN
cana-2345	39	23	are	be	AUX
cana-2345	39	24	captured	capture	VERB
cana-2345	39	25	and	and	CCONJ
cana-2345	39	26	analyzed	analyze	VERB
cana-2345	39	27	in	in	ADP
cana-2345	39	28	text	text	NOUN
cana-2345	39	29	.	.	PUNCT
cana-2345	40	1	this	this	DET
cana-2345	40	2	research	research	NOUN
cana-2345	40	3	introduces	introduce	VERB
cana-2345	40	4	a	a	DET
cana-2345	40	5	framework	framework	NOUN
cana-2345	40	6	that	that	PRON
cana-2345	40	7	combines	combine	VERB
cana-2345	40	8	neural	neural	ADJ
cana-2345	40	9	network	network	NOUN
cana-2345	40	10	models	model	NOUN
cana-2345	40	11	like	like	ADP
cana-2345	40	12	bi	bi	NOUN
cana-2345	40	13	-	-	NOUN
cana-2345	40	14	lstm	lstm	NOUN
cana-2345	40	15	and	and	CCONJ
cana-2345	40	16	gru	gru	VERB
cana-2345	40	17	with	with	ADP
cana-2345	40	18	traditional	traditional	ADJ
cana-2345	40	19	lexical	lexical	ADJ
cana-2345	40	20	methods	method	NOUN
cana-2345	40	21	to	to	PART
cana-2345	40	22	improve	improve	VERB
cana-2345	40	23	sentiment	sentiment	NOUN
cana-2345	40	24	analysis	analysis	NOUN
cana-2345	40	25	.	.	PUNCT
cana-2345	41	1	the	the	DET
cana-2345	41	2	key	key	ADJ
cana-2345	41	3	objectives	objective	NOUN
cana-2345	41	4	are	be	AUX
cana-2345	41	5	:	:	PUNCT
cana-2345	41	6	•	•	ADP
cana-2345	41	7	to	to	PART
cana-2345	41	8	enhance	enhance	VERB
cana-2345	41	9	sentiment	sentiment	NOUN
cana-2345	41	10	analysis	analysis	NOUN
cana-2345	41	11	by	by	ADP
cana-2345	41	12	integrating	integrate	VERB
cana-2345	41	13	neural	neural	ADJ
cana-2345	41	14	and	and	CCONJ
cana-2345	41	15	lexical	lexical	ADJ
cana-2345	41	16	methods	method	NOUN
cana-2345	41	17	,	,	PUNCT
cana-2345	41	18	capturing	capture	VERB
cana-2345	41	19	subtle	subtle	ADJ
cana-2345	41	20	and	and	CCONJ
cana-2345	41	21	complex	complex	ADJ
cana-2345	41	22	expressions	expression	NOUN
cana-2345	41	23	in	in	ADP
cana-2345	41	24	text	text	NOUN
cana-2345	41	25	[	[	X
cana-2345	41	26	4	4	NUM
cana-2345	41	27	]	]	PUNCT
cana-2345	41	28	.	.	PUNCT
cana-2345	42	1	•	•	NUM
cana-2345	42	2	to	to	PART
cana-2345	42	3	improve	improve	VERB
cana-2345	42	4	emotion	emotion	NOUN
cana-2345	42	5	recognition	recognition	NOUN
cana-2345	42	6	and	and	CCONJ
cana-2345	42	7	handle	handle	VERB
cana-2345	42	8	multi	multi	ADJ
cana-2345	42	9	-	-	ADJ
cana-2345	42	10	label	label	ADJ
cana-2345	42	11	classification	classification	NOUN
cana-2345	42	12	where	where	SCONJ
cana-2345	42	13	multiple	multiple	ADJ
cana-2345	42	14	sentiments	sentiment	NOUN
cana-2345	42	15	are	be	AUX
cana-2345	42	16	expressed	express	VERB
cana-2345	42	17	together	together	ADV
cana-2345	42	18	[	[	X
cana-2345	42	19	6][8	6][8	NOUN
cana-2345	42	20	]	]	X
cana-2345	42	21	.	.	PUNCT
cana-2345	43	1	•	•	NOUN
cana-2345	43	2	to	to	PART
cana-2345	43	3	create	create	VERB
cana-2345	43	4	a	a	DET
cana-2345	43	5	flexible	flexible	ADJ
cana-2345	43	6	framework	framework	NOUN
cana-2345	43	7	that	that	PRON
cana-2345	43	8	can	can	AUX
cana-2345	43	9	be	be	AUX
cana-2345	43	10	adapted	adapt	VERB
cana-2345	43	11	to	to	ADP
cana-2345	43	12	various	various	ADJ
cana-2345	43	13	domains	domain	NOUN
cana-2345	43	14	and	and	CCONJ
cana-2345	43	15	datasets	dataset	NOUN
cana-2345	43	16	by	by	ADP
cana-2345	43	17	incorporating	incorporate	VERB
cana-2345	43	18	domain	domain	NOUN
cana-2345	43	19	-	-	PUNCT
cana-2345	43	20	specific	specific	ADJ
cana-2345	43	21	lexicons	lexicon	NOUN
cana-2345	43	22	and	and	CCONJ
cana-2345	43	23	embeddings	embedding	NOUN
cana-2345	43	24	[	[	X
cana-2345	43	25	4	4	NUM
cana-2345	43	26	]	]	PUNCT
cana-2345	43	27	.	.	PUNCT
cana-2345	44	1	this	this	DET
cana-2345	44	2	study	study	NOUN
cana-2345	44	3	presents	present	VERB
cana-2345	44	4	a	a	DET
cana-2345	44	5	new	new	ADJ
cana-2345	44	6	approach	approach	NOUN
cana-2345	44	7	to	to	ADP
cana-2345	44	8	sentiment	sentiment	NOUN
cana-2345	44	9	analysis	analysis	NOUN
cana-2345	44	10	,	,	PUNCT
cana-2345	44	11	combining	combine	VERB
cana-2345	44	12	neural	neural	ADJ
cana-2345	44	13	models	model	NOUN
cana-2345	44	14	and	and	CCONJ
cana-2345	44	15	lexical	lexical	ADJ
cana-2345	44	16	techniques	technique	NOUN
cana-2345	44	17	to	to	PART
cana-2345	44	18	overcome	overcome	VERB
cana-2345	44	19	common	common	ADJ
cana-2345	44	20	challenges	challenge	NOUN
cana-2345	44	21	.	.	PUNCT
cana-2345	45	1	the	the	DET
cana-2345	45	2	proposed	propose	VERB
cana-2345	45	3	framework	framework	NOUN
cana-2345	45	4	improves	improve	VERB
cana-2345	45	5	accuracy	accuracy	NOUN
cana-2345	45	6	in	in	ADP
cana-2345	45	7	identifying	identify	VERB
cana-2345	45	8	complex	complex	ADJ
cana-2345	45	9	and	and	CCONJ
cana-2345	45	10	nuanced	nuanced	ADJ
cana-2345	45	11	sentiments	sentiment	NOUN
cana-2345	45	12	in	in	ADP
cana-2345	45	13	text	text	NOUN
cana-2345	45	14	.	.	PUNCT
cana-2345	46	1	this	this	PRON
cana-2345	46	2	makes	make	VERB
cana-2345	46	3	it	it	PRON
cana-2345	46	4	useful	useful	ADJ
cana-2345	46	5	for	for	ADP
cana-2345	46	6	practical	practical	ADJ
cana-2345	46	7	applications	application	NOUN
cana-2345	46	8	in	in	ADP
cana-2345	46	9	fields	field	NOUN
cana-2345	46	10	where	where	SCONJ
cana-2345	46	11	understanding	understand	VERB
cana-2345	46	12	public	public	ADJ
cana-2345	46	13	opinion	opinion	NOUN
cana-2345	46	14	and	and	CCONJ
cana-2345	46	15	emotional	emotional	ADJ
cana-2345	46	16	expression	expression	NOUN
cana-2345	46	17	is	be	AUX
cana-2345	46	18	important	important	ADJ
cana-2345	46	19	.	.	PUNCT
cana-2345	47	1	the	the	DET
cana-2345	47	2	research	research	NOUN
cana-2345	47	3	contributes	contribute	VERB
cana-2345	47	4	a	a	DET
cana-2345	47	5	flexible	flexible	ADJ
cana-2345	47	6	tool	tool	NOUN
cana-2345	47	7	that	that	PRON
cana-2345	47	8	can	can	AUX
cana-2345	47	9	be	be	AUX
cana-2345	47	10	tailored	tailor	VERB
cana-2345	47	11	to	to	ADP
cana-2345	47	12	specific	specific	ADJ
cana-2345	47	13	industries	industry	NOUN
cana-2345	47	14	,	,	PUNCT
cana-2345	47	15	enhancing	enhance	VERB
cana-2345	47	16	the	the	DET
cana-2345	47	17	value	value	NOUN
cana-2345	47	18	of	of	ADP
cana-2345	47	19	sentiment	sentiment	NOUN
cana-2345	47	20	analysis	analysis	NOUN
cana-2345	47	21	in	in	ADP
cana-2345	47	22	real	real	ADJ
cana-2345	47	23	-	-	PUNCT
cana-2345	47	24	world	world	NOUN
cana-2345	47	25	scenarios	scenario	NOUN
cana-2345	47	26	[	[	X
cana-2345	47	27	9][10	9][10	NUM
cana-2345	47	28	]	]	PUNCT
cana-2345	47	29	.	.	PUNCT
cana-2345	48	1	the	the	DET
cana-2345	48	2	paper	paper	NOUN
cana-2345	48	3	is	be	AUX
cana-2345	48	4	organized	organize	VERB
cana-2345	48	5	as	as	SCONJ
cana-2345	48	6	follows	follow	VERB
cana-2345	48	7	:	:	PUNCT
cana-2345	48	8	section	section	NOUN
cana-2345	48	9	2	2	NUM
cana-2345	48	10	reviews	review	NOUN
cana-2345	48	11	related	relate	VERB
cana-2345	48	12	work	work	NOUN
cana-2345	48	13	and	and	CCONJ
cana-2345	48	14	discusses	discuss	VERB
cana-2345	48	15	key	key	ADJ
cana-2345	48	16	deep	deep	ADJ
cana-2345	48	17	learning	learning	NOUN
cana-2345	48	18	architectures	architecture	NOUN
cana-2345	48	19	used	use	VERB
cana-2345	48	20	in	in	ADP
cana-2345	48	21	sentiment	sentiment	NOUN
cana-2345	48	22	analysis	analysis	NOUN
cana-2345	48	23	.	.	PUNCT
cana-2345	49	1	section	section	NOUN
cana-2345	49	2	3	3	NUM
cana-2345	49	3	details	detail	NOUN
cana-2345	49	4	the	the	DET
cana-2345	49	5	proposed	proposed	ADJ
cana-2345	49	6	nonlinear	nonlinear	ADJ
cana-2345	49	7	framework	framework	NOUN
cana-2345	49	8	,	,	PUNCT
cana-2345	49	9	explaining	explain	VERB
cana-2345	49	10	the	the	DET
cana-2345	49	11	integration	integration	NOUN
cana-2345	49	12	of	of	ADP
cana-2345	49	13	neural	neural	ADJ
cana-2345	49	14	and	and	CCONJ
cana-2345	49	15	lexical	lexical	ADJ
cana-2345	49	16	methods	method	NOUN
cana-2345	49	17	.	.	PUNCT
cana-2345	50	1	section	section	NOUN
cana-2345	50	2	4	4	NUM
cana-2345	50	3	covers	cover	VERB
cana-2345	50	4	the	the	DET
cana-2345	50	5	experiments	experiment	NOUN
cana-2345	50	6	and	and	CCONJ
cana-2345	50	7	evaluation	evaluation	NOUN
cana-2345	50	8	of	of	ADP
cana-2345	50	9	the	the	DET
cana-2345	50	10	framework	framework	NOUN
cana-2345	50	11	's	's	PART
cana-2345	50	12	performance	performance	NOUN
cana-2345	50	13	.	.	PUNCT
cana-2345	51	1	section	section	NOUN
cana-2345	51	2	5	5	NUM
cana-2345	51	3	discusses	discuss	VERB
cana-2345	51	4	the	the	DET
cana-2345	51	5	results	result	NOUN
cana-2345	51	6	,	,	PUNCT
cana-2345	51	7	highlighting	highlight	VERB
cana-2345	51	8	how	how	SCONJ
cana-2345	51	9	the	the	DET
cana-2345	51	10	framework	framework	NOUN
cana-2345	51	11	improves	improve	VERB
cana-2345	51	12	sentiment	sentiment	NOUN
cana-2345	51	13	classification	classification	NOUN
cana-2345	51	14	.	.	PUNCT
cana-2345	52	1	section	section	NOUN
cana-2345	52	2	6	6	NUM
cana-2345	52	3	concludes	conclude	VERB
cana-2345	52	4	with	with	ADP
cana-2345	52	5	the	the	DET
cana-2345	52	6	main	main	ADJ
cana-2345	52	7	findings	finding	NOUN
cana-2345	52	8	and	and	CCONJ
cana-2345	52	9	suggests	suggest	VERB
cana-2345	52	10	future	future	ADJ
cana-2345	52	11	directions	direction	NOUN
cana-2345	52	12	for	for	ADP
cana-2345	52	13	research	research	NOUN
cana-2345	52	14	.	.	PUNCT
cana-2345	53	1	this	this	DET
cana-2345	53	2	layout	layout	NOUN
cana-2345	53	3	aims	aim	VERB
cana-2345	53	4	to	to	PART
cana-2345	53	5	clearly	clearly	ADV
cana-2345	53	6	present	present	VERB
cana-2345	53	7	the	the	DET
cana-2345	53	8	research	research	NOUN
cana-2345	53	9	's	's	PART
cana-2345	53	10	goals	goal	NOUN
cana-2345	53	11	,	,	PUNCT
cana-2345	53	12	methods	method	NOUN
cana-2345	53	13	,	,	PUNCT
cana-2345	53	14	and	and	CCONJ
cana-2345	53	15	contributions	contribution	NOUN
cana-2345	53	16	.	.	PUNCT
cana-2345	54	1	communications	communication	NOUN
cana-2345	54	2	on	on	ADP
cana-2345	54	3	applied	apply	VERB
cana-2345	54	4	nonlinear	nonlinear	ADJ
cana-2345	54	5	analysis	analysis	NOUN
cana-2345	54	6	issn	issn	NOUN
cana-2345	54	7	:	:	PUNCT
cana-2345	54	8	1074	1074	NUM
cana-2345	54	9	-	-	PUNCT
cana-2345	54	10	133x	133x	NUM
cana-2345	54	11	vol	vol	NOUN
cana-2345	54	12	32	32	NUM
cana-2345	54	13	no	no	NOUN
cana-2345	54	14	.	.	PUNCT
cana-2345	55	1	1s	1s	NUM
cana-2345	55	2	(	(	PUNCT
cana-2345	55	3	2025	2025	NUM
cana-2345	55	4	)	)	PUNCT
cana-2345	55	5	575	575	NUM
cana-2345	55	6	https://internationalpubls.com	https://internationalpubls.com	SYM
cana-2345	55	7	1	1	NUM
cana-2345	55	8	related	relate	VERB
cana-2345	55	9	work	work	NOUN
cana-2345	55	10	recent	recent	ADJ
cana-2345	55	11	work	work	NOUN
cana-2345	55	12	in	in	ADP
cana-2345	55	13	sentiment	sentiment	NOUN
cana-2345	55	14	analysis	analysis	NOUN
cana-2345	55	15	focuses	focus	VERB
cana-2345	55	16	on	on	ADP
cana-2345	55	17	deep	deep	ADJ
cana-2345	55	18	learning	learning	NOUN
cana-2345	55	19	models	model	NOUN
cana-2345	55	20	to	to	PART
cana-2345	55	21	better	well	ADV
cana-2345	55	22	understand	understand	VERB
cana-2345	55	23	and	and	CCONJ
cana-2345	55	24	classify	classify	VERB
cana-2345	55	25	text	text	NOUN
cana-2345	55	26	data	datum	NOUN
cana-2345	55	27	.	.	PUNCT
cana-2345	56	1	different	different	ADJ
cana-2345	56	2	methods	method	NOUN
cana-2345	56	3	have	have	AUX
cana-2345	56	4	been	be	AUX
cana-2345	56	5	tried	try	VERB
cana-2345	56	6	to	to	PART
cana-2345	56	7	improve	improve	VERB
cana-2345	56	8	accuracy	accuracy	NOUN
cana-2345	56	9	,	,	PUNCT
cana-2345	56	10	handle	handle	VERB
cana-2345	56	11	various	various	ADJ
cana-2345	56	12	text	text	NOUN
cana-2345	56	13	types	type	NOUN
cana-2345	56	14	,	,	PUNCT
cana-2345	56	15	and	and	CCONJ
cana-2345	56	16	merge	merge	VERB
cana-2345	56	17	several	several	ADJ
cana-2345	56	18	techniques	technique	NOUN
cana-2345	56	19	.	.	PUNCT
cana-2345	57	1	research	research	NOUN
cana-2345	57	2	can	can	AUX
cana-2345	57	3	be	be	AUX
cana-2345	57	4	grouped	group	VERB
cana-2345	57	5	into	into	ADP
cana-2345	57	6	key	key	ADJ
cana-2345	57	7	areas	area	NOUN
cana-2345	57	8	based	base	VERB
cana-2345	57	9	on	on	ADP
cana-2345	57	10	the	the	DET
cana-2345	57	11	methods	method	NOUN
cana-2345	57	12	used	use	VERB
cana-2345	57	13	and	and	CCONJ
cana-2345	57	14	what	what	PRON
cana-2345	57	15	each	each	DET
cana-2345	57	16	study	study	NOUN
cana-2345	57	17	brings	bring	VERB
cana-2345	57	18	to	to	ADP
cana-2345	57	19	the	the	DET
cana-2345	57	20	field	field	NOUN
cana-2345	57	21	.	.	PUNCT
cana-2345	58	1	some	some	DET
cana-2345	58	2	studies	study	NOUN
cana-2345	58	3	mix	mix	VERB
cana-2345	58	4	neural	neural	ADJ
cana-2345	58	5	network	network	NOUN
cana-2345	58	6	models	model	NOUN
cana-2345	58	7	to	to	PART
cana-2345	58	8	improve	improve	VERB
cana-2345	58	9	how	how	SCONJ
cana-2345	58	10	sentiment	sentiment	NOUN
cana-2345	58	11	is	be	AUX
cana-2345	58	12	understood	understand	VERB
cana-2345	58	13	.	.	PUNCT
cana-2345	59	1	adilakshmi	adilakshmi	PROPN
cana-2345	59	2	et	et	PROPN
cana-2345	59	3	al	al	PROPN
cana-2345	59	4	.	.	PUNCT
cana-2345	60	1	[	[	X
cana-2345	60	2	11	11	NUM
cana-2345	60	3	]	]	PUNCT
cana-2345	60	4	used	use	VERB
cana-2345	60	5	a	a	DET
cana-2345	60	6	method	method	NOUN
cana-2345	60	7	called	call	VERB
cana-2345	60	8	rmdl	rmdl	NOUN
cana-2345	60	9	,	,	PUNCT
cana-2345	60	10	combining	combine	VERB
cana-2345	60	11	several	several	ADJ
cana-2345	60	12	neural	neural	ADJ
cana-2345	60	13	networks	network	NOUN
cana-2345	60	14	and	and	CCONJ
cana-2345	60	15	transfer	transfer	NOUN
cana-2345	60	16	learning	learn	VERB
cana-2345	60	17	to	to	PART
cana-2345	60	18	improve	improve	VERB
cana-2345	60	19	classification	classification	NOUN
cana-2345	60	20	in	in	ADP
cana-2345	60	21	context	context	NOUN
cana-2345	60	22	-	-	PUNCT
cana-2345	60	23	rich	rich	ADJ
cana-2345	60	24	text	text	NOUN
cana-2345	60	25	.	.	PUNCT
cana-2345	61	1	rose	rise	VERB
cana-2345	61	2	et	et	PROPN
cana-2345	61	3	al	al	PROPN
cana-2345	61	4	.	.	PUNCT
cana-2345	62	1	[	[	X
cana-2345	62	2	4	4	X
cana-2345	62	3	]	]	PUNCT
cana-2345	62	4	explored	explore	VERB
cana-2345	62	5	hybrid	hybrid	ADJ
cana-2345	62	6	models	model	NOUN
cana-2345	62	7	,	,	PUNCT
cana-2345	62	8	mixing	mix	VERB
cana-2345	62	9	traditional	traditional	ADJ
cana-2345	62	10	sentiment	sentiment	NOUN
cana-2345	62	11	analysis	analysis	NOUN
cana-2345	62	12	methods	method	NOUN
cana-2345	62	13	with	with	ADP
cana-2345	62	14	deep	deep	ADJ
cana-2345	62	15	learning	learning	NOUN
cana-2345	62	16	to	to	PART
cana-2345	62	17	handle	handle	VERB
cana-2345	62	18	complex	complex	ADJ
cana-2345	62	19	emotions	emotion	NOUN
cana-2345	62	20	in	in	ADP
cana-2345	62	21	text	text	NOUN
cana-2345	62	22	.	.	PUNCT
cana-2345	63	1	sachin	sachin	PROPN
cana-2345	63	2	sambhaji	sambhaji	PROPN
cana-2345	63	3	patil	patil	PROPN
cana-2345	63	4	et	et	PROPN
cana-2345	63	5	al	al	PROPN
cana-2345	63	6	.	.	PUNCT
cana-2345	64	1	[	[	X
cana-2345	64	2	12	12	NUM
cana-2345	64	3	]	]	PUNCT
cana-2345	64	4	used	use	VERB
cana-2345	64	5	cnns	cnn	NOUN
cana-2345	64	6	to	to	PART
cana-2345	64	7	capture	capture	VERB
cana-2345	64	8	the	the	DET
cana-2345	64	9	relationships	relationship	NOUN
cana-2345	64	10	between	between	ADP
cana-2345	64	11	words	word	NOUN
cana-2345	64	12	and	and	CCONJ
cana-2345	64	13	added	add	VERB
cana-2345	64	14	advanced	advanced	ADJ
cana-2345	64	15	pooling	pooling	NOUN
cana-2345	64	16	techniques	technique	NOUN
cana-2345	64	17	to	to	PART
cana-2345	64	18	boost	boost	VERB
cana-2345	64	19	accuracy	accuracy	NOUN
cana-2345	64	20	.	.	PUNCT
cana-2345	65	1	these	these	DET
cana-2345	65	2	approaches	approach	NOUN
cana-2345	65	3	highlight	highlight	VERB
cana-2345	65	4	how	how	SCONJ
cana-2345	65	5	mixing	mix	VERB
cana-2345	65	6	different	different	ADJ
cana-2345	65	7	models	model	NOUN
cana-2345	65	8	helps	help	VERB
cana-2345	65	9	understand	understand	VERB
cana-2345	65	10	text	text	NOUN
cana-2345	65	11	from	from	ADP
cana-2345	65	12	multiple	multiple	ADJ
cana-2345	65	13	viewpoints	viewpoint	NOUN
cana-2345	65	14	.	.	PUNCT
cana-2345	66	1	multi	multi	ADJ
cana-2345	66	2	-	-	NOUN
cana-2345	66	3	source	source	NOUN
cana-2345	66	4	and	and	CCONJ
cana-2345	66	5	transfer	transfer	NOUN
cana-2345	66	6	learning	learning	NOUN
cana-2345	66	7	approaches	approach	NOUN
cana-2345	66	8	have	have	AUX
cana-2345	66	9	been	be	AUX
cana-2345	66	10	used	use	VERB
cana-2345	66	11	to	to	PART
cana-2345	66	12	merge	merge	VERB
cana-2345	66	13	data	datum	NOUN
cana-2345	66	14	from	from	ADP
cana-2345	66	15	different	different	ADJ
cana-2345	66	16	places	place	NOUN
cana-2345	66	17	,	,	PUNCT
cana-2345	66	18	improving	improve	VERB
cana-2345	66	19	how	how	SCONJ
cana-2345	66	20	text	text	NOUN
cana-2345	66	21	is	be	AUX
cana-2345	66	22	classified	classified	ADJ
cana-2345	66	23	.	.	PUNCT
cana-2345	67	1	nguyen	nguyen	PROPN
cana-2345	67	2	et	et	PROPN
cana-2345	67	3	al	al	PROPN
cana-2345	67	4	.	.	PUNCT
cana-2345	68	1	[	[	X
cana-2345	68	2	13	13	NUM
cana-2345	68	3	]	]	PUNCT
cana-2345	68	4	,	,	PUNCT
cana-2345	68	5	suganya	suganya	ADJ
cana-2345	68	6	,	,	PUNCT
cana-2345	68	7	v.	v.	ADV
cana-2345	68	8	,	,	PUNCT
cana-2345	68	9	et	et	PROPN
cana-2345	68	10	al	al	PROPN
cana-2345	68	11	.	.	PROPN
cana-2345	68	12	,	,	PUNCT
cana-2345	69	1	[	[	X
cana-2345	69	2	14	14	NUM
cana-2345	69	3	]	]	PUNCT
cana-2345	69	4	introduced	introduce	VERB
cana-2345	69	5	lifa	lifa	NOUN
cana-2345	69	6	,	,	PUNCT
cana-2345	69	7	a	a	DET
cana-2345	69	8	framework	framework	NOUN
cana-2345	69	9	that	that	PRON
cana-2345	69	10	uses	use	VERB
cana-2345	69	11	transfer	transfer	NOUN
cana-2345	69	12	learning	learn	VERB
cana-2345	69	13	to	to	PART
cana-2345	69	14	pull	pull	VERB
cana-2345	69	15	together	together	ADV
cana-2345	69	16	information	information	NOUN
cana-2345	69	17	from	from	ADP
cana-2345	69	18	various	various	ADJ
cana-2345	69	19	pretrained	pretraine	VERB
cana-2345	69	20	models	model	NOUN
cana-2345	69	21	,	,	PUNCT
cana-2345	69	22	boosting	boost	VERB
cana-2345	69	23	sentiment	sentiment	NOUN
cana-2345	69	24	analysis	analysis	NOUN
cana-2345	69	25	across	across	ADP
cana-2345	69	26	different	different	ADJ
cana-2345	69	27	fields	field	NOUN
cana-2345	69	28	.	.	PUNCT
cana-2345	70	1	this	this	DET
cana-2345	70	2	technique	technique	NOUN
cana-2345	70	3	shows	show	VERB
cana-2345	70	4	the	the	DET
cana-2345	70	5	flexibility	flexibility	NOUN
cana-2345	70	6	of	of	ADP
cana-2345	70	7	transfer	transfer	NOUN
cana-2345	70	8	learning	learning	NOUN
cana-2345	70	9	in	in	ADP
cana-2345	70	10	refining	refining	NOUN
cana-2345	70	11	models	model	NOUN
cana-2345	70	12	,	,	PUNCT
cana-2345	70	13	though	though	SCONJ
cana-2345	70	14	it	it	PRON
cana-2345	70	15	also	also	ADV
cana-2345	70	16	needs	need	VERB
cana-2345	70	17	quality	quality	NOUN
cana-2345	70	18	data	datum	NOUN
cana-2345	70	19	to	to	PART
cana-2345	70	20	work	work	VERB
cana-2345	70	21	well	well	ADV
cana-2345	70	22	.	.	PUNCT
cana-2345	71	1	adilakshmi	adilakshmi	PROPN
cana-2345	71	2	et	et	PROPN
cana-2345	71	3	al	al	PROPN
cana-2345	71	4	.	.	PUNCT
cana-2345	72	1	[	[	X
cana-2345	72	2	11	11	NUM
cana-2345	72	3	]	]	PUNCT
cana-2345	72	4	also	also	ADV
cana-2345	72	5	showed	show	VERB
cana-2345	72	6	how	how	SCONJ
cana-2345	72	7	transfer	transfer	NOUN
cana-2345	72	8	learning	learning	NOUN
cana-2345	72	9	could	could	AUX
cana-2345	72	10	optimize	optimize	VERB
cana-2345	72	11	neural	neural	ADJ
cana-2345	72	12	networks	network	NOUN
cana-2345	72	13	,	,	PUNCT
cana-2345	72	14	cutting	cut	VERB
cana-2345	72	15	down	down	ADP
cana-2345	72	16	training	training	NOUN
cana-2345	72	17	time	time	NOUN
cana-2345	72	18	and	and	CCONJ
cana-2345	72	19	adapting	adapt	VERB
cana-2345	72	20	models	model	NOUN
cana-2345	72	21	to	to	ADP
cana-2345	72	22	new	new	ADJ
cana-2345	72	23	data	datum	NOUN
cana-2345	72	24	.	.	PUNCT
cana-2345	73	1	handling	handle	VERB
cana-2345	73	2	multiple	multiple	ADJ
cana-2345	73	3	emotions	emotion	NOUN
cana-2345	73	4	in	in	ADP
cana-2345	73	5	a	a	DET
cana-2345	73	6	single	single	ADJ
cana-2345	73	7	text	text	NOUN
cana-2345	73	8	is	be	AUX
cana-2345	73	9	challenging	challenge	VERB
cana-2345	73	10	.	.	PUNCT
cana-2345	74	1	priyanka	priyanka	PROPN
cana-2345	74	2	et	et	PROPN
cana-2345	74	3	al	al	PROPN
cana-2345	74	4	.	.	PUNCT
cana-2345	75	1	[	[	X
cana-2345	75	2	15	15	NUM
cana-2345	75	3	]	]	X
cana-2345	75	4	used	use	VERB
cana-2345	75	5	neural	neural	ADJ
cana-2345	75	6	networks	network	NOUN
cana-2345	75	7	to	to	PART
cana-2345	75	8	classify	classify	VERB
cana-2345	75	9	emotions	emotion	NOUN
cana-2345	75	10	and	and	CCONJ
cana-2345	75	11	recognize	recognize	VERB
cana-2345	75	12	complex	complex	ADJ
cana-2345	75	13	expressions	expression	NOUN
cana-2345	75	14	in	in	ADP
cana-2345	75	15	text	text	NOUN
cana-2345	75	16	.	.	PUNCT
cana-2345	76	1	radha	radha	PROPN
cana-2345	76	2	et	et	PROPN
cana-2345	76	3	al	al	PROPN
cana-2345	76	4	.	.	PUNCT
cana-2345	77	1	[	[	X
cana-2345	77	2	7	7	NUM
cana-2345	77	3	]	]	PUNCT
cana-2345	77	4	developed	develop	VERB
cana-2345	77	5	deep	deep	ADJ
cana-2345	77	6	learning	learning	NOUN
cana-2345	77	7	models	model	NOUN
cana-2345	77	8	to	to	PART
cana-2345	77	9	identify	identify	VERB
cana-2345	77	10	multiple	multiple	ADJ
cana-2345	77	11	emotions	emotion	NOUN
cana-2345	77	12	using	use	VERB
cana-2345	77	13	advanced	advanced	ADJ
cana-2345	77	14	embeddings	embedding	NOUN
cana-2345	77	15	,	,	PUNCT
cana-2345	77	16	showing	show	VERB
cana-2345	77	17	promise	promise	NOUN
cana-2345	77	18	in	in	ADP
cana-2345	77	19	managing	manage	VERB
cana-2345	77	20	different	different	ADJ
cana-2345	77	21	emotional	emotional	ADJ
cana-2345	77	22	layers	layer	NOUN
cana-2345	77	23	within	within	ADP
cana-2345	77	24	the	the	DET
cana-2345	77	25	same	same	ADJ
cana-2345	77	26	text	text	NOUN
cana-2345	77	27	.	.	PUNCT
cana-2345	78	1	chen	chen	PROPN
cana-2345	78	2	et	et	PROPN
cana-2345	78	3	al	al	PROPN
cana-2345	78	4	.	.	PUNCT
cana-2345	79	1	[	[	X
cana-2345	79	2	16	16	NUM
cana-2345	79	3	]	]	PUNCT
cana-2345	79	4	introduced	introduce	VERB
cana-2345	79	5	contrastive	contrastive	ADJ
cana-2345	79	6	learning	learning	NOUN
cana-2345	79	7	to	to	PART
cana-2345	79	8	refine	refine	VERB
cana-2345	79	9	how	how	SCONJ
cana-2345	79	10	models	model	NOUN
cana-2345	79	11	distinguish	distinguish	VERB
cana-2345	79	12	subtle	subtle	ADJ
cana-2345	79	13	emotional	emotional	ADJ
cana-2345	79	14	shifts	shift	NOUN
cana-2345	79	15	,	,	PUNCT
cana-2345	79	16	improving	improve	VERB
cana-2345	79	17	performance	performance	NOUN
cana-2345	79	18	in	in	ADP
cana-2345	79	19	scenarios	scenario	NOUN
cana-2345	79	20	where	where	SCONJ
cana-2345	79	21	data	datum	NOUN
cana-2345	79	22	is	be	AUX
cana-2345	79	23	limited	limited	ADJ
cana-2345	79	24	.	.	PUNCT
cana-2345	80	1	sentiment	sentiment	NOUN
cana-2345	80	2	analysis	analysis	NOUN
cana-2345	80	3	often	often	ADV
cana-2345	80	4	requires	require	VERB
cana-2345	80	5	adapting	adapt	VERB
cana-2345	80	6	models	model	NOUN
cana-2345	80	7	to	to	ADP
cana-2345	80	8	specific	specific	ADJ
cana-2345	80	9	language	language	NOUN
cana-2345	80	10	challenges	challenge	NOUN
cana-2345	80	11	.	.	PUNCT
cana-2345	81	1	jadon	jadon	PROPN
cana-2345	81	2	et	et	PROPN
cana-2345	81	3	al	al	PROPN
cana-2345	81	4	.	.	PUNCT
cana-2345	82	1	[	[	X
cana-2345	82	2	17	17	NUM
cana-2345	82	3	]	]	PUNCT
cana-2345	82	4	developed	develop	VERB
cana-2345	82	5	hybrid	hybrid	NOUN
cana-2345	82	6	models	model	NOUN
cana-2345	82	7	to	to	PART
cana-2345	82	8	handle	handle	VERB
cana-2345	82	9	mixed	mixed	ADJ
cana-2345	82	10	languages	language	NOUN
cana-2345	82	11	like	like	ADP
cana-2345	82	12	hinglish	hinglish	NOUN
cana-2345	82	13	,	,	PUNCT
cana-2345	82	14	capturing	capture	VERB
cana-2345	82	15	the	the	DET
cana-2345	82	16	unique	unique	ADJ
cana-2345	82	17	patterns	pattern	NOUN
cana-2345	82	18	of	of	ADP
cana-2345	82	19	such	such	ADJ
cana-2345	82	20	text	text	NOUN
cana-2345	82	21	.	.	PUNCT
cana-2345	83	1	gamal	gamal	PROPN
cana-2345	83	2	et	et	PROPN
cana-2345	83	3	al	al	PROPN
cana-2345	83	4	.	.	PUNCT
cana-2345	84	1	[	[	X
cana-2345	84	2	18	18	NUM
cana-2345	84	3	]	]	PUNCT
cana-2345	84	4	focused	focus	VERB
cana-2345	84	5	on	on	ADP
cana-2345	84	6	arabic	arabic	ADJ
cana-2345	84	7	language	language	NOUN
cana-2345	84	8	processing	processing	NOUN
cana-2345	84	9	,	,	PUNCT
cana-2345	84	10	using	use	VERB
cana-2345	84	11	neural	neural	ADJ
cana-2345	84	12	networks	network	NOUN
cana-2345	84	13	to	to	PART
cana-2345	84	14	manage	manage	VERB
cana-2345	84	15	dialects	dialect	NOUN
cana-2345	84	16	and	and	CCONJ
cana-2345	84	17	informal	informal	ADJ
cana-2345	84	18	styles	style	NOUN
cana-2345	84	19	,	,	PUNCT
cana-2345	84	20	demonstrating	demonstrate	VERB
cana-2345	84	21	that	that	SCONJ
cana-2345	84	22	tailored	tailor	VERB
cana-2345	84	23	models	model	NOUN
cana-2345	84	24	can	can	AUX
cana-2345	84	25	handle	handle	VERB
cana-2345	84	26	diverse	diverse	ADJ
cana-2345	84	27	language	language	NOUN
cana-2345	84	28	needs	need	NOUN
cana-2345	84	29	.	.	PUNCT
cana-2345	85	1	ensemble	ensemble	ADJ
cana-2345	85	2	and	and	CCONJ
cana-2345	85	3	active	active	ADJ
cana-2345	85	4	learning	learning	NOUN
cana-2345	85	5	methods	method	NOUN
cana-2345	85	6	bring	bring	VERB
cana-2345	85	7	together	together	ADV
cana-2345	85	8	different	different	ADJ
cana-2345	85	9	models	model	NOUN
cana-2345	85	10	to	to	PART
cana-2345	85	11	create	create	VERB
cana-2345	85	12	more	more	ADJ
cana-2345	85	13	robust	robust	ADJ
cana-2345	85	14	systems	system	NOUN
cana-2345	85	15	.	.	PUNCT
cana-2345	86	1	garg	garg	NOUN
cana-2345	86	2	et	et	PROPN
cana-2345	86	3	al	al	PROPN
cana-2345	86	4	.	.	PUNCT
cana-2345	87	1	[	[	X
cana-2345	87	2	19	19	NUM
cana-2345	87	3	]	]	PUNCT
cana-2345	87	4	combined	combine	VERB
cana-2345	87	5	multiple	multiple	ADJ
cana-2345	87	6	neural	neural	ADJ
cana-2345	87	7	networks	network	NOUN
cana-2345	87	8	in	in	ADP
cana-2345	87	9	an	an	DET
cana-2345	87	10	ensemble	ensemble	ADJ
cana-2345	87	11	approach	approach	NOUN
cana-2345	87	12	,	,	PUNCT
cana-2345	87	13	boosting	boost	VERB
cana-2345	87	14	classification	classification	NOUN
cana-2345	87	15	by	by	ADP
cana-2345	87	16	integrating	integrate	VERB
cana-2345	87	17	various	various	ADJ
cana-2345	87	18	perspectives	perspective	NOUN
cana-2345	87	19	of	of	ADP
cana-2345	87	20	the	the	DET
cana-2345	87	21	text	text	NOUN
cana-2345	87	22	.	.	PUNCT
cana-2345	88	1	raja	raja	PROPN
cana-2345	88	2	et	et	PROPN
cana-2345	88	3	al	al	PROPN
cana-2345	88	4	.	.	PUNCT
cana-2345	89	1	[	[	X
cana-2345	89	2	20	20	NUM
cana-2345	89	3	]	]	PUNCT
cana-2345	89	4	used	use	VERB
cana-2345	89	5	active	active	ADJ
cana-2345	89	6	learning	learning	NOUN
cana-2345	89	7	strategies	strategy	NOUN
cana-2345	89	8	to	to	PART
cana-2345	89	9	selectively	selectively	ADV
cana-2345	89	10	train	train	NOUN
cana-2345	89	11	models	model	NOUN
cana-2345	89	12	on	on	ADP
cana-2345	89	13	the	the	DET
cana-2345	89	14	most	most	ADV
cana-2345	89	15	informative	informative	ADJ
cana-2345	89	16	data	datum	NOUN
cana-2345	89	17	points	point	NOUN
cana-2345	89	18	,	,	PUNCT
cana-2345	89	19	enhancing	enhance	VERB
cana-2345	89	20	accuracy	accuracy	NOUN
cana-2345	89	21	,	,	PUNCT
cana-2345	89	22	especially	especially	ADV
cana-2345	89	23	in	in	ADP
cana-2345	89	24	unbalanced	unbalanced	ADJ
cana-2345	89	25	datasets	dataset	NOUN
cana-2345	89	26	.	.	PUNCT
cana-2345	90	1	these	these	DET
cana-2345	90	2	methods	method	NOUN
cana-2345	90	3	show	show	VERB
cana-2345	90	4	how	how	SCONJ
cana-2345	90	5	combining	combine	VERB
cana-2345	90	6	and	and	CCONJ
cana-2345	90	7	refining	refining	NOUN
cana-2345	90	8	learning	learning	NOUN
cana-2345	90	9	techniques	technique	NOUN
cana-2345	90	10	can	can	AUX
cana-2345	90	11	lead	lead	VERB
cana-2345	90	12	to	to	ADP
cana-2345	90	13	better	well	ADJ
cana-2345	90	14	sentiment	sentiment	NOUN
cana-2345	90	15	classification	classification	NOUN
cana-2345	90	16	results	result	NOUN
cana-2345	90	17	.	.	PUNCT
cana-2345	91	1	some	some	DET
cana-2345	91	2	studies	study	NOUN
cana-2345	91	3	explore	explore	VERB
cana-2345	91	4	ways	way	NOUN
cana-2345	91	5	to	to	PART
cana-2345	91	6	improve	improve	VERB
cana-2345	91	7	sentiment	sentiment	NOUN
cana-2345	91	8	classification	classification	NOUN
cana-2345	91	9	by	by	ADP
cana-2345	91	10	focusing	focus	VERB
cana-2345	91	11	on	on	ADP
cana-2345	91	12	the	the	DET
cana-2345	91	13	relationships	relationship	NOUN
cana-2345	91	14	between	between	ADP
cana-2345	91	15	words	word	NOUN
cana-2345	91	16	and	and	CCONJ
cana-2345	91	17	refining	refining	NOUN
cana-2345	91	18	feature	feature	NOUN
cana-2345	91	19	extraction	extraction	NOUN
cana-2345	91	20	.	.	PUNCT
cana-2345	92	1	liu	liu	PROPN
cana-2345	92	2	et	et	PROPN
cana-2345	92	3	al	al	PROPN
cana-2345	92	4	.	.	PUNCT
cana-2345	93	1	[	[	X
cana-2345	93	2	5	5	NUM
cana-2345	93	3	]	]	PUNCT
cana-2345	93	4	,	,	PUNCT
cana-2345	93	5	and	and	CCONJ
cana-2345	93	6	kiran	kiran	PROPN
cana-2345	93	7	,	,	PUNCT
cana-2345	93	8	k.	k.	PROPN
cana-2345	93	9	,	,	PUNCT
cana-2345	93	10	abhishek	abhishek	PROPN
cana-2345	93	11	appaji	appaji	PROPN
cana-2345	93	12	et	et	PROPN
cana-2345	93	13	al	al	PROPN
cana-2345	93	14	.	.	PROPN
cana-2345	93	15	,	,	PUNCT
cana-2345	94	1	[	[	X
cana-2345	94	2	21	21	NUM
cana-2345	94	3	]	]	PUNCT
cana-2345	94	4	proposed	propose	VERB
cana-2345	94	5	a	a	DET
cana-2345	94	6	dependency	dependency	NOUN
cana-2345	94	7	-	-	PUNCT
cana-2345	94	8	based	base	VERB
cana-2345	94	9	model	model	NOUN
cana-2345	94	10	that	that	PRON
cana-2345	94	11	combines	combine	VERB
cana-2345	94	12	grammar	grammar	NOUN
cana-2345	94	13	rules	rule	NOUN
cana-2345	94	14	with	with	ADP
cana-2345	94	15	neural	neural	ADJ
cana-2345	94	16	networks	network	NOUN
cana-2345	94	17	to	to	PART
cana-2345	94	18	capture	capture	VERB
cana-2345	94	19	more	more	ADV
cana-2345	94	20	detailed	detailed	ADJ
cana-2345	94	21	meanings	meaning	NOUN
cana-2345	94	22	in	in	ADP
cana-2345	94	23	complex	complex	ADJ
cana-2345	94	24	sentences	sentence	NOUN
cana-2345	94	25	.	.	PUNCT
cana-2345	95	1	chen	chen	PROPN
cana-2345	95	2	et	et	PROPN
cana-2345	95	3	al	al	PROPN
cana-2345	95	4	.	.	PUNCT
cana-2345	96	1	[	[	X
cana-2345	96	2	16	16	NUM
cana-2345	96	3	]	]	PUNCT
cana-2345	96	4	used	use	VERB
cana-2345	96	5	contrastive	contrastive	ADJ
cana-2345	96	6	learning	learning	NOUN
cana-2345	96	7	communications	communication	NOUN
cana-2345	96	8	on	on	ADP
cana-2345	96	9	applied	apply	VERB
cana-2345	96	10	nonlinear	nonlinear	ADJ
cana-2345	96	11	analysis	analysis	NOUN
cana-2345	96	12	issn	issn	NOUN
cana-2345	96	13	:	:	PUNCT
cana-2345	96	14	1074	1074	NUM
cana-2345	96	15	-	-	PUNCT
cana-2345	96	16	133x	133x	NUM
cana-2345	96	17	vol	vol	NOUN
cana-2345	96	18	32	32	NUM
cana-2345	96	19	no	no	NOUN
cana-2345	96	20	.	.	PUNCT
cana-2345	97	1	1s	1s	NUM
cana-2345	97	2	(	(	PUNCT
cana-2345	97	3	2025	2025	NUM
cana-2345	97	4	)	)	PUNCT
cana-2345	97	5	576	576	NUM
cana-2345	97	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2345	97	7	techniques	technique	NOUN
cana-2345	97	8	to	to	PART
cana-2345	97	9	sharpen	sharpen	VERB
cana-2345	97	10	how	how	SCONJ
cana-2345	97	11	models	model	NOUN
cana-2345	97	12	differentiate	differentiate	VERB
cana-2345	97	13	subtle	subtle	ADJ
cana-2345	97	14	sentiment	sentiment	NOUN
cana-2345	97	15	changes	change	NOUN
cana-2345	97	16	,	,	PUNCT
cana-2345	97	17	which	which	PRON
cana-2345	97	18	is	be	AUX
cana-2345	97	19	particularly	particularly	ADV
cana-2345	97	20	useful	useful	ADJ
cana-2345	97	21	when	when	SCONJ
cana-2345	97	22	working	work	VERB
cana-2345	97	23	with	with	ADP
cana-2345	97	24	less	less	ADJ
cana-2345	97	25	data	datum	NOUN
cana-2345	97	26	.	.	PUNCT
cana-2345	98	1	social	social	ADJ
cana-2345	98	2	media	medium	NOUN
cana-2345	98	3	presents	present	VERB
cana-2345	98	4	unique	unique	ADJ
cana-2345	98	5	challenges	challenge	NOUN
cana-2345	98	6	due	due	ADP
cana-2345	98	7	to	to	ADP
cana-2345	98	8	its	its	PRON
cana-2345	98	9	fast	fast	ADV
cana-2345	98	10	-	-	PUNCT
cana-2345	98	11	changing	change	VERB
cana-2345	98	12	and	and	CCONJ
cana-2345	98	13	noisy	noisy	ADJ
cana-2345	98	14	nature	nature	NOUN
cana-2345	98	15	.	.	PUNCT
cana-2345	99	1	rangarjan	rangarjan	VERB
cana-2345	99	2	et	et	PROPN
cana-2345	99	3	al	al	PROPN
cana-2345	99	4	.	.	PUNCT
cana-2345	100	1	[	[	X
cana-2345	100	2	22	22	NUM
cana-2345	100	3	]	]	PUNCT
cana-2345	100	4	developed	develop	VERB
cana-2345	100	5	frameworks	framework	NOUN
cana-2345	100	6	integrating	integrate	VERB
cana-2345	100	7	neural	neural	ADJ
cana-2345	100	8	networks	network	NOUN
cana-2345	100	9	to	to	PART
cana-2345	100	10	capture	capture	VERB
cana-2345	100	11	real	real	ADJ
cana-2345	100	12	-	-	PUNCT
cana-2345	100	13	time	time	NOUN
cana-2345	100	14	sentiment	sentiment	NOUN
cana-2345	100	15	trends	trend	NOUN
cana-2345	100	16	,	,	PUNCT
cana-2345	100	17	providing	provide	VERB
cana-2345	100	18	useful	useful	ADJ
cana-2345	100	19	insights	insight	NOUN
cana-2345	100	20	for	for	ADP
cana-2345	100	21	applications	application	NOUN
cana-2345	100	22	like	like	ADP
cana-2345	100	23	tracking	track	VERB
cana-2345	100	24	public	public	ADJ
cana-2345	100	25	opinion	opinion	NOUN
cana-2345	100	26	.	.	PUNCT
cana-2345	101	1	kathiravan	kathiravan	PROPN
cana-2345	101	2	et	et	PROPN
cana-2345	101	3	al	al	PROPN
cana-2345	101	4	.	.	PUNCT
cana-2345	102	1	[	[	X
cana-2345	102	2	6	6	NUM
cana-2345	102	3	]	]	PUNCT
cana-2345	102	4	and	and	CCONJ
cana-2345	102	5	hernández	hernández	PROPN
cana-2345	102	6	,	,	PUNCT
cana-2345	102	7	nayeli	nayeli	PROPN
cana-2345	102	8	et	et	PROPN
cana-2345	102	9	al	al	PROPN
cana-2345	102	10	.	.	PROPN
cana-2345	102	11	,	,	PUNCT
cana-2345	103	1	[	[	X
cana-2345	103	2	23	23	NUM
cana-2345	103	3	]	]	PUNCT
cana-2345	103	4	focused	focus	VERB
cana-2345	103	5	on	on	ADP
cana-2345	103	6	deep	deep	ADJ
cana-2345	103	7	learning	learning	NOUN
cana-2345	103	8	methods	method	NOUN
cana-2345	103	9	like	like	ADP
cana-2345	103	10	lstm	lstm	NOUN
cana-2345	103	11	and	and	CCONJ
cana-2345	103	12	gru	gru	VERB
cana-2345	103	13	to	to	PART
cana-2345	103	14	analyze	analyze	VERB
cana-2345	103	15	large	large	ADJ
cana-2345	103	16	volumes	volume	NOUN
cana-2345	103	17	of	of	ADP
cana-2345	103	18	social	social	ADJ
cana-2345	103	19	media	medium	NOUN
cana-2345	103	20	text	text	NOUN
cana-2345	103	21	,	,	PUNCT
cana-2345	103	22	capturing	capture	VERB
cana-2345	103	23	hidden	hidden	ADJ
cana-2345	103	24	emotional	emotional	ADJ
cana-2345	103	25	cues	cue	NOUN
cana-2345	103	26	and	and	CCONJ
cana-2345	103	27	improving	improve	VERB
cana-2345	103	28	classification	classification	NOUN
cana-2345	103	29	efficiency	efficiency	NOUN
cana-2345	103	30	.	.	PUNCT
cana-2345	104	1	the	the	DET
cana-2345	104	2	research	research	NOUN
cana-2345	104	3	shows	show	VERB
cana-2345	104	4	that	that	SCONJ
cana-2345	104	5	combining	combine	VERB
cana-2345	104	6	various	various	ADJ
cana-2345	104	7	deep	deep	ADJ
cana-2345	104	8	learning	learning	NOUN
cana-2345	104	9	and	and	CCONJ
cana-2345	104	10	traditional	traditional	ADJ
cana-2345	104	11	approaches	approach	NOUN
cana-2345	104	12	improves	improve	VERB
cana-2345	104	13	sentiment	sentiment	NOUN
cana-2345	104	14	analysis	analysis	NOUN
cana-2345	104	15	in	in	ADP
cana-2345	104	16	different	different	ADJ
cana-2345	104	17	contexts	contexts	NOUN
cana-2345	104	18	.	.	PUNCT
cana-2345	105	1	by	by	ADP
cana-2345	105	2	using	use	VERB
cana-2345	105	3	hybrid	hybrid	ADJ
cana-2345	105	4	models	model	NOUN
cana-2345	105	5	,	,	PUNCT
cana-2345	105	6	transfer	transfer	NOUN
cana-2345	105	7	learning	learning	NOUN
cana-2345	105	8	,	,	PUNCT
cana-2345	105	9	and	and	CCONJ
cana-2345	105	10	techniques	technique	NOUN
cana-2345	105	11	like	like	ADP
cana-2345	105	12	contrastive	contrastive	ADJ
cana-2345	105	13	learning	learning	NOUN
cana-2345	105	14	,	,	PUNCT
cana-2345	105	15	sentiment	sentiment	NOUN
cana-2345	105	16	classification	classification	NOUN
cana-2345	105	17	becomes	become	VERB
cana-2345	105	18	more	more	ADV
cana-2345	105	19	precise	precise	ADJ
cana-2345	105	20	and	and	CCONJ
cana-2345	105	21	adaptable	adaptable	ADJ
cana-2345	105	22	.	.	PUNCT
cana-2345	106	1	however	however	ADV
cana-2345	106	2	,	,	PUNCT
cana-2345	106	3	balancing	balance	VERB
cana-2345	106	4	accuracy	accuracy	NOUN
cana-2345	106	5	with	with	ADP
cana-2345	106	6	the	the	DET
cana-2345	106	7	need	need	NOUN
cana-2345	106	8	for	for	ADP
cana-2345	106	9	simpler	simple	ADJ
cana-2345	106	10	,	,	PUNCT
cana-2345	106	11	faster	fast	ADJ
cana-2345	106	12	models	model	NOUN
cana-2345	106	13	remains	remain	VERB
cana-2345	106	14	a	a	DET
cana-2345	106	15	challenge	challenge	NOUN
cana-2345	106	16	,	,	PUNCT
cana-2345	106	17	highlighting	highlight	VERB
cana-2345	106	18	the	the	DET
cana-2345	106	19	ongoing	ongoing	ADJ
cana-2345	106	20	need	need	NOUN
cana-2345	106	21	to	to	PART
cana-2345	106	22	refine	refine	VERB
cana-2345	106	23	these	these	DET
cana-2345	106	24	methods	method	NOUN
cana-2345	106	25	for	for	ADP
cana-2345	106	26	broader	broad	ADJ
cana-2345	106	27	,	,	PUNCT
cana-2345	106	28	real	real	ADJ
cana-2345	106	29	-	-	PUNCT
cana-2345	106	30	world	world	NOUN
cana-2345	106	31	use	use	NOUN
cana-2345	106	32	.	.	PUNCT
cana-2345	107	1	2	2	NUM
cana-2345	107	2	methods	method	NOUN
cana-2345	107	3	and	and	CCONJ
cana-2345	107	4	materials	material	NOUN
cana-2345	107	5	the	the	DET
cana-2345	107	6	proposed	propose	VERB
cana-2345	107	7	nonlinear	nonlinear	ADJ
cana-2345	107	8	sentiment	sentiment	NOUN
cana-2345	107	9	analysis	analysis	NOUN
cana-2345	107	10	model	model	NOUN
cana-2345	107	11	architecture	architecture	NOUN
cana-2345	107	12	is	be	AUX
cana-2345	107	13	designed	design	VERB
cana-2345	107	14	to	to	PART
cana-2345	107	15	capture	capture	VERB
cana-2345	107	16	complex	complex	ADJ
cana-2345	107	17	dependencies	dependency	NOUN
cana-2345	107	18	in	in	ADP
cana-2345	107	19	textual	textual	ADJ
cana-2345	107	20	data	datum	NOUN
cana-2345	107	21	through	through	ADP
cana-2345	107	22	a	a	DET
cana-2345	107	23	sequence	sequence	NOUN
cana-2345	107	24	of	of	ADP
cana-2345	107	25	specialized	specialized	ADJ
cana-2345	107	26	layers	layer	NOUN
cana-2345	107	27	.	.	PUNCT
cana-2345	108	1	initially	initially	ADV
cana-2345	108	2	,	,	PUNCT
cana-2345	108	3	raw	raw	ADJ
cana-2345	108	4	text	text	NOUN
cana-2345	108	5	undergoes	undergoe	NOUN
cana-2345	108	6	preprocessing	preprocesse	VERB
cana-2345	108	7	and	and	CCONJ
cana-2345	108	8	embedding	embed	VERB
cana-2345	108	9	using	use	VERB
cana-2345	108	10	bert	bert	NOUN
cana-2345	108	11	,	,	PUNCT
cana-2345	108	12	transforming	transform	VERB
cana-2345	108	13	words	word	NOUN
cana-2345	108	14	into	into	ADP
cana-2345	108	15	dense	dense	ADJ
cana-2345	108	16	,	,	PUNCT
cana-2345	108	17	context	context	NOUN
cana-2345	108	18	-	-	PUNCT
cana-2345	108	19	rich	rich	ADJ
cana-2345	108	20	vectors	vector	NOUN
cana-2345	108	21	.	.	PUNCT
cana-2345	109	1	these	these	DET
cana-2345	109	2	vectors	vector	NOUN
cana-2345	109	3	feed	feed	VERB
cana-2345	109	4	into	into	ADP
cana-2345	109	5	a	a	DET
cana-2345	109	6	bidirectional	bidirectional	ADJ
cana-2345	109	7	lstm	lstm	NOUN
cana-2345	109	8	,	,	PUNCT
cana-2345	109	9	which	which	PRON
cana-2345	109	10	captures	capture	VERB
cana-2345	109	11	sequential	sequential	ADJ
cana-2345	109	12	context	context	NOUN
cana-2345	109	13	from	from	ADP
cana-2345	109	14	both	both	PRON
cana-2345	109	15	past	past	NOUN
cana-2345	109	16	and	and	CCONJ
cana-2345	109	17	future	future	ADJ
cana-2345	109	18	words	word	NOUN
cana-2345	109	19	,	,	PUNCT
cana-2345	109	20	creating	create	VERB
cana-2345	109	21	comprehensive	comprehensive	ADJ
cana-2345	109	22	hidden	hide	VERB
cana-2345	109	23	states	state	NOUN
cana-2345	109	24	for	for	ADP
cana-2345	109	25	each	each	DET
cana-2345	109	26	word	word	NOUN
cana-2345	109	27	.	.	PUNCT
cana-2345	110	1	these	these	DET
cana-2345	110	2	hidden	hide	VERB
cana-2345	110	3	states	state	NOUN
cana-2345	110	4	are	be	AUX
cana-2345	110	5	further	far	ADV
cana-2345	110	6	refined	refine	VERB
cana-2345	110	7	by	by	ADP
cana-2345	110	8	a	a	DET
cana-2345	110	9	convolutional	convolutional	ADJ
cana-2345	110	10	neural	neural	ADJ
cana-2345	110	11	network	network	NOUN
cana-2345	110	12	(	(	PUNCT
cana-2345	110	13	cnn	cnn	PROPN
cana-2345	110	14	)	)	PUNCT
cana-2345	110	15	that	that	PRON
cana-2345	110	16	identifies	identify	VERB
cana-2345	110	17	crucial	crucial	ADJ
cana-2345	110	18	local	local	ADJ
cana-2345	110	19	patterns	pattern	NOUN
cana-2345	110	20	and	and	CCONJ
cana-2345	110	21	phrases	phrase	NOUN
cana-2345	110	22	,	,	PUNCT
cana-2345	110	23	emphasizing	emphasize	VERB
cana-2345	110	24	sentiment	sentiment	NOUN
cana-2345	110	25	-	-	PUNCT
cana-2345	110	26	carrying	carry	VERB
cana-2345	110	27	elements	element	NOUN
cana-2345	110	28	like	like	ADP
cana-2345	110	29	adjectives	adjective	NOUN
cana-2345	110	30	and	and	CCONJ
cana-2345	110	31	adverbs	adverb	NOUN
cana-2345	110	32	through	through	ADP
cana-2345	110	33	convolution	convolution	NOUN
cana-2345	110	34	and	and	CCONJ
cana-2345	110	35	pooling	pooling	NOUN
cana-2345	110	36	.	.	PUNCT
cana-2345	111	1	the	the	DET
cana-2345	111	2	attention	attention	NOUN
cana-2345	111	3	mechanism	mechanism	NOUN
cana-2345	111	4	then	then	ADV
cana-2345	111	5	dynamically	dynamically	ADV
cana-2345	111	6	adjusts	adjust	VERB
cana-2345	111	7	focus	focus	NOUN
cana-2345	111	8	on	on	ADP
cana-2345	111	9	sentiment	sentiment	NOUN
cana-2345	111	10	-	-	PUNCT
cana-2345	111	11	critical	critical	ADJ
cana-2345	111	12	words	word	NOUN
cana-2345	111	13	,	,	PUNCT
cana-2345	111	14	enhancing	enhance	VERB
cana-2345	111	15	their	their	PRON
cana-2345	111	16	influence	influence	NOUN
cana-2345	111	17	in	in	ADP
cana-2345	111	18	the	the	DET
cana-2345	111	19	analysis	analysis	NOUN
cana-2345	111	20	.	.	PUNCT
cana-2345	112	1	a	a	DET
cana-2345	112	2	dense	dense	ADJ
cana-2345	112	3	nonlinear	nonlinear	ADJ
cana-2345	112	4	feature	feature	NOUN
cana-2345	112	5	fusion	fusion	NOUN
cana-2345	112	6	layer	layer	NOUN
cana-2345	112	7	integrates	integrate	VERB
cana-2345	112	8	the	the	DET
cana-2345	112	9	outputs	output	NOUN
cana-2345	112	10	from	from	ADP
cana-2345	112	11	bilstm	bilstm	NOUN
cana-2345	112	12	,	,	PUNCT
cana-2345	112	13	cnn	cnn	PROPN
cana-2345	112	14	,	,	PUNCT
cana-2345	112	15	and	and	CCONJ
cana-2345	112	16	attention	attention	NOUN
cana-2345	112	17	,	,	PUNCT
cana-2345	112	18	applying	apply	VERB
cana-2345	112	19	relu	relu	NOUN
cana-2345	112	20	activation	activation	NOUN
cana-2345	112	21	to	to	PART
cana-2345	112	22	capture	capture	VERB
cana-2345	112	23	intricate	intricate	ADJ
cana-2345	112	24	feature	feature	NOUN
cana-2345	112	25	interdependencies	interdependency	NOUN
cana-2345	112	26	while	while	SCONJ
cana-2345	112	27	dropout	dropout	NOUN
cana-2345	112	28	regularization	regularization	NOUN
cana-2345	112	29	prevents	prevent	VERB
cana-2345	112	30	overfitting	overfitte	VERB
cana-2345	112	31	.	.	PUNCT
cana-2345	113	1	finally	finally	ADV
cana-2345	113	2	,	,	PUNCT
cana-2345	113	3	the	the	DET
cana-2345	113	4	sentiment	sentiment	NOUN
cana-2345	113	5	prediction	prediction	NOUN
cana-2345	113	6	layer	layer	NOUN
cana-2345	113	7	uses	use	VERB
cana-2345	113	8	a	a	DET
cana-2345	113	9	fully	fully	ADV
cana-2345	113	10	connected	connect	VERB
cana-2345	113	11	network	network	NOUN
cana-2345	113	12	followed	follow	VERB
cana-2345	113	13	by	by	ADP
cana-2345	113	14	a	a	DET
cana-2345	113	15	softmax	softmax	NOUN
cana-2345	113	16	or	or	CCONJ
cana-2345	113	17	sigmoid	sigmoid	NOUN
cana-2345	113	18	activation	activation	NOUN
cana-2345	113	19	to	to	PART
cana-2345	113	20	produce	produce	VERB
cana-2345	113	21	the	the	DET
cana-2345	113	22	sentiment	sentiment	NOUN
cana-2345	113	23	classification	classification	NOUN
cana-2345	113	24	with	with	ADP
cana-2345	113	25	high	high	ADJ
cana-2345	113	26	precision	precision	NOUN
cana-2345	113	27	,	,	PUNCT
cana-2345	113	28	translating	translate	VERB
cana-2345	113	29	complex	complex	ADJ
cana-2345	113	30	text	text	NOUN
cana-2345	113	31	patterns	pattern	NOUN
cana-2345	113	32	into	into	ADP
cana-2345	113	33	clear	clear	ADJ
cana-2345	113	34	,	,	PUNCT
cana-2345	113	35	confident	confident	ADJ
cana-2345	113	36	sentiment	sentiment	NOUN
cana-2345	113	37	scores	score	NOUN
cana-2345	113	38	.	.	PUNCT
cana-2345	114	1	this	this	DET
cana-2345	114	2	architecture	architecture	NOUN
cana-2345	114	3	excels	excel	VERB
cana-2345	114	4	in	in	ADP
cana-2345	114	5	extracting	extract	VERB
cana-2345	114	6	and	and	CCONJ
cana-2345	114	7	leveraging	leverage	VERB
cana-2345	114	8	subtle	subtle	ADJ
cana-2345	114	9	,	,	PUNCT
cana-2345	114	10	non	non	ADJ
cana-2345	114	11	-	-	ADJ
cana-2345	114	12	linear	linear	ADJ
cana-2345	114	13	relationships	relationship	NOUN
cana-2345	114	14	within	within	ADP
cana-2345	114	15	text	text	NOUN
cana-2345	114	16	for	for	ADP
cana-2345	114	17	highly	highly	ADV
cana-2345	114	18	accurate	accurate	ADJ
cana-2345	114	19	sentiment	sentiment	NOUN
cana-2345	114	20	predictions	prediction	NOUN
cana-2345	114	21	.	.	PUNCT
cana-2345	115	1	communications	communication	NOUN
cana-2345	115	2	on	on	ADP
cana-2345	115	3	applied	apply	VERB
cana-2345	115	4	nonlinear	nonlinear	ADJ
cana-2345	115	5	analysis	analysis	NOUN
cana-2345	115	6	issn	issn	NOUN
cana-2345	115	7	:	:	PUNCT
cana-2345	115	8	1074	1074	NUM
cana-2345	115	9	-	-	PUNCT
cana-2345	115	10	133x	133x	NUM
cana-2345	115	11	vol	vol	NOUN
cana-2345	115	12	32	32	NUM
cana-2345	115	13	no	no	NOUN
cana-2345	115	14	.	.	PUNCT
cana-2345	116	1	1s	1s	NUM
cana-2345	116	2	(	(	PUNCT
cana-2345	116	3	2025	2025	NUM
cana-2345	116	4	)	)	PUNCT
cana-2345	116	5	577	577	NUM
cana-2345	116	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2345	116	7	figure	figure	NOUN
cana-2345	116	8	1	1	NUM
cana-2345	116	9	:	:	PUNCT
cana-2345	116	10	architecture	architecture	NOUN
cana-2345	116	11	of	of	ADP
cana-2345	116	12	nonlinear	nonlinear	ADJ
cana-2345	116	13	sentiment	sentiment	NOUN
cana-2345	116	14	analysis	analysis	NOUN
cana-2345	116	15	model	model	NOUN
cana-2345	116	16	the	the	DET
cana-2345	116	17	diagram	diagram	NOUN
cana-2345	116	18	in	in	ADP
cana-2345	116	19	figure	figure	NOUN
cana-2345	116	20	1	1	NUM
cana-2345	116	21	illustrates	illustrate	VERB
cana-2345	116	22	the	the	DET
cana-2345	116	23	architecture	architecture	NOUN
cana-2345	116	24	of	of	ADP
cana-2345	116	25	a	a	DET
cana-2345	116	26	nonlinear	nonlinear	ADJ
cana-2345	116	27	sentiment	sentiment	NOUN
cana-2345	116	28	analysis	analysis	NOUN
cana-2345	116	29	model	model	NOUN
cana-2345	116	30	,	,	PUNCT
cana-2345	116	31	showing	show	VERB
cana-2345	116	32	the	the	DET
cana-2345	116	33	sequential	sequential	ADJ
cana-2345	116	34	processing	processing	NOUN
cana-2345	116	35	through	through	ADP
cana-2345	116	36	layers	layer	NOUN
cana-2345	116	37	from	from	ADP
cana-2345	116	38	input	input	NOUN
cana-2345	116	39	to	to	ADP
cana-2345	116	40	prediction	prediction	NOUN
cana-2345	116	41	.	.	PUNCT
cana-2345	117	1	components	component	NOUN
cana-2345	117	2	within	within	ADP
cana-2345	117	3	each	each	DET
cana-2345	117	4	layer	layer	NOUN
cana-2345	117	5	,	,	PUNCT
cana-2345	117	6	such	such	ADJ
cana-2345	117	7	as	as	ADP
cana-2345	117	8	the	the	DET
cana-2345	117	9	bidirectional	bidirectional	ADJ
cana-2345	117	10	lstm	lstm	PROPN
cana-2345	117	11	,	,	PUNCT
cana-2345	117	12	cnn	cnn	PROPN
cana-2345	117	13	,	,	PUNCT
cana-2345	117	14	attention	attention	NOUN
cana-2345	117	15	,	,	PUNCT
cana-2345	117	16	and	and	CCONJ
cana-2345	117	17	dense	dense	ADJ
cana-2345	117	18	layers	layer	NOUN
cana-2345	117	19	,	,	PUNCT
cana-2345	117	20	are	be	AUX
cana-2345	117	21	arranged	arrange	VERB
cana-2345	117	22	left	left	ADJ
cana-2345	117	23	-	-	PUNCT
cana-2345	117	24	to	to	ADP
cana-2345	117	25	-	-	PUNCT
cana-2345	117	26	right	right	ADJ
cana-2345	117	27	,	,	PUNCT
cana-2345	117	28	highlighting	highlight	VERB
cana-2345	117	29	key	key	ADJ
cana-2345	117	30	data	datum	NOUN
cana-2345	117	31	transformations	transformation	NOUN
cana-2345	117	32	,	,	PUNCT
cana-2345	117	33	conditions	condition	NOUN
cana-2345	117	34	,	,	PUNCT
cana-2345	117	35	and	and	CCONJ
cana-2345	117	36	iterations	iteration	NOUN
cana-2345	117	37	that	that	PRON
cana-2345	117	38	refine	refine	VERB
cana-2345	117	39	text	text	NOUN
cana-2345	117	40	inputs	input	NOUN
cana-2345	117	41	into	into	ADP
cana-2345	117	42	accurate	accurate	ADJ
cana-2345	117	43	sentiment	sentiment	NOUN
cana-2345	117	44	predictions	prediction	NOUN
cana-2345	117	45	.	.	PUNCT
cana-2345	118	1	this	this	DET
cana-2345	118	2	architecture	architecture	NOUN
cana-2345	118	3	integrates	integrate	VERB
cana-2345	118	4	advanced	advanced	ADJ
cana-2345	118	5	layers	layer	NOUN
cana-2345	118	6	designed	design	VERB
cana-2345	118	7	to	to	PART
cana-2345	118	8	capture	capture	VERB
cana-2345	118	9	complex	complex	ADJ
cana-2345	118	10	,	,	PUNCT
cana-2345	118	11	non	non	ADJ
cana-2345	118	12	-	-	ADJ
cana-2345	118	13	linear	linear	ADJ
cana-2345	118	14	relationships	relationship	NOUN
cana-2345	118	15	between	between	ADP
cana-2345	118	16	textual	textual	ADJ
cana-2345	118	17	features	feature	NOUN
cana-2345	118	18	and	and	CCONJ
cana-2345	118	19	sentiments	sentiment	NOUN
cana-2345	118	20	.	.	PUNCT
cana-2345	119	1	each	each	DET
cana-2345	119	2	component	component	NOUN
cana-2345	119	3	is	be	AUX
cana-2345	119	4	optimized	optimize	VERB
cana-2345	119	5	for	for	ADP
cana-2345	119	6	its	its	PRON
cana-2345	119	7	specific	specific	ADJ
cana-2345	119	8	role	role	NOUN
cana-2345	119	9	in	in	ADP
cana-2345	119	10	the	the	DET
cana-2345	119	11	overall	overall	ADJ
cana-2345	119	12	architecture	architecture	NOUN
cana-2345	119	13	.	.	PUNCT
cana-2345	120	1	2.1	2.1	NUM
cana-2345	120	2	input	input	NOUN
cana-2345	120	3	layer	layer	NOUN
cana-2345	120	4	:	:	PUNCT
cana-2345	120	5	text	text	NOUN
cana-2345	120	6	preprocessing	preprocessing	NOUN
cana-2345	120	7	and	and	CCONJ
cana-2345	120	8	embedding	embed	VERB
cana-2345	120	9	the	the	DET
cana-2345	120	10	input	input	NOUN
cana-2345	120	11	layer	layer	NOUN
cana-2345	120	12	transforms	transform	VERB
cana-2345	120	13	raw	raw	ADJ
cana-2345	120	14	text	text	NOUN
cana-2345	120	15	into	into	ADP
cana-2345	120	16	numerical	numerical	ADJ
cana-2345	120	17	vectors	vector	NOUN
cana-2345	120	18	that	that	PRON
cana-2345	120	19	represent	represent	VERB
cana-2345	120	20	semantic	semantic	ADJ
cana-2345	120	21	meanings	meaning	NOUN
cana-2345	120	22	,	,	PUNCT
cana-2345	120	23	serving	serve	VERB
cana-2345	120	24	as	as	ADP
cana-2345	120	25	the	the	DET
cana-2345	120	26	foundation	foundation	NOUN
cana-2345	120	27	for	for	ADP
cana-2345	120	28	further	further	ADJ
cana-2345	120	29	analysis	analysis	NOUN
cana-2345	120	30	.	.	PUNCT
cana-2345	121	1	the	the	DET
cana-2345	121	2	embedding	embed	VERB
cana-2345	121	3	layer	layer	NOUN
cana-2345	121	4	converts	convert	VERB
cana-2345	121	5	words	word	NOUN
cana-2345	121	6	into	into	ADP
cana-2345	121	7	dense	dense	ADJ
cana-2345	121	8	vectors	vector	NOUN
cana-2345	121	9	that	that	PRON
cana-2345	121	10	capture	capture	VERB
cana-2345	121	11	semantic	semantic	ADJ
cana-2345	121	12	and	and	CCONJ
cana-2345	121	13	syntactic	syntactic	ADJ
cana-2345	121	14	nuances	nuance	NOUN
cana-2345	121	15	using	use	VERB
cana-2345	121	16	pre	pre	ADJ
cana-2345	121	17	-	-	ADJ
cana-2345	121	18	trained	train	VERB
cana-2345	121	19	models	model	NOUN
cana-2345	121	20	such	such	ADJ
cana-2345	121	21	as	as	ADP
cana-2345	121	22	bert	bert	PROPN
cana-2345	121	23	(	(	PUNCT
cana-2345	121	24	bidirectional	bidirectional	ADJ
cana-2345	121	25	encoder	encoder	NOUN
cana-2345	121	26	representations	representation	VERB
cana-2345	121	27	from	from	ADP
cana-2345	121	28	transformers	transformer	NOUN
cana-2345	121	29	)	)	PUNCT
cana-2345	121	30	.	.	PUNCT
cana-2345	122	1	bert	bert	PROPN
cana-2345	122	2	embeddings	embedding	NOUN
cana-2345	122	3	provide	provide	VERB
cana-2345	122	4	contextual	contextual	ADJ
cana-2345	122	5	word	word	NOUN
cana-2345	122	6	representations	representation	NOUN
cana-2345	122	7	,	,	PUNCT
cana-2345	122	8	accounting	account	VERB
cana-2345	122	9	for	for	ADP
cana-2345	122	10	the	the	DET
cana-2345	122	11	word	word	NOUN
cana-2345	122	12	’s	’s	PART
cana-2345	122	13	role	role	NOUN
cana-2345	122	14	within	within	ADP
cana-2345	122	15	sentences	sentence	NOUN
cana-2345	122	16	.	.	PUNCT
cana-2345	123	1	let	let	VERB
cana-2345	123	2	𝑇	𝑇	PROPN
cana-2345	123	3	=	=	PROPN
cana-2345	124	1	[	[	X
cana-2345	124	2	𝑤1	𝑤1	VERB
cana-2345	124	3	,	,	PUNCT
cana-2345	124	4	𝑤2	𝑤2	NOUN
cana-2345	124	5	,	,	PUNCT
cana-2345	124	6	.	.	PUNCT
cana-2345	124	7	.	.	PUNCT
cana-2345	124	8	.	.	PUNCT
cana-2345	125	1	,	,	PUNCT
cana-2345	125	2	𝑤𝑛	𝑤𝑛	AUX
cana-2345	125	3	]	]	PUNCT
cana-2345	125	4	be	be	VERB
cana-2345	125	5	the	the	DET
cana-2345	125	6	sequence	sequence	NOUN
cana-2345	125	7	of	of	ADP
cana-2345	125	8	words	word	NOUN
cana-2345	125	9	in	in	ADP
cana-2345	125	10	the	the	DET
cana-2345	125	11	input	input	NOUN
cana-2345	125	12	text	text	NOUN
cana-2345	125	13	.	.	PUNCT
cana-2345	126	1	the	the	DET
cana-2345	126	2	embedding	embed	VERB
cana-2345	126	3	layer	layer	NOUN
cana-2345	126	4	maps	map	NOUN
cana-2345	126	5	each	each	DET
cana-2345	126	6	word	word	NOUN
cana-2345	126	7	𝑤𝑖	𝑤𝑖	ADP
cana-2345	126	8	to	to	ADP
cana-2345	126	9	a	a	DET
cana-2345	126	10	high	high	ADV
cana-2345	126	11	-	-	PUNCT
cana-2345	126	12	dimensional	dimensional	ADJ
cana-2345	126	13	vector	vector	NOUN
cana-2345	126	14	𝑒𝑖	𝑒𝑖	ADP
cana-2345	126	15	∈	∈	PROPN
cana-2345	126	16	ℝ𝑑	ℝ𝑑	PROPN
cana-2345	126	17	,	,	PUNCT
cana-2345	126	18	where	where	SCONJ
cana-2345	126	19	𝑑	𝑑	NOUN
cana-2345	126	20	is	be	AUX
cana-2345	126	21	the	the	DET
cana-2345	126	22	embedding	embed	VERB
cana-2345	126	23	dimension	dimension	NOUN
cana-2345	126	24	:	:	PUNCT
cana-2345	126	25	eq	eq	NOUN
cana-2345	126	26	1	1	NUM
cana-2345	126	27	𝐸	𝐸	NOUN
cana-2345	126	28	=	=	PUNCT
cana-2345	126	29	𝑓(𝑇	𝑓(𝑇	X
cana-2345	126	30	)	)	PUNCT
cana-2345	126	31	=	=	PUNCT
cana-2345	127	1	[	[	X
cana-2345	127	2	𝑒1	𝑒1	NOUN
cana-2345	127	3	,	,	PUNCT
cana-2345	127	4	𝑒2	𝑒2	PROPN
cana-2345	127	5	,	,	PUNCT
cana-2345	127	6	.	.	PUNCT
cana-2345	127	7	.	.	PUNCT
cana-2345	128	1	.	.	PUNCT
cana-2345	129	1	,	,	PUNCT
cana-2345	129	2	𝑒𝑛	𝑒𝑛	X
cana-2345	129	3	]	]	PUNCT
cana-2345	129	4	....	....	PUNCT
cana-2345	130	1	(	(	PUNCT
cana-2345	130	2	eq	eq	NOUN
cana-2345	130	3	1	1	NUM
cana-2345	130	4	)	)	PUNCT
cana-2345	130	5	where	where	SCONJ
cana-2345	130	6	𝑓	𝑓	PROPN
cana-2345	130	7	represents	represent	VERB
cana-2345	130	8	the	the	DET
cana-2345	130	9	embedding	embed	VERB
cana-2345	130	10	function	function	NOUN
cana-2345	130	11	provided	provide	VERB
cana-2345	130	12	by	by	ADP
cana-2345	130	13	bert	bert	PROPN
cana-2345	130	14	.	.	PUNCT
cana-2345	131	1	communications	communication	NOUN
cana-2345	131	2	on	on	ADP
cana-2345	131	3	applied	apply	VERB
cana-2345	131	4	nonlinear	nonlinear	ADJ
cana-2345	131	5	analysis	analysis	NOUN
cana-2345	131	6	issn	issn	NOUN
cana-2345	131	7	:	:	PUNCT
cana-2345	131	8	1074	1074	NUM
cana-2345	131	9	-	-	PUNCT
cana-2345	131	10	133x	133x	NUM
cana-2345	131	11	vol	vol	NOUN
cana-2345	131	12	32	32	NUM
cana-2345	131	13	no	no	NOUN
cana-2345	131	14	.	.	PUNCT
cana-2345	132	1	1s	1s	NUM
cana-2345	132	2	(	(	PUNCT
cana-2345	132	3	2025	2025	NUM
cana-2345	132	4	)	)	PUNCT
cana-2345	132	5	578	578	NUM
cana-2345	132	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2345	132	7	2.2	2.2	NUM
cana-2345	132	8	feature	feature	NOUN
cana-2345	132	9	extraction	extraction	NOUN
cana-2345	132	10	layer	layer	NOUN
cana-2345	132	11	:	:	PUNCT
cana-2345	132	12	bidirectional	bidirectional	ADJ
cana-2345	132	13	lstm	lstm	NOUN
cana-2345	132	14	this	this	DET
cana-2345	132	15	layer	layer	NOUN
cana-2345	132	16	captures	capture	VERB
cana-2345	132	17	long	long	ADJ
cana-2345	132	18	-	-	PUNCT
cana-2345	132	19	term	term	NOUN
cana-2345	132	20	dependencies	dependency	NOUN
cana-2345	132	21	in	in	ADP
cana-2345	132	22	both	both	CCONJ
cana-2345	132	23	forward	forward	ADJ
cana-2345	132	24	and	and	CCONJ
cana-2345	132	25	backward	backward	ADJ
cana-2345	132	26	directions	direction	NOUN
cana-2345	132	27	,	,	PUNCT
cana-2345	132	28	providing	provide	VERB
cana-2345	132	29	a	a	DET
cana-2345	132	30	deep	deep	ADJ
cana-2345	132	31	understanding	understanding	NOUN
cana-2345	132	32	of	of	ADP
cana-2345	132	33	the	the	DET
cana-2345	132	34	context	context	NOUN
cana-2345	132	35	surrounding	surround	VERB
cana-2345	132	36	each	each	DET
cana-2345	132	37	word	word	NOUN
cana-2345	132	38	.	.	PUNCT
cana-2345	133	1	a	a	DET
cana-2345	133	2	bidirectional	bidirectional	ADJ
cana-2345	133	3	lstm	lstm	NOUN
cana-2345	133	4	(	(	PUNCT
cana-2345	133	5	long	long	ADJ
cana-2345	133	6	short	short	ADJ
cana-2345	133	7	-	-	PUNCT
cana-2345	133	8	term	term	NOUN
cana-2345	133	9	memory	memory	NOUN
cana-2345	133	10	)	)	PUNCT
cana-2345	133	11	network	network	NOUN
cana-2345	133	12	is	be	AUX
cana-2345	133	13	used	use	VERB
cana-2345	133	14	to	to	PART
cana-2345	133	15	model	model	VERB
cana-2345	133	16	sequential	sequential	ADJ
cana-2345	133	17	data	datum	NOUN
cana-2345	133	18	,	,	PUNCT
cana-2345	133	19	effectively	effectively	ADV
cana-2345	133	20	managing	manage	VERB
cana-2345	133	21	long	long	ADJ
cana-2345	133	22	-	-	PUNCT
cana-2345	133	23	range	range	NOUN
cana-2345	133	24	dependencies	dependency	NOUN
cana-2345	133	25	in	in	ADP
cana-2345	133	26	text	text	NOUN
cana-2345	133	27	.	.	PUNCT
cana-2345	134	1	this	this	DET
cana-2345	134	2	approach	approach	NOUN
cana-2345	134	3	captures	capture	VERB
cana-2345	134	4	context	context	NOUN
cana-2345	134	5	from	from	ADP
cana-2345	134	6	both	both	CCONJ
cana-2345	134	7	the	the	DET
cana-2345	134	8	past	past	NOUN
cana-2345	134	9	and	and	CCONJ
cana-2345	134	10	future	future	ADJ
cana-2345	134	11	,	,	PUNCT
cana-2345	134	12	enhancing	enhance	VERB
cana-2345	134	13	sentiment	sentiment	NOUN
cana-2345	134	14	prediction	prediction	NOUN
cana-2345	134	15	by	by	ADP
cana-2345	134	16	considering	consider	VERB
cana-2345	134	17	the	the	DET
cana-2345	134	18	full	full	ADJ
cana-2345	134	19	sentence	sentence	NOUN
cana-2345	134	20	structure	structure	NOUN
cana-2345	134	21	.	.	PUNCT
cana-2345	135	1	given	give	VERB
cana-2345	135	2	input	input	NOUN
cana-2345	135	3	sequence	sequence	NOUN
cana-2345	135	4	𝐸	𝐸	NOUN
cana-2345	135	5	=	=	PUNCT
cana-2345	136	1	[	[	X
cana-2345	136	2	𝑒1	𝑒1	NOUN
cana-2345	136	3	,	,	PUNCT
cana-2345	136	4	𝑒2	𝑒2	PROPN
cana-2345	136	5	,	,	PUNCT
cana-2345	136	6	.	.	PUNCT
cana-2345	136	7	.	.	PUNCT
cana-2345	136	8	.	.	PUNCT
cana-2345	137	1	,	,	PUNCT
cana-2345	137	2	𝑒𝑛	𝑒𝑛	X
cana-2345	137	3	]	]	X
cana-2345	137	4	,	,	PUNCT
cana-2345	137	5	the	the	DET
cana-2345	137	6	bilstm	bilstm	NOUN
cana-2345	137	7	generates	generate	VERB
cana-2345	137	8	hidden	hide	VERB
cana-2345	137	9	states	state	NOUN
cana-2345	137	10	𝐻	𝐻	NOUN
cana-2345	137	11	=	=	PUNCT
cana-2345	138	1	[	[	X
cana-2345	138	2	ℎ1	ℎ1	PROPN
cana-2345	138	3	,	,	PUNCT
cana-2345	138	4	ℎ2	ℎ2	NOUN
cana-2345	138	5	,	,	PUNCT
cana-2345	138	6	.	.	PUNCT
cana-2345	138	7	.	.	PUNCT
cana-2345	139	1	.	.	PUNCT
cana-2345	140	1	,	,	PUNCT
cana-2345	140	2	ℎ𝑛	ℎ𝑛	NOUN
cana-2345	140	3	]	]	X
cana-2345	140	4	where	where	SCONJ
cana-2345	140	5	each	each	DET
cana-2345	140	6	ℎ𝑖	ℎ𝑖	NOUN
cana-2345	140	7	is	be	AUX
cana-2345	140	8	the	the	DET
cana-2345	140	9	concatenation	concatenation	NOUN
cana-2345	140	10	of	of	ADP
cana-2345	140	11	forward	forward	ADV
cana-2345	140	12	and	and	CCONJ
cana-2345	140	13	backward	backward	ADJ
cana-2345	140	14	lstm	lstm	ADJ
cana-2345	140	15	outputs	output	NOUN
cana-2345	140	16	:	:	PUNCT
cana-2345	140	17	eq	eq	NOUN
cana-2345	140	18	2	2	NUM
cana-2345	140	19	ℎ𝑖	ℎ𝑖	NOUN
cana-2345	140	20	=	=	SYM
cana-2345	140	21	𝐿𝑆𝑇𝑀	𝐿𝑆𝑇𝑀	PROPN
cana-2345	140	22	⃖	⃖	PROPN
cana-2345	140	23	(	(	PUNCT
cana-2345	140	24	𝑒1	𝑒1	NOUN
cana-2345	140	25	,	,	PUNCT
cana-2345	140	26	.	.	PUNCT
cana-2345	140	27	.	.	PUNCT
cana-2345	140	28	.	.	PUNCT
cana-2345	141	1	,	,	PUNCT
cana-2345	141	2	𝑒𝑖	𝑒𝑖	X
cana-2345	141	3	)	)	PUNCT
cana-2345	141	4	⊕	⊕	PROPN
cana-2345	141	5	𝐿𝑆𝑇𝑀	𝐿𝑆𝑇𝑀	PROPN
cana-2345	141	6	(	(	PUNCT
cana-2345	141	7	𝑒𝑛	𝑒𝑛	NOUN
cana-2345	141	8	,	,	PUNCT
cana-2345	141	9	.	.	PUNCT
cana-2345	141	10	.	.	PUNCT
cana-2345	141	11	.	.	PUNCT
cana-2345	142	1	,	,	PUNCT
cana-2345	142	2	𝑒𝑖	𝑒𝑖	X
cana-2345	142	3	)	)	PUNCT
cana-2345	142	4	....	....	PUNCT
cana-2345	143	1	(	(	PUNCT
cana-2345	143	2	eq	eq	NOUN
cana-2345	143	3	2	2	NUM
cana-2345	143	4	)	)	PUNCT
cana-2345	143	5	2.3	2.3	NUM
cana-2345	143	6	feature	feature	NOUN
cana-2345	143	7	extraction	extraction	NOUN
cana-2345	143	8	layer	layer	NOUN
cana-2345	143	9	:	:	PUNCT
cana-2345	143	10	convolutional	convolutional	ADJ
cana-2345	143	11	neural	neural	ADJ
cana-2345	143	12	network	network	NOUN
cana-2345	143	13	(	(	PUNCT
cana-2345	143	14	cnn	cnn	PROPN
cana-2345	143	15	)	)	PUNCT
cana-2345	143	16	cnn	cnn	PROPN
cana-2345	143	17	layers	layer	NOUN
cana-2345	143	18	identify	identify	VERB
cana-2345	143	19	key	key	ADJ
cana-2345	143	20	phrases	phrase	NOUN
cana-2345	143	21	and	and	CCONJ
cana-2345	143	22	local	local	ADJ
cana-2345	143	23	features	feature	NOUN
cana-2345	143	24	crucial	crucial	ADJ
cana-2345	143	25	for	for	ADP
cana-2345	143	26	sentiment	sentiment	NOUN
cana-2345	143	27	analysis	analysis	NOUN
cana-2345	143	28	by	by	ADP
cana-2345	143	29	scanning	scan	VERB
cana-2345	143	30	through	through	ADP
cana-2345	143	31	the	the	DET
cana-2345	143	32	input	input	NOUN
cana-2345	143	33	with	with	ADP
cana-2345	143	34	filters	filter	NOUN
cana-2345	143	35	.	.	PUNCT
cana-2345	144	1	multiple	multiple	ADJ
cana-2345	144	2	convolutional	convolutional	ADJ
cana-2345	144	3	filters	filter	NOUN
cana-2345	144	4	of	of	ADP
cana-2345	144	5	varying	vary	VERB
cana-2345	144	6	sizes	size	NOUN
cana-2345	144	7	(	(	PUNCT
cana-2345	144	8	e.g.	e.g.	ADV
cana-2345	144	9	,	,	PUNCT
cana-2345	144	10	2	2	NUM
cana-2345	144	11	,	,	PUNCT
cana-2345	144	12	3	3	NUM
cana-2345	144	13	,	,	PUNCT
cana-2345	144	14	and	and	CCONJ
cana-2345	144	15	4	4	X
cana-2345	144	16	)	)	PUNCT
cana-2345	144	17	slide	slide	NOUN
cana-2345	144	18	over	over	ADP
cana-2345	144	19	the	the	DET
cana-2345	144	20	input	input	NOUN
cana-2345	144	21	sequence	sequence	NOUN
cana-2345	144	22	,	,	PUNCT
cana-2345	144	23	capturing	capture	VERB
cana-2345	144	24	local	local	ADJ
cana-2345	144	25	dependencies	dependency	NOUN
cana-2345	144	26	and	and	CCONJ
cana-2345	144	27	significant	significant	ADJ
cana-2345	144	28	patterns	pattern	NOUN
cana-2345	144	29	that	that	PRON
cana-2345	144	30	may	may	AUX
cana-2345	144	31	indicate	indicate	VERB
cana-2345	144	32	sentiment	sentiment	NOUN
cana-2345	144	33	.	.	PUNCT
cana-2345	145	1	max	max	PROPN
cana-2345	145	2	pooling	pooling	PROPN
cana-2345	145	3	follows	follow	VERB
cana-2345	145	4	to	to	PART
cana-2345	145	5	retain	retain	VERB
cana-2345	145	6	the	the	DET
cana-2345	145	7	most	most	ADV
cana-2345	145	8	prominent	prominent	ADJ
cana-2345	145	9	features	feature	NOUN
cana-2345	145	10	,	,	PUNCT
cana-2345	145	11	compressing	compress	VERB
cana-2345	145	12	the	the	DET
cana-2345	145	13	representation	representation	NOUN
cana-2345	145	14	.	.	PUNCT
cana-2345	146	1	for	for	ADP
cana-2345	146	2	a	a	DET
cana-2345	146	3	filter	filter	NOUN
cana-2345	146	4	𝑊	𝑊	NOUN
cana-2345	146	5	of	of	ADP
cana-2345	146	6	size	size	NOUN
cana-2345	146	7	𝑘	𝑘	PROPN
cana-2345	146	8	,	,	PUNCT
cana-2345	146	9	the	the	DET
cana-2345	146	10	convolution	convolution	NOUN
cana-2345	146	11	operation	operation	NOUN
cana-2345	146	12	generates	generate	VERB
cana-2345	146	13	feature	feature	NOUN
cana-2345	146	14	map	map	NOUN
cana-2345	146	15	𝑐𝑖	𝑐𝑖	NOUN
cana-2345	146	16	as	as	ADP
cana-2345	146	17	:	:	PUNCT
cana-2345	146	18	eq	eq	NOUN
cana-2345	146	19	3	3	NUM
cana-2345	146	20	𝑐𝑖	𝑐𝑖	NOUN
cana-2345	146	21	=	=	SYM
cana-2345	146	22	𝑅𝑒𝐿𝑈(𝑊	𝑅𝑒𝐿𝑈(𝑊	PROPN
cana-2345	146	23	∗	∗	NOUN
cana-2345	146	24	ℎ𝑖:𝑖+𝑘−1	ℎ𝑖:𝑖+𝑘−1	VERB
cana-2345	147	1	+	+	CCONJ
cana-2345	147	2	𝑏	𝑏	NOUN
cana-2345	147	3	)	)	PUNCT
cana-2345	147	4	....	....	PUNCT
cana-2345	148	1	(	(	PUNCT
cana-2345	148	2	eq	eq	NOUN
cana-2345	148	3	3	3	NUM
cana-2345	148	4	)	)	PUNCT
cana-2345	148	5	where	where	SCONJ
cana-2345	148	6	𝑏	𝑏	NOUN
cana-2345	148	7	is	be	AUX
cana-2345	148	8	a	a	DET
cana-2345	148	9	bias	bias	NOUN
cana-2345	148	10	term	term	NOUN
cana-2345	148	11	,	,	PUNCT
cana-2345	148	12	and	and	CCONJ
cana-2345	148	13	∗	∗	NOUN
cana-2345	148	14	denotes	denote	NOUN
cana-2345	148	15	convolution	convolution	NOUN
cana-2345	148	16	.	.	PUNCT
cana-2345	149	1	1	1	X
cana-2345	149	2	.	.	X
cana-2345	149	3	initialize	initialize	VERB
cana-2345	149	4	multiple	multiple	ADJ
cana-2345	149	5	filters	filter	NOUN
cana-2345	149	6	of	of	ADP
cana-2345	149	7	different	different	ADJ
cana-2345	149	8	sizes	size	NOUN
cana-2345	149	9	.	.	PUNCT
cana-2345	150	1	2	2	X
cana-2345	150	2	.	.	X
cana-2345	150	3	convolve	convolve	NOUN
cana-2345	150	4	each	each	DET
cana-2345	150	5	filter	filter	NOUN
cana-2345	150	6	over	over	ADP
cana-2345	150	7	the	the	DET
cana-2345	150	8	bilstm	bilstm	NOUN
cana-2345	150	9	output	output	NOUN
cana-2345	150	10	.	.	PUNCT
cana-2345	151	1	3	3	X
cana-2345	151	2	.	.	X
cana-2345	151	3	apply	apply	VERB
cana-2345	151	4	relu	relu	NOUN
cana-2345	151	5	activation	activation	NOUN
cana-2345	151	6	.	.	PUNCT
cana-2345	152	1	4	4	X
cana-2345	152	2	.	.	X
cana-2345	152	3	perform	perform	VERB
cana-2345	152	4	max	max	PROPN
cana-2345	152	5	pooling	pool	VERB
cana-2345	152	6	over	over	ADP
cana-2345	152	7	each	each	DET
cana-2345	152	8	feature	feature	NOUN
cana-2345	152	9	map	map	NOUN
cana-2345	152	10	.	.	PUNCT
cana-2345	153	1	5	5	X
cana-2345	153	2	.	.	X
cana-2345	153	3	concatenate	concatenate	NOUN
cana-2345	153	4	pooled	pool	VERB
cana-2345	153	5	features	feature	NOUN
cana-2345	153	6	to	to	PART
cana-2345	153	7	form	form	VERB
cana-2345	153	8	the	the	DET
cana-2345	153	9	final	final	ADJ
cana-2345	153	10	output	output	NOUN
cana-2345	153	11	vector	vector	NOUN
cana-2345	153	12	.	.	PUNCT
cana-2345	154	1	2.4	2.4	NUM
cana-2345	154	2	attention	attention	NOUN
cana-2345	154	3	mechanism	mechanism	NOUN
cana-2345	154	4	layer	layer	NOUN
cana-2345	154	5	the	the	DET
cana-2345	154	6	attention	attention	NOUN
cana-2345	154	7	layer	layer	NOUN
cana-2345	154	8	selectively	selectively	ADV
cana-2345	154	9	focuses	focus	VERB
cana-2345	154	10	on	on	ADP
cana-2345	154	11	significant	significant	ADJ
cana-2345	154	12	words	word	NOUN
cana-2345	154	13	,	,	PUNCT
cana-2345	154	14	enhancing	enhance	VERB
cana-2345	154	15	the	the	DET
cana-2345	154	16	model	model	NOUN
cana-2345	154	17	’s	’s	PART
cana-2345	154	18	interpretability	interpretability	NOUN
cana-2345	154	19	and	and	CCONJ
cana-2345	154	20	accuracy	accuracy	NOUN
cana-2345	154	21	.	.	PUNCT
cana-2345	155	1	attention	attention	NOUN
cana-2345	155	2	scores	score	NOUN
cana-2345	155	3	are	be	AUX
cana-2345	155	4	calculated	calculate	VERB
cana-2345	155	5	for	for	ADP
cana-2345	155	6	each	each	DET
cana-2345	155	7	word	word	NOUN
cana-2345	155	8	,	,	PUNCT
cana-2345	155	9	amplifying	amplify	VERB
cana-2345	155	10	those	those	PRON
cana-2345	155	11	contributing	contribute	VERB
cana-2345	155	12	most	most	ADJ
cana-2345	155	13	to	to	ADP
cana-2345	155	14	the	the	DET
cana-2345	155	15	sentiment	sentiment	NOUN
cana-2345	155	16	.	.	PUNCT
cana-2345	156	1	this	this	DET
cana-2345	156	2	dynamic	dynamic	ADJ
cana-2345	156	3	weighting	weighting	NOUN
cana-2345	156	4	improves	improve	VERB
cana-2345	156	5	the	the	DET
cana-2345	156	6	model	model	NOUN
cana-2345	156	7	’s	’s	PART
cana-2345	156	8	focus	focus	NOUN
cana-2345	156	9	on	on	ADP
cana-2345	156	10	sentiment	sentiment	NOUN
cana-2345	156	11	-	-	PUNCT
cana-2345	156	12	relevant	relevant	ADJ
cana-2345	156	13	text	text	NOUN
cana-2345	156	14	portions	portion	NOUN
cana-2345	156	15	.	.	PUNCT
cana-2345	157	1	given	give	VERB
cana-2345	157	2	hidden	hide	VERB
cana-2345	157	3	states	state	NOUN
cana-2345	157	4	𝐻	𝐻	PROPN
cana-2345	157	5	,	,	PUNCT
cana-2345	157	6	attention	attention	NOUN
cana-2345	157	7	weights	weight	NOUN
cana-2345	157	8	𝛼𝑖	𝛼𝑖	PROPN
cana-2345	157	9	are	be	AUX
cana-2345	157	10	computed	compute	VERB
cana-2345	157	11	:	:	PUNCT
cana-2345	157	12	eq	eq	NOUN
cana-2345	157	13	4	4	NUM
cana-2345	157	14	𝛼𝑖	𝛼𝑖	PROPN
cana-2345	157	15	=	=	SYM
cana-2345	157	16	𝑒𝑥𝑝(ℎ𝑖	𝑒𝑥𝑝(ℎ𝑖	PROPN
cana-2345	157	17	𝑇⋅𝑊𝑎	𝑇⋅𝑊𝑎	PROPN
cana-2345	157	18	)	)	PUNCT
cana-2345	157	19	∑𝑛	∑𝑛	PROPN
cana-2345	157	20	𝑗=1	𝑗=1	PROPN
cana-2345	157	21	𝑒𝑥𝑝(ℎ𝑗	𝑒𝑥𝑝(ℎ𝑗	VERB
cana-2345	157	22	𝑇⋅𝑊𝑎	𝑇⋅𝑊𝑎	PROPN
cana-2345	157	23	)	)	PUNCT
cana-2345	157	24	....	....	PUNCT
cana-2345	158	1	(	(	PUNCT
cana-2345	158	2	eq	eq	NOUN
cana-2345	158	3	4	4	NUM
cana-2345	158	4	)	)	PUNCT
cana-2345	158	5	where	where	SCONJ
cana-2345	158	6	𝑊𝑎	𝑊𝑎	PROPN
cana-2345	158	7	is	be	AUX
cana-2345	158	8	the	the	DET
cana-2345	158	9	attention	attention	NOUN
cana-2345	158	10	weight	weight	NOUN
cana-2345	158	11	matrix	matrix	NOUN
cana-2345	158	12	.	.	PUNCT
cana-2345	159	1	the	the	DET
cana-2345	159	2	context	context	PROPN
cana-2345	159	3	vector	vector	NOUN
cana-2345	159	4	𝐶	𝐶	PROPN
cana-2345	159	5	is	be	AUX
cana-2345	159	6	then	then	ADV
cana-2345	159	7	:	:	PUNCT
cana-2345	159	8	eq	eq	NOUN
cana-2345	159	9	5	5	NUM
cana-2345	159	10	𝐶	𝐶	PROPN
cana-2345	159	11	=	=	SYM
cana-2345	159	12	∑𝑛	∑𝑛	PROPN
cana-2345	159	13	𝑖=1	𝑖=1	PROPN
cana-2345	159	14	𝛼𝑖ℎ𝑖	𝛼𝑖ℎ𝑖	NOUN
cana-2345	159	15	....	....	PUNCT
cana-2345	160	1	(	(	PUNCT
cana-2345	160	2	eq	eq	NOUN
cana-2345	160	3	5	5	NUM
cana-2345	160	4	)	)	PUNCT
cana-2345	160	5	communications	communication	NOUN
cana-2345	160	6	on	on	ADP
cana-2345	160	7	applied	apply	VERB
cana-2345	160	8	nonlinear	nonlinear	ADJ
cana-2345	160	9	analysis	analysis	NOUN
cana-2345	160	10	issn	issn	NOUN
cana-2345	160	11	:	:	PUNCT
cana-2345	160	12	1074	1074	NUM
cana-2345	160	13	-	-	PUNCT
cana-2345	160	14	133x	133x	NUM
cana-2345	160	15	vol	vol	NOUN
cana-2345	160	16	32	32	NUM
cana-2345	160	17	no	no	NOUN
cana-2345	160	18	.	.	PUNCT
cana-2345	161	1	1s	1s	NUM
cana-2345	161	2	(	(	PUNCT
cana-2345	161	3	2025	2025	NUM
cana-2345	161	4	)	)	PUNCT
cana-2345	161	5	579	579	NUM
cana-2345	161	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2345	161	7	2.5	2.5	NUM
cana-2345	161	8	nonlinear	nonlinear	ADJ
cana-2345	161	9	feature	feature	NOUN
cana-2345	161	10	fusion	fusion	NOUN
cana-2345	161	11	layer	layer	NOUN
cana-2345	161	12	this	this	DET
cana-2345	161	13	layer	layer	NOUN
cana-2345	161	14	combines	combine	VERB
cana-2345	161	15	all	all	PRON
cana-2345	161	16	extracted	extract	VERB
cana-2345	161	17	features	feature	NOUN
cana-2345	161	18	using	use	VERB
cana-2345	161	19	dense	dense	ADJ
cana-2345	161	20	neural	neural	ADJ
cana-2345	161	21	networks	network	NOUN
cana-2345	161	22	,	,	PUNCT
cana-2345	161	23	applying	apply	VERB
cana-2345	161	24	nonlinear	nonlinear	ADJ
cana-2345	161	25	activation	activation	NOUN
cana-2345	161	26	functions	function	NOUN
cana-2345	161	27	to	to	PART
cana-2345	161	28	learn	learn	VERB
cana-2345	161	29	complex	complex	ADJ
cana-2345	161	30	sentiment	sentiment	NOUN
cana-2345	161	31	patterns	pattern	NOUN
cana-2345	161	32	.	.	PUNCT
cana-2345	162	1	dense	dense	ADJ
cana-2345	162	2	layers	layer	NOUN
cana-2345	162	3	with	with	ADP
cana-2345	162	4	relu	relu	NOUN
cana-2345	162	5	activation	activation	NOUN
cana-2345	162	6	merge	merge	VERB
cana-2345	162	7	the	the	DET
cana-2345	162	8	cnn	cnn	NOUN
cana-2345	162	9	and	and	CCONJ
cana-2345	162	10	bilstm	bilstm	NOUN
cana-2345	162	11	outputs	output	NOUN
cana-2345	162	12	,	,	PUNCT
cana-2345	162	13	capturing	capture	VERB
cana-2345	162	14	nonlinear	nonlinear	ADJ
cana-2345	162	15	dependencies	dependency	NOUN
cana-2345	162	16	between	between	ADP
cana-2345	162	17	extracted	extract	VERB
cana-2345	162	18	features	feature	NOUN
cana-2345	162	19	.	.	PUNCT
cana-2345	163	1	dropout	dropout	NOUN
cana-2345	163	2	regularization	regularization	NOUN
cana-2345	163	3	ensures	ensure	VERB
cana-2345	163	4	the	the	DET
cana-2345	163	5	model	model	NOUN
cana-2345	163	6	does	do	AUX
cana-2345	163	7	not	not	PART
cana-2345	163	8	overfit	overfit	VERB
cana-2345	163	9	.	.	PUNCT
cana-2345	164	1	for	for	ADP
cana-2345	164	2	dense	dense	ADJ
cana-2345	164	3	layer	layer	NOUN
cana-2345	164	4	input	input	NOUN
cana-2345	164	5	𝑥	𝑥	PROPN
cana-2345	164	6	,	,	PUNCT
cana-2345	164	7	the	the	DET
cana-2345	164	8	output	output	NOUN
cana-2345	164	9	is	be	AUX
cana-2345	164	10	:	:	PUNCT
cana-2345	164	11	eq	eq	ADP
cana-2345	164	12	6	6	NUM
cana-2345	164	13	𝑦	𝑦	NOUN
cana-2345	164	14	=	=	PUNCT
cana-2345	164	15	𝑅𝑒𝐿𝑈(𝑊𝑑	𝑅𝑒𝐿𝑈(𝑊𝑑	PUNCT
cana-2345	164	16	⋅	⋅	X
cana-2345	164	17	𝑥	𝑥	PROPN
cana-2345	164	18	+	+	NUM
cana-2345	164	19	𝑏𝑑	𝑏𝑑	PROPN
cana-2345	164	20	)	)	PUNCT
cana-2345	164	21	....	....	PUNCT
cana-2345	165	1	(	(	PUNCT
cana-2345	165	2	eq	eq	NOUN
cana-2345	165	3	6	6	NUM
cana-2345	165	4	)	)	PUNCT
cana-2345	165	5	where	where	SCONJ
cana-2345	165	6	𝑊𝑑	𝑊𝑑	PROPN
cana-2345	165	7	and	and	CCONJ
cana-2345	165	8	𝑏𝑑	𝑏𝑑	PROPN
cana-2345	165	9	are	be	AUX
cana-2345	165	10	the	the	DET
cana-2345	165	11	weight	weight	NOUN
cana-2345	165	12	matrix	matrix	NOUN
cana-2345	165	13	and	and	CCONJ
cana-2345	165	14	bias	bias	NOUN
cana-2345	165	15	vector	vector	NOUN
cana-2345	165	16	,	,	PUNCT
cana-2345	165	17	respectively	respectively	ADV
cana-2345	165	18	.	.	PUNCT
cana-2345	166	1	2.6	2.6	NUM
cana-2345	166	2	sentiment	sentiment	NOUN
cana-2345	166	3	prediction	prediction	NOUN
cana-2345	166	4	layer	layer	NOUN
cana-2345	166	5	the	the	DET
cana-2345	166	6	final	final	ADJ
cana-2345	166	7	layer	layer	NOUN
cana-2345	166	8	produces	produce	VERB
cana-2345	166	9	the	the	DET
cana-2345	166	10	sentiment	sentiment	NOUN
cana-2345	166	11	classification	classification	NOUN
cana-2345	166	12	,	,	PUNCT
cana-2345	166	13	providing	provide	VERB
cana-2345	166	14	a	a	DET
cana-2345	166	15	probability	probability	NOUN
cana-2345	166	16	distribution	distribution	NOUN
cana-2345	166	17	over	over	ADP
cana-2345	166	18	sentiment	sentiment	NOUN
cana-2345	166	19	classes	class	NOUN
cana-2345	166	20	.	.	PUNCT
cana-2345	167	1	a	a	DET
cana-2345	167	2	fully	fully	ADV
cana-2345	167	3	connected	connect	VERB
cana-2345	167	4	layer	layer	NOUN
cana-2345	167	5	aggregates	aggregate	VERB
cana-2345	167	6	the	the	DET
cana-2345	167	7	feature	feature	NOUN
cana-2345	167	8	fusion	fusion	NOUN
cana-2345	167	9	layer	layer	NOUN
cana-2345	167	10	’s	’s	PART
cana-2345	167	11	outputs	output	NOUN
cana-2345	167	12	,	,	PUNCT
cana-2345	167	13	feeding	feed	VERB
cana-2345	167	14	into	into	ADP
cana-2345	167	15	a	a	DET
cana-2345	167	16	softmax	softmax	NOUN
cana-2345	167	17	activation	activation	NOUN
cana-2345	167	18	for	for	ADP
cana-2345	167	19	multi	multi	ADJ
cana-2345	167	20	-	-	ADJ
cana-2345	167	21	class	class	ADJ
cana-2345	167	22	classification	classification	NOUN
cana-2345	167	23	or	or	CCONJ
cana-2345	167	24	sigmoid	sigmoid	NOUN
cana-2345	167	25	for	for	ADP
cana-2345	167	26	binary	binary	ADJ
cana-2345	167	27	classification	classification	NOUN
cana-2345	167	28	.	.	PUNCT
cana-2345	168	1	for	for	ADP
cana-2345	168	2	multi	multi	ADJ
cana-2345	168	3	-	-	ADJ
cana-2345	168	4	class	class	ADJ
cana-2345	168	5	classification	classification	NOUN
cana-2345	168	6	,	,	PUNCT
cana-2345	168	7	the	the	DET
cana-2345	168	8	softmax	softmax	NOUN
cana-2345	168	9	function	function	NOUN
cana-2345	168	10	outputs	output	VERB
cana-2345	168	11	probabilities	probability	NOUN
cana-2345	168	12	:	:	PUNCT
cana-2345	168	13	eq	eq	NOUN
cana-2345	168	14	7	7	NUM
cana-2345	168	15	𝑃(𝑦𝑖	𝑃(𝑦𝑖	ADV
cana-2345	168	16	)	)	PUNCT
cana-2345	168	17	=	=	SYM
cana-2345	168	18	𝑒𝑥𝑝(𝑧𝑖	𝑒𝑥𝑝(𝑧𝑖	NOUN
cana-2345	168	19	)	)	PUNCT
cana-2345	168	20	∑𝐾	∑𝐾	PROPN
cana-2345	168	21	𝑗=1	𝑗=1	PROPN
cana-2345	168	22	𝑒𝑥𝑝(𝑧𝑗	𝑒𝑥𝑝(𝑧𝑗	PROPN
cana-2345	168	23	)	)	PUNCT
cana-2345	168	24	....	....	PUNCT
cana-2345	169	1	(	(	PUNCT
cana-2345	169	2	eq	eq	NOUN
cana-2345	169	3	7	7	NUM
cana-2345	169	4	)	)	PUNCT
cana-2345	169	5	where	where	SCONJ
cana-2345	169	6	𝑧	𝑧	PROPN
cana-2345	169	7	is	be	AUX
cana-2345	169	8	the	the	DET
cana-2345	169	9	input	input	NOUN
cana-2345	169	10	to	to	ADP
cana-2345	169	11	the	the	DET
cana-2345	169	12	softmax	softmax	NOUN
cana-2345	169	13	layer	layer	NOUN
cana-2345	169	14	and	and	CCONJ
cana-2345	169	15	𝐾	𝐾	PROPN
cana-2345	169	16	is	be	AUX
cana-2345	169	17	the	the	DET
cana-2345	169	18	number	number	NOUN
cana-2345	169	19	of	of	ADP
cana-2345	169	20	classes	class	NOUN
cana-2345	169	21	.	.	PUNCT
cana-2345	170	1	3	3	NUM
cana-2345	170	2	experimental	experimental	ADJ
cana-2345	170	3	study	study	NOUN
cana-2345	170	4	the	the	DET
cana-2345	170	5	experimental	experimental	ADJ
cana-2345	170	6	study	study	NOUN
cana-2345	170	7	aims	aim	VERB
cana-2345	170	8	to	to	PART
cana-2345	170	9	evaluate	evaluate	VERB
cana-2345	170	10	the	the	DET
cana-2345	170	11	performance	performance	NOUN
cana-2345	170	12	of	of	ADP
cana-2345	170	13	the	the	DET
cana-2345	170	14	proposed	propose	VERB
cana-2345	170	15	npsc	npsc	NOUN
cana-2345	170	16	(	(	PUNCT
cana-2345	170	17	nonlinear	nonlinear	ADJ
cana-2345	170	18	precise	precise	ADJ
cana-2345	170	19	sentiment	sentiment	NOUN
cana-2345	170	20	classification	classification	NOUN
cana-2345	170	21	)	)	PUNCT
cana-2345	170	22	framework	framework	NOUN
cana-2345	170	23	against	against	ADP
cana-2345	170	24	contemporary	contemporary	ADJ
cana-2345	170	25	sentiment	sentiment	NOUN
cana-2345	170	26	classification	classification	NOUN
cana-2345	170	27	models	model	NOUN
cana-2345	170	28	using	use	VERB
cana-2345	170	29	the	the	DET
cana-2345	170	30	imdb	imdb	NOUN
cana-2345	170	31	movie	movie	NOUN
cana-2345	170	32	review	review	NOUN
cana-2345	170	33	dataset	dataset	VERB
cana-2345	170	34	.	.	PUNCT
cana-2345	171	1	this	this	DET
cana-2345	171	2	section	section	NOUN
cana-2345	171	3	details	detail	VERB
cana-2345	171	4	the	the	DET
cana-2345	171	5	experimental	experimental	ADJ
cana-2345	171	6	setup	setup	NOUN
cana-2345	171	7	,	,	PUNCT
cana-2345	171	8	including	include	VERB
cana-2345	171	9	dataset	dataset	VERB
cana-2345	171	10	descriptions	description	NOUN
cana-2345	171	11	,	,	PUNCT
cana-2345	171	12	model	model	NOUN
cana-2345	171	13	configurations	configuration	NOUN
cana-2345	171	14	,	,	PUNCT
cana-2345	171	15	evaluation	evaluation	NOUN
cana-2345	171	16	metrics	metric	NOUN
cana-2345	171	17	,	,	PUNCT
cana-2345	171	18	and	and	CCONJ
cana-2345	171	19	a	a	DET
cana-2345	171	20	comparative	comparative	ADJ
cana-2345	171	21	analysis	analysis	NOUN
cana-2345	171	22	of	of	ADP
cana-2345	171	23	results	result	NOUN
cana-2345	171	24	.	.	PUNCT
cana-2345	172	1	dataset	dataset	NOUN
cana-2345	172	2	:	:	PUNCT
cana-2345	172	3	the	the	DET
cana-2345	172	4	imdb	imdb	PROPN
cana-2345	172	5	movie	movie	NOUN
cana-2345	172	6	review	review	NOUN
cana-2345	172	7	dataset	dataset	NOUN
cana-2345	172	8	,	,	PUNCT
cana-2345	172	9	sourced	source	VERB
cana-2345	172	10	from	from	ADP
cana-2345	172	11	kaggle	kaggle	PROPN
cana-2345	172	12	,	,	PUNCT
cana-2345	172	13	was	be	AUX
cana-2345	172	14	used	use	VERB
cana-2345	172	15	for	for	ADP
cana-2345	172	16	the	the	DET
cana-2345	172	17	experiments	experiment	NOUN
cana-2345	172	18	.	.	PUNCT
cana-2345	173	1	this	this	DET
cana-2345	173	2	dataset	dataset	NOUN
cana-2345	173	3	contains	contain	VERB
cana-2345	173	4	50,000	50,000	NUM
cana-2345	173	5	movie	movie	NOUN
cana-2345	173	6	reviews	review	NOUN
cana-2345	173	7	split	split	VERB
cana-2345	173	8	evenly	evenly	ADV
cana-2345	173	9	between	between	ADP
cana-2345	173	10	positive	positive	ADJ
cana-2345	173	11	and	and	CCONJ
cana-2345	173	12	negative	negative	ADJ
cana-2345	173	13	sentiments	sentiment	NOUN
cana-2345	173	14	,	,	PUNCT
cana-2345	173	15	providing	provide	VERB
cana-2345	173	16	a	a	DET
cana-2345	173	17	balanced	balanced	ADJ
cana-2345	173	18	classification	classification	NOUN
cana-2345	173	19	task	task	NOUN
cana-2345	173	20	.	.	PUNCT
cana-2345	174	1	the	the	DET
cana-2345	174	2	reviews	review	NOUN
cana-2345	174	3	vary	vary	VERB
cana-2345	174	4	in	in	ADP
cana-2345	174	5	length	length	NOUN
cana-2345	174	6	and	and	CCONJ
cana-2345	174	7	complexity	complexity	NOUN
cana-2345	174	8	,	,	PUNCT
cana-2345	174	9	making	make	VERB
cana-2345	174	10	the	the	DET
cana-2345	174	11	dataset	dataset	NOUN
cana-2345	174	12	suitable	suitable	ADJ
cana-2345	174	13	for	for	ADP
cana-2345	174	14	testing	test	VERB
cana-2345	174	15	the	the	DET
cana-2345	174	16	model	model	NOUN
cana-2345	174	17	's	's	PART
cana-2345	174	18	ability	ability	NOUN
cana-2345	174	19	to	to	PART
cana-2345	174	20	handle	handle	VERB
cana-2345	174	21	diverse	diverse	ADJ
cana-2345	174	22	and	and	CCONJ
cana-2345	174	23	nuanced	nuanced	ADJ
cana-2345	174	24	textual	textual	ADJ
cana-2345	174	25	data	datum	NOUN
cana-2345	174	26	.	.	PUNCT
cana-2345	175	1	each	each	PRON
cana-2345	175	2	review	review	VERB
cana-2345	175	3	underwent	underwent	NOUN
cana-2345	175	4	preprocessing	preprocessing	NOUN
cana-2345	175	5	,	,	PUNCT
cana-2345	175	6	including	include	VERB
cana-2345	175	7	text	text	NOUN
cana-2345	175	8	cleaning	cleaning	NOUN
cana-2345	175	9	,	,	PUNCT
cana-2345	175	10	tokenization	tokenization	NOUN
cana-2345	175	11	,	,	PUNCT
cana-2345	175	12	removal	removal	NOUN
cana-2345	175	13	of	of	ADP
cana-2345	175	14	stop	stop	NOUN
cana-2345	175	15	words	word	NOUN
cana-2345	175	16	,	,	PUNCT
cana-2345	175	17	and	and	CCONJ
cana-2345	175	18	conversion	conversion	NOUN
cana-2345	175	19	into	into	ADP
cana-2345	175	20	bert	bert	PROPN
cana-2345	175	21	embeddings	embedding	NOUN
cana-2345	175	22	,	,	PUNCT
cana-2345	175	23	which	which	PRON
cana-2345	175	24	served	serve	VERB
cana-2345	175	25	as	as	ADP
cana-2345	175	26	inputs	input	NOUN
cana-2345	175	27	to	to	ADP
cana-2345	175	28	the	the	DET
cana-2345	175	29	npsc	npsc	PROPN
cana-2345	175	30	model	model	PROPN
cana-2345	175	31	.	.	PUNCT
cana-2345	176	1	model	model	NOUN
cana-2345	176	2	configuration	configuration	NOUN
cana-2345	176	3	and	and	CCONJ
cana-2345	176	4	training	training	NOUN
cana-2345	176	5	:	:	PUNCT
cana-2345	176	6	the	the	DET
cana-2345	176	7	npsc	npsc	PROPN
cana-2345	176	8	model	model	NOUN
cana-2345	176	9	architecture	architecture	NOUN
cana-2345	176	10	utilized	utilize	VERB
cana-2345	176	11	bert	bert	NOUN
cana-2345	176	12	embeddings	embedding	NOUN
cana-2345	176	13	to	to	PART
cana-2345	176	14	capture	capture	VERB
cana-2345	176	15	word	word	NOUN
cana-2345	176	16	semantics	semantic	NOUN
cana-2345	176	17	,	,	PUNCT
cana-2345	176	18	followed	follow	VERB
cana-2345	176	19	by	by	ADP
cana-2345	176	20	a	a	DET
cana-2345	176	21	bidirectional	bidirectional	ADJ
cana-2345	176	22	lstm	lstm	NOUN
cana-2345	176	23	(	(	PUNCT
cana-2345	176	24	bilstm	bilstm	NOUN
cana-2345	176	25	)	)	PUNCT
cana-2345	176	26	to	to	PART
cana-2345	176	27	understand	understand	VERB
cana-2345	176	28	sequential	sequential	ADJ
cana-2345	176	29	dependencies	dependency	NOUN
cana-2345	176	30	from	from	ADP
cana-2345	176	31	both	both	DET
cana-2345	176	32	directions	direction	NOUN
cana-2345	176	33	.	.	PUNCT
cana-2345	177	1	a	a	DET
cana-2345	177	2	convolutional	convolutional	ADJ
cana-2345	177	3	neural	neural	ADJ
cana-2345	177	4	network	network	NOUN
cana-2345	177	5	(	(	PUNCT
cana-2345	177	6	cnn	cnn	PROPN
cana-2345	177	7	)	)	PUNCT
cana-2345	177	8	layer	layer	NOUN
cana-2345	177	9	extracted	extract	VERB
cana-2345	177	10	local	local	ADJ
cana-2345	177	11	features	feature	NOUN
cana-2345	177	12	,	,	PUNCT
cana-2345	177	13	and	and	CCONJ
cana-2345	177	14	an	an	DET
cana-2345	177	15	attention	attention	NOUN
cana-2345	177	16	mechanism	mechanism	NOUN
cana-2345	177	17	focused	focus	VERB
cana-2345	177	18	on	on	ADP
cana-2345	177	19	sentiment	sentiment	NOUN
cana-2345	177	20	-	-	PUNCT
cana-2345	177	21	relevant	relevant	ADJ
cana-2345	177	22	words	word	NOUN
cana-2345	177	23	.	.	PUNCT
cana-2345	178	1	the	the	DET
cana-2345	178	2	outputs	output	NOUN
cana-2345	178	3	were	be	AUX
cana-2345	178	4	combined	combine	VERB
cana-2345	178	5	in	in	ADP
cana-2345	178	6	a	a	DET
cana-2345	178	7	nonlinear	nonlinear	ADJ
cana-2345	178	8	fusion	fusion	NOUN
cana-2345	178	9	layer	layer	NOUN
cana-2345	178	10	with	with	ADP
cana-2345	178	11	relu	relu	NOUN
cana-2345	178	12	activation	activation	NOUN
cana-2345	178	13	,	,	PUNCT
cana-2345	178	14	followed	follow	VERB
cana-2345	178	15	by	by	ADP
cana-2345	178	16	a	a	DET
cana-2345	178	17	dense	dense	ADJ
cana-2345	178	18	layer	layer	NOUN
cana-2345	178	19	with	with	ADP
cana-2345	178	20	dropout	dropout	NOUN
cana-2345	178	21	regularization	regularization	NOUN
cana-2345	178	22	to	to	PART
cana-2345	178	23	prevent	prevent	VERB
cana-2345	178	24	overfitting	overfitting	NOUN
cana-2345	178	25	.	.	PUNCT
cana-2345	179	1	the	the	DET
cana-2345	179	2	final	final	ADJ
cana-2345	179	3	sentiment	sentiment	NOUN
cana-2345	179	4	prediction	prediction	NOUN
cana-2345	179	5	was	be	AUX
cana-2345	179	6	made	make	VERB
cana-2345	179	7	using	use	VERB
cana-2345	179	8	a	a	DET
cana-2345	179	9	fully	fully	ADV
cana-2345	179	10	connected	connect	VERB
cana-2345	179	11	layer	layer	NOUN
cana-2345	179	12	with	with	ADP
cana-2345	179	13	softmax	softmax	NOUN
cana-2345	179	14	activation	activation	NOUN
cana-2345	179	15	.	.	PUNCT
cana-2345	180	1	training	training	NOUN
cana-2345	180	2	was	be	AUX
cana-2345	180	3	conducted	conduct	VERB
cana-2345	180	4	using	use	VERB
cana-2345	180	5	the	the	DET
cana-2345	180	6	adam	adam	PROPN
cana-2345	180	7	optimizer	optimizer	NOUN
cana-2345	180	8	with	with	ADP
cana-2345	180	9	a	a	DET
cana-2345	180	10	learning	learn	VERB
cana-2345	180	11	rate	rate	NOUN
cana-2345	180	12	of	of	ADP
cana-2345	180	13	0.001	0.001	NUM
cana-2345	180	14	,	,	PUNCT
cana-2345	180	15	a	a	DET
cana-2345	180	16	batch	batch	NOUN
cana-2345	180	17	size	size	NOUN
cana-2345	180	18	of	of	ADP
cana-2345	180	19	64	64	NUM
cana-2345	180	20	,	,	PUNCT
cana-2345	180	21	and	and	CCONJ
cana-2345	180	22	early	early	ADV
cana-2345	180	23	stopping	stopping	NOUN
cana-2345	180	24	based	base	VERB
cana-2345	180	25	on	on	ADP
cana-2345	180	26	validation	validation	NOUN
cana-2345	180	27	loss	loss	NOUN
cana-2345	180	28	to	to	PART
cana-2345	180	29	avoid	avoid	VERB
cana-2345	180	30	overfitting	overfitte	VERB
cana-2345	180	31	.	.	PUNCT
cana-2345	181	1	the	the	DET
cana-2345	181	2	model	model	NOUN
cana-2345	181	3	was	be	AUX
cana-2345	181	4	trained	train	VERB
cana-2345	181	5	for	for	ADP
cana-2345	181	6	up	up	ADP
cana-2345	181	7	to	to	PART
cana-2345	181	8	50	50	NUM
cana-2345	181	9	epochs	epoch	NOUN
cana-2345	181	10	,	,	PUNCT
cana-2345	181	11	with	with	ADP
cana-2345	181	12	dropout	dropout	NOUN
cana-2345	181	13	rates	rate	NOUN
cana-2345	181	14	of	of	ADP
cana-2345	181	15	0.5	0.5	NUM
cana-2345	181	16	applied	apply	VERB
cana-2345	181	17	in	in	ADP
cana-2345	181	18	the	the	DET
cana-2345	181	19	dense	dense	ADJ
cana-2345	181	20	layers	layer	NOUN
cana-2345	181	21	.	.	PUNCT
cana-2345	182	1	communications	communication	NOUN
cana-2345	182	2	on	on	ADP
cana-2345	182	3	applied	apply	VERB
cana-2345	182	4	nonlinear	nonlinear	ADJ
cana-2345	182	5	analysis	analysis	NOUN
cana-2345	182	6	issn	issn	NOUN
cana-2345	182	7	:	:	PUNCT
cana-2345	182	8	1074	1074	NUM
cana-2345	182	9	-	-	PUNCT
cana-2345	182	10	133x	133x	NUM
cana-2345	182	11	vol	vol	NOUN
cana-2345	182	12	32	32	NUM
cana-2345	182	13	no	no	NOUN
cana-2345	182	14	.	.	PUNCT
cana-2345	183	1	1s	1s	NUM
cana-2345	183	2	(	(	PUNCT
cana-2345	183	3	2025	2025	NUM
cana-2345	183	4	)	)	PUNCT
cana-2345	183	5	580	580	NUM
cana-2345	183	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2345	183	7	performance	performance	NOUN
cana-2345	183	8	was	be	AUX
cana-2345	183	9	assessed	assess	VERB
cana-2345	183	10	using	use	VERB
cana-2345	183	11	several	several	ADJ
cana-2345	183	12	metrics	metric	NOUN
cana-2345	183	13	:	:	PUNCT
cana-2345	183	14	accuracy	accuracy	NOUN
cana-2345	183	15	,	,	PUNCT
cana-2345	183	16	precision	precision	NOUN
cana-2345	183	17	,	,	PUNCT
cana-2345	183	18	recall	recall	NOUN
cana-2345	183	19	,	,	PUNCT
cana-2345	183	20	and	and	CCONJ
cana-2345	183	21	f1	f1	NOUN
cana-2345	183	22	-	-	PUNCT
cana-2345	183	23	score	score	NOUN
cana-2345	183	24	.	.	PUNCT
cana-2345	184	1	these	these	DET
cana-2345	184	2	metrics	metric	NOUN
cana-2345	184	3	provided	provide	VERB
cana-2345	184	4	a	a	DET
cana-2345	184	5	comprehensive	comprehensive	ADJ
cana-2345	184	6	evaluation	evaluation	NOUN
cana-2345	184	7	of	of	ADP
cana-2345	184	8	each	each	DET
cana-2345	184	9	model	model	NOUN
cana-2345	184	10	's	's	PART
cana-2345	184	11	classification	classification	NOUN
cana-2345	184	12	capabilities	capability	NOUN
cana-2345	184	13	,	,	PUNCT
cana-2345	184	14	focusing	focus	VERB
cana-2345	184	15	on	on	ADP
cana-2345	184	16	both	both	CCONJ
cana-2345	184	17	the	the	DET
cana-2345	184	18	correct	correct	ADJ
cana-2345	184	19	identification	identification	NOUN
cana-2345	184	20	of	of	ADP
cana-2345	184	21	sentiments	sentiment	NOUN
cana-2345	184	22	and	and	CCONJ
cana-2345	184	23	the	the	DET
cana-2345	184	24	minimization	minimization	NOUN
cana-2345	184	25	of	of	ADP
cana-2345	184	26	false	false	ADJ
cana-2345	184	27	positives	positive	NOUN
cana-2345	184	28	and	and	CCONJ
cana-2345	184	29	negatives	negative	NOUN
cana-2345	184	30	.	.	PUNCT
cana-2345	185	1	two	two	NUM
cana-2345	185	2	contemporary	contemporary	ADJ
cana-2345	185	3	models	model	NOUN
cana-2345	185	4	,	,	PUNCT
cana-2345	185	5	rmdl	rmdl	NOUN
cana-2345	185	6	[	[	X
cana-2345	185	7	11	11	NUM
cana-2345	185	8	]	]	X
cana-2345	185	9	(	(	PUNCT
cana-2345	185	10	random	random	ADJ
cana-2345	185	11	multimodel	multimodel	VERB
cana-2345	185	12	deep	deep	ADJ
cana-2345	185	13	learning	learning	NOUN
cana-2345	185	14	)	)	PUNCT
cana-2345	185	15	and	and	CCONJ
cana-2345	185	16	clm	clm	X
cana-2345	186	1	[	[	X
cana-2345	186	2	16	16	NUM
cana-2345	186	3	]	]	X
cana-2345	186	4	(	(	PUNCT
cana-2345	186	5	convolutional	convolutional	ADJ
cana-2345	186	6	lstm	lstm	NOUN
cana-2345	186	7	model	model	NOUN
cana-2345	186	8	)	)	PUNCT
cana-2345	186	9	,	,	PUNCT
cana-2345	186	10	were	be	AUX
cana-2345	186	11	selected	select	VERB
cana-2345	186	12	as	as	ADP
cana-2345	186	13	baselines	baseline	NOUN
cana-2345	186	14	for	for	ADP
cana-2345	186	15	comparison	comparison	NOUN
cana-2345	186	16	with	with	ADP
cana-2345	186	17	npsc	npsc	PROPN
cana-2345	186	18	.	.	PUNCT
cana-2345	187	1	rmdl	rmdl	NOUN
cana-2345	187	2	uses	use	VERB
cana-2345	187	3	an	an	DET
cana-2345	187	4	ensemble	ensemble	NOUN
cana-2345	187	5	of	of	ADP
cana-2345	187	6	deep	deep	ADJ
cana-2345	187	7	learning	learning	NOUN
cana-2345	187	8	models	model	NOUN
cana-2345	187	9	,	,	PUNCT
cana-2345	187	10	including	include	VERB
cana-2345	187	11	lstms	lstms	NOUN
cana-2345	187	12	and	and	CCONJ
cana-2345	187	13	cnns	cnn	NOUN
cana-2345	187	14	,	,	PUNCT
cana-2345	187	15	while	while	SCONJ
cana-2345	187	16	clm	clm	NOUN
cana-2345	187	17	combines	combine	VERB
cana-2345	187	18	convolutional	convolutional	ADJ
cana-2345	187	19	layers	layer	NOUN
cana-2345	187	20	with	with	ADP
cana-2345	187	21	lstm	lstm	ADJ
cana-2345	187	22	networks	network	NOUN
cana-2345	187	23	to	to	PART
cana-2345	187	24	capture	capture	VERB
cana-2345	187	25	both	both	DET
cana-2345	187	26	local	local	ADJ
cana-2345	187	27	and	and	CCONJ
cana-2345	187	28	sequential	sequential	ADJ
cana-2345	187	29	patterns	pattern	NOUN
cana-2345	187	30	in	in	ADP
cana-2345	187	31	text	text	NOUN
cana-2345	187	32	data	datum	NOUN
cana-2345	187	33	.	.	PUNCT
cana-2345	188	1	these	these	DET
cana-2345	188	2	models	model	NOUN
cana-2345	188	3	represent	represent	VERB
cana-2345	188	4	advanced	advanced	ADJ
cana-2345	188	5	yet	yet	ADV
cana-2345	188	6	distinct	distinct	ADJ
cana-2345	188	7	approaches	approach	NOUN
cana-2345	188	8	in	in	ADP
cana-2345	188	9	sentiment	sentiment	NOUN
cana-2345	188	10	classification	classification	NOUN
cana-2345	188	11	,	,	PUNCT
cana-2345	188	12	making	make	VERB
cana-2345	188	13	them	they	PRON
cana-2345	188	14	suitable	suitable	ADJ
cana-2345	188	15	comparators	comparator	NOUN
cana-2345	188	16	for	for	ADP
cana-2345	188	17	evaluating	evaluate	VERB
cana-2345	188	18	npsc	npsc	NOUN
cana-2345	188	19	.	.	PUNCT
cana-2345	189	1	3.1	3.1	NUM
cana-2345	189	2	results	result	NOUN
cana-2345	189	3	and	and	CCONJ
cana-2345	189	4	analysis	analysis	NOUN
cana-2345	189	5	the	the	DET
cana-2345	189	6	results	result	NOUN
cana-2345	189	7	demonstrated	demonstrate	VERB
cana-2345	189	8	that	that	SCONJ
cana-2345	189	9	npsc	npsc	PROPN
cana-2345	189	10	outperformed	outperform	VERB
cana-2345	189	11	both	both	DET
cana-2345	189	12	rmdl	rmdl	NOUN
cana-2345	190	1	[	[	X
cana-2345	190	2	11	11	NUM
cana-2345	190	3	]	]	PUNCT
cana-2345	190	4	and	and	CCONJ
cana-2345	190	5	clm	clm	X
cana-2345	191	1	[	[	X
cana-2345	191	2	16	16	NUM
cana-2345	191	3	]	]	PUNCT
cana-2345	191	4	across	across	ADP
cana-2345	191	5	all	all	DET
cana-2345	191	6	evaluated	evaluated	ADJ
cana-2345	191	7	metrics	metric	NOUN
cana-2345	191	8	,	,	PUNCT
cana-2345	191	9	as	as	SCONJ
cana-2345	191	10	shown	show	VERB
cana-2345	191	11	in	in	ADP
cana-2345	191	12	table	table	NOUN
cana-2345	191	13	1	1	NUM
cana-2345	191	14	.	.	PUNCT
cana-2345	192	1	npsc	npsc	PROPN
cana-2345	192	2	achieved	achieve	VERB
cana-2345	192	3	the	the	DET
cana-2345	192	4	highest	high	ADJ
cana-2345	192	5	accuracy	accuracy	NOUN
cana-2345	192	6	,	,	PUNCT
cana-2345	192	7	precision	precision	NOUN
cana-2345	192	8	,	,	PUNCT
cana-2345	192	9	recall	recall	NOUN
cana-2345	192	10	,	,	PUNCT
cana-2345	192	11	and	and	CCONJ
cana-2345	192	12	f1	f1	NOUN
cana-2345	192	13	-	-	PUNCT
cana-2345	192	14	score	score	NOUN
cana-2345	192	15	,	,	PUNCT
cana-2345	192	16	highlighting	highlight	VERB
cana-2345	192	17	its	its	PRON
cana-2345	192	18	superior	superior	ADJ
cana-2345	192	19	capability	capability	NOUN
cana-2345	192	20	in	in	ADP
cana-2345	192	21	precise	precise	ADJ
cana-2345	192	22	sentiment	sentiment	NOUN
cana-2345	192	23	classification	classification	NOUN
cana-2345	192	24	.	.	PUNCT
cana-2345	193	1	while	while	SCONJ
cana-2345	193	2	clm	clm	PROPN
cana-2345	193	3	]	]	PUNCT
cana-2345	193	4	showed	show	VERB
cana-2345	193	5	better	well	ADJ
cana-2345	193	6	performance	performance	NOUN
cana-2345	193	7	than	than	ADP
cana-2345	193	8	rmdl	rmdl	NOUN
cana-2345	193	9	,	,	PUNCT
cana-2345	193	10	it	it	PRON
cana-2345	193	11	still	still	ADV
cana-2345	193	12	lagged	lag	VERB
cana-2345	193	13	behind	behind	ADP
cana-2345	193	14	npsc	npsc	NOUN
cana-2345	193	15	,	,	PUNCT
cana-2345	193	16	particularly	particularly	ADV
cana-2345	193	17	in	in	ADP
cana-2345	193	18	precision	precision	NOUN
cana-2345	193	19	and	and	CCONJ
cana-2345	193	20	recall	recall	NOUN
cana-2345	193	21	,	,	PUNCT
cana-2345	193	22	indicating	indicate	VERB
cana-2345	193	23	that	that	SCONJ
cana-2345	193	24	npsc	npsc	PROPN
cana-2345	193	25	’s	’s	PART
cana-2345	193	26	nonlinear	nonlinear	ADJ
cana-2345	193	27	architecture	architecture	NOUN
cana-2345	193	28	effectively	effectively	ADV
cana-2345	193	29	captures	capture	NOUN
cana-2345	193	30	and	and	CCONJ
cana-2345	193	31	leverages	leverage	VERB
cana-2345	193	32	complex	complex	ADJ
cana-2345	193	33	relationships	relationship	NOUN
cana-2345	193	34	in	in	ADP
cana-2345	193	35	textual	textual	ADJ
cana-2345	193	36	data	datum	NOUN
cana-2345	193	37	.	.	PUNCT
cana-2345	194	1	table	table	NOUN
cana-2345	194	2	1	1	NUM
cana-2345	194	3	:	:	PUNCT
cana-2345	194	4	performance	performance	NOUN
cana-2345	194	5	comparison	comparison	NOUN
cana-2345	194	6	of	of	ADP
cana-2345	194	7	npsc	npsc	NOUN
cana-2345	194	8	,	,	PUNCT
cana-2345	194	9	rmdl	rmdl	NOUN
cana-2345	194	10	,	,	PUNCT
cana-2345	194	11	and	and	CCONJ
cana-2345	194	12	clm	clm	PROPN
cana-2345	194	13	model	model	NOUN
cana-2345	194	14	accuracy	accuracy	NOUN
cana-2345	194	15	(	(	PUNCT
cana-2345	194	16	%	%	INTJ
cana-2345	194	17	)	)	PUNCT
cana-2345	194	18	precision	precision	NOUN
cana-2345	194	19	(	(	PUNCT
cana-2345	194	20	%	%	INTJ
cana-2345	194	21	)	)	PUNCT
cana-2345	194	22	recall	recall	NOUN
cana-2345	194	23	(	(	PUNCT
cana-2345	194	24	%	%	NOUN
cana-2345	194	25	)	)	PUNCT
cana-2345	194	26	f1	f1	NOUN
cana-2345	194	27	-	-	PUNCT
cana-2345	194	28	score	score	NOUN
cana-2345	194	29	(	(	PUNCT
cana-2345	194	30	%	%	INTJ
cana-2345	194	31	)	)	PUNCT
cana-2345	194	32	npsc	npsc	VERB
cana-2345	194	33	94.2	94.2	NUM
cana-2345	194	34	93.8	93.8	NUM
cana-2345	194	35	94.5	94.5	NUM
cana-2345	194	36	94.1	94.1	NUM
cana-2345	194	37	clm	clm	NOUN
cana-2345	194	38	91.6	91.6	NUM
cana-2345	194	39	90.9	90.9	NUM
cana-2345	194	40	91.2	91.2	NUM
cana-2345	194	41	91.0	91.0	NUM
cana-2345	194	42	rmdl	rmdl	NOUN
cana-2345	194	43	88.3	88.3	NUM
cana-2345	194	44	87.5	87.5	NUM
cana-2345	194	45	88.0	88.0	NUM
cana-2345	194	46	87.7	87.7	NUM
cana-2345	194	47	ablation	ablation	NOUN
cana-2345	194	48	study	study	NOUN
cana-2345	194	49	:	:	PUNCT
cana-2345	194	50	the	the	DET
cana-2345	194	51	ablation	ablation	NOUN
cana-2345	194	52	study	study	NOUN
cana-2345	194	53	was	be	AUX
cana-2345	194	54	conducted	conduct	VERB
cana-2345	194	55	to	to	PART
cana-2345	194	56	assess	assess	VERB
cana-2345	194	57	the	the	DET
cana-2345	194	58	contribution	contribution	NOUN
cana-2345	194	59	of	of	ADP
cana-2345	194	60	individual	individual	ADJ
cana-2345	194	61	components	component	NOUN
cana-2345	194	62	within	within	ADP
cana-2345	194	63	npsc	npsc	NOUN
cana-2345	194	64	by	by	ADP
cana-2345	194	65	systematically	systematically	ADV
cana-2345	194	66	removing	remove	VERB
cana-2345	194	67	key	key	ADJ
cana-2345	194	68	elements	element	NOUN
cana-2345	194	69	and	and	CCONJ
cana-2345	194	70	observing	observe	VERB
cana-2345	194	71	the	the	DET
cana-2345	194	72	impact	impact	NOUN
cana-2345	194	73	on	on	ADP
cana-2345	194	74	performance	performance	NOUN
cana-2345	194	75	.	.	PUNCT
cana-2345	195	1	table	table	NOUN
cana-2345	195	2	2	2	NUM
cana-2345	195	3	shows	show	VERB
cana-2345	195	4	the	the	DET
cana-2345	195	5	results	result	NOUN
cana-2345	195	6	of	of	ADP
cana-2345	195	7	removing	remove	VERB
cana-2345	195	8	the	the	DET
cana-2345	195	9	cnn	cnn	PROPN
cana-2345	195	10	layer	layer	NOUN
cana-2345	195	11	,	,	PUNCT
cana-2345	195	12	attention	attention	NOUN
cana-2345	195	13	mechanism	mechanism	NOUN
cana-2345	195	14	,	,	PUNCT
cana-2345	195	15	and	and	CCONJ
cana-2345	195	16	the	the	DET
cana-2345	195	17	nonlinear	nonlinear	ADJ
cana-2345	195	18	feature	feature	NOUN
cana-2345	195	19	fusion	fusion	NOUN
cana-2345	195	20	layer	layer	NOUN
cana-2345	195	21	.	.	PUNCT
cana-2345	196	1	table	table	NOUN
cana-2345	196	2	2	2	NUM
cana-2345	196	3	:	:	PUNCT
cana-2345	196	4	ablation	ablation	NOUN
cana-2345	196	5	study	study	NOUN
cana-2345	196	6	results	result	NOUN
cana-2345	196	7	of	of	ADP
cana-2345	196	8	npsc	npsc	ADJ
cana-2345	196	9	model	model	NOUN
cana-2345	196	10	configuration	configuration	NOUN
cana-2345	196	11	accuracy	accuracy	NOUN
cana-2345	196	12	(	(	PUNCT
cana-2345	196	13	%	%	INTJ
cana-2345	196	14	)	)	PUNCT
cana-2345	196	15	precision	precision	NOUN
cana-2345	196	16	(	(	PUNCT
cana-2345	196	17	%	%	INTJ
cana-2345	196	18	)	)	PUNCT
cana-2345	196	19	recall	recall	NOUN
cana-2345	196	20	(	(	PUNCT
cana-2345	196	21	%	%	NOUN
cana-2345	196	22	)	)	PUNCT
cana-2345	196	23	f1	f1	NOUN
cana-2345	196	24	-	-	PUNCT
cana-2345	196	25	score	score	NOUN
cana-2345	196	26	(	(	PUNCT
cana-2345	196	27	%	%	INTJ
cana-2345	196	28	)	)	PUNCT
cana-2345	196	29	full	full	ADJ
cana-2345	196	30	npsc	npsc	NOUN
cana-2345	196	31	94.2	94.2	NUM
cana-2345	196	32	93.8	93.8	NUM
cana-2345	196	33	94.5	94.5	NUM
cana-2345	196	34	94.1	94.1	NUM
cana-2345	196	35	npsc	npsc	NOUN
cana-2345	196	36	without	without	ADP
cana-2345	196	37	cnn	cnn	PROPN
cana-2345	196	38	90.1	90.1	NUM
cana-2345	196	39	89.7	89.7	NUM
cana-2345	196	40	89.9	89.9	NUM
cana-2345	196	41	89.8	89.8	NUM
cana-2345	196	42	npsc	npsc	NOUN
cana-2345	196	43	without	without	ADP
cana-2345	196	44	attention	attention	NOUN
cana-2345	196	45	88.9	88.9	NUM
cana-2345	196	46	88.2	88.2	NUM
cana-2345	196	47	88.6	88.6	NUM
cana-2345	196	48	88.4	88.4	NUM
cana-2345	196	49	npsc	npsc	NOUN
cana-2345	196	50	without	without	ADP
cana-2345	196	51	nonlinear	nonlinear	ADJ
cana-2345	196	52	fusion	fusion	NOUN
cana-2345	196	53	86.5	86.5	NUM
cana-2345	196	54	85.8	85.8	NUM
cana-2345	196	55	86.1	86.1	NUM
cana-2345	196	56	85.9	85.9	NUM
cana-2345	196	57	the	the	DET
cana-2345	196	58	results	result	NOUN
cana-2345	196	59	highlight	highlight	VERB
cana-2345	196	60	that	that	SCONJ
cana-2345	196	61	each	each	DET
cana-2345	196	62	component	component	NOUN
cana-2345	196	63	plays	play	VERB
cana-2345	196	64	a	a	DET
cana-2345	196	65	critical	critical	ADJ
cana-2345	196	66	role	role	NOUN
cana-2345	196	67	in	in	ADP
cana-2345	196	68	the	the	DET
cana-2345	196	69	overall	overall	ADJ
cana-2345	196	70	performance	performance	NOUN
cana-2345	196	71	of	of	ADP
cana-2345	196	72	npsc	npsc	NOUN
cana-2345	196	73	.	.	PUNCT
cana-2345	197	1	removing	remove	VERB
cana-2345	197	2	the	the	DET
cana-2345	197	3	cnn	cnn	NOUN
cana-2345	197	4	layer	layer	NOUN
cana-2345	197	5	significantly	significantly	ADV
cana-2345	197	6	reduced	reduce	VERB
cana-2345	197	7	the	the	DET
cana-2345	197	8	accuracy	accuracy	NOUN
cana-2345	197	9	and	and	CCONJ
cana-2345	197	10	precision	precision	NOUN
cana-2345	197	11	,	,	PUNCT
cana-2345	197	12	demonstrating	demonstrate	VERB
cana-2345	197	13	its	its	PRON
cana-2345	197	14	importance	importance	NOUN
cana-2345	197	15	in	in	ADP
cana-2345	197	16	capturing	capture	VERB
cana-2345	197	17	local	local	ADJ
cana-2345	197	18	patterns	pattern	NOUN
cana-2345	197	19	.	.	PUNCT
cana-2345	198	1	the	the	DET
cana-2345	198	2	absence	absence	NOUN
cana-2345	198	3	of	of	ADP
cana-2345	198	4	the	the	DET
cana-2345	198	5	attention	attention	NOUN
cana-2345	198	6	mechanism	mechanism	NOUN
cana-2345	198	7	led	lead	VERB
cana-2345	198	8	to	to	ADP
cana-2345	198	9	a	a	DET
cana-2345	198	10	noticeable	noticeable	ADJ
cana-2345	198	11	drop	drop	NOUN
cana-2345	198	12	in	in	ADP
cana-2345	198	13	recall	recall	NOUN
cana-2345	198	14	,	,	PUNCT
cana-2345	198	15	emphasizing	emphasize	VERB
cana-2345	198	16	its	its	PRON
cana-2345	198	17	role	role	NOUN
cana-2345	198	18	in	in	ADP
cana-2345	198	19	focusing	focus	VERB
cana-2345	198	20	on	on	ADP
cana-2345	198	21	sentiment	sentiment	NOUN
cana-2345	198	22	-	-	PUNCT
cana-2345	198	23	critical	critical	ADJ
cana-2345	198	24	elements	element	NOUN
cana-2345	198	25	.	.	PUNCT
cana-2345	199	1	the	the	DET
cana-2345	199	2	nonlinear	nonlinear	ADJ
cana-2345	199	3	fusion	fusion	NOUN
cana-2345	199	4	layer	layer	NOUN
cana-2345	199	5	was	be	AUX
cana-2345	199	6	particularly	particularly	ADV
cana-2345	199	7	impactful	impactful	ADJ
cana-2345	199	8	,	,	PUNCT
cana-2345	199	9	as	as	SCONJ
cana-2345	199	10	its	its	PRON
cana-2345	199	11	removal	removal	NOUN
cana-2345	199	12	caused	cause	VERB
cana-2345	199	13	the	the	DET
cana-2345	199	14	most	most	ADV
cana-2345	199	15	significant	significant	ADJ
cana-2345	199	16	decrease	decrease	NOUN
cana-2345	199	17	in	in	ADP
cana-2345	199	18	performance	performance	NOUN
cana-2345	199	19	metrics	metric	NOUN
cana-2345	199	20	,	,	PUNCT
cana-2345	199	21	underlining	underline	VERB
cana-2345	199	22	its	its	PRON
cana-2345	199	23	essential	essential	ADJ
cana-2345	199	24	function	function	NOUN
cana-2345	199	25	in	in	ADP
cana-2345	199	26	integrating	integrate	VERB
cana-2345	199	27	features	feature	NOUN
cana-2345	199	28	to	to	PART
cana-2345	199	29	capture	capture	VERB
cana-2345	199	30	complex	complex	ADJ
cana-2345	199	31	dependencies	dependency	NOUN
cana-2345	199	32	.	.	PUNCT
cana-2345	200	1	communications	communication	NOUN
cana-2345	200	2	on	on	ADP
cana-2345	200	3	applied	apply	VERB
cana-2345	200	4	nonlinear	nonlinear	ADJ
cana-2345	200	5	analysis	analysis	NOUN
cana-2345	200	6	issn	issn	NOUN
cana-2345	200	7	:	:	PUNCT
cana-2345	200	8	1074	1074	NUM
cana-2345	200	9	-	-	PUNCT
cana-2345	200	10	133x	133x	NUM
cana-2345	200	11	vol	vol	NOUN
cana-2345	200	12	32	32	NUM
cana-2345	200	13	no	no	NOUN
cana-2345	200	14	.	.	PUNCT
cana-2345	201	1	1s	1s	NUM
cana-2345	201	2	(	(	PUNCT
cana-2345	201	3	2025	2025	NUM
cana-2345	201	4	)	)	PUNCT
cana-2345	201	5	581	581	NUM
cana-2345	201	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2345	201	7	error	error	NOUN
cana-2345	201	8	analysis	analysis	NOUN
cana-2345	201	9	:	:	PUNCT
cana-2345	201	10	an	an	DET
cana-2345	201	11	error	error	NOUN
cana-2345	201	12	analysis	analysis	NOUN
cana-2345	201	13	was	be	AUX
cana-2345	201	14	performed	perform	VERB
cana-2345	201	15	to	to	PART
cana-2345	201	16	examine	examine	VERB
cana-2345	201	17	the	the	DET
cana-2345	201	18	misclassifications	misclassification	NOUN
cana-2345	201	19	made	make	VERB
cana-2345	201	20	by	by	ADP
cana-2345	201	21	npsc	npsc	NOUN
cana-2345	201	22	,	,	PUNCT
cana-2345	201	23	with	with	ADP
cana-2345	201	24	a	a	DET
cana-2345	201	25	focus	focus	NOUN
cana-2345	201	26	on	on	ADP
cana-2345	201	27	understanding	understand	VERB
cana-2345	201	28	the	the	DET
cana-2345	201	29	nature	nature	NOUN
cana-2345	201	30	of	of	ADP
cana-2345	201	31	errors	error	NOUN
cana-2345	201	32	and	and	CCONJ
cana-2345	201	33	identifying	identify	VERB
cana-2345	201	34	potential	potential	ADJ
cana-2345	201	35	areas	area	NOUN
cana-2345	201	36	for	for	ADP
cana-2345	201	37	improvement	improvement	NOUN
cana-2345	201	38	.	.	PUNCT
cana-2345	202	1	the	the	DET
cana-2345	202	2	analysis	analysis	NOUN
cana-2345	202	3	revealed	reveal	VERB
cana-2345	202	4	that	that	SCONJ
cana-2345	202	5	most	most	ADJ
cana-2345	202	6	errors	error	NOUN
cana-2345	202	7	occurred	occur	VERB
cana-2345	202	8	in	in	ADP
cana-2345	202	9	reviews	review	NOUN
cana-2345	202	10	with	with	ADP
cana-2345	202	11	mixed	mixed	ADJ
cana-2345	202	12	or	or	CCONJ
cana-2345	202	13	ambiguous	ambiguous	ADJ
cana-2345	202	14	sentiments	sentiment	NOUN
cana-2345	202	15	,	,	PUNCT
cana-2345	202	16	where	where	SCONJ
cana-2345	202	17	both	both	DET
cana-2345	202	18	positive	positive	ADJ
cana-2345	202	19	and	and	CCONJ
cana-2345	202	20	negative	negative	ADJ
cana-2345	202	21	elements	element	NOUN
cana-2345	202	22	were	be	AUX
cana-2345	202	23	present	present	ADJ
cana-2345	202	24	within	within	ADP
cana-2345	202	25	the	the	DET
cana-2345	202	26	same	same	ADJ
cana-2345	202	27	review	review	NOUN
cana-2345	202	28	.	.	PUNCT
cana-2345	203	1	for	for	ADP
cana-2345	203	2	instance	instance	NOUN
cana-2345	203	3	,	,	PUNCT
cana-2345	203	4	reviews	review	NOUN
cana-2345	203	5	that	that	PRON
cana-2345	203	6	praised	praise	VERB
cana-2345	203	7	certain	certain	ADJ
cana-2345	203	8	aspects	aspect	NOUN
cana-2345	203	9	of	of	ADP
cana-2345	203	10	a	a	DET
cana-2345	203	11	movie	movie	NOUN
cana-2345	203	12	while	while	SCONJ
cana-2345	203	13	criticizing	criticize	VERB
cana-2345	203	14	others	other	NOUN
cana-2345	203	15	posed	pose	VERB
cana-2345	203	16	a	a	DET
cana-2345	203	17	challenge	challenge	NOUN
cana-2345	203	18	for	for	ADP
cana-2345	203	19	precise	precise	ADJ
cana-2345	203	20	classification	classification	NOUN
cana-2345	203	21	.	.	PUNCT
cana-2345	204	1	table	table	NOUN
cana-2345	204	2	3	3	NUM
cana-2345	204	3	:	:	PUNCT
cana-2345	204	4	error	error	NOUN
cana-2345	204	5	analysis	analysis	NOUN
cana-2345	204	6	results	result	VERB
cana-2345	204	7	error	error	NOUN
cana-2345	204	8	type	type	NOUN
cana-2345	204	9	npsc	npsc	NOUN
cana-2345	204	10	misclassifications	misclassification	NOUN
cana-2345	204	11	(	(	PUNCT
cana-2345	204	12	%	%	INTJ
cana-2345	204	13	)	)	PUNCT
cana-2345	204	14	clm	clm	PROPN
cana-2345	204	15	misclassifications	misclassification	NOUN
cana-2345	204	16	(	(	PUNCT
cana-2345	204	17	%	%	INTJ
cana-2345	204	18	)	)	PUNCT
cana-2345	204	19	rmdl	rmdl	NOUN
cana-2345	204	20	misclassifications	misclassification	NOUN
cana-2345	204	21	(	(	PUNCT
cana-2345	204	22	%	%	INTJ
cana-2345	204	23	)	)	PUNCT
cana-2345	204	24	mixed	mixed	ADJ
cana-2345	204	25	sentiments	sentiment	NOUN
cana-2345	204	26	5.6	5.6	NUM
cana-2345	204	27	7.4	7.4	NUM
cana-2345	204	28	9.2	9.2	NUM
cana-2345	204	29	neutral	neutral	ADJ
cana-2345	204	30	or	or	CCONJ
cana-2345	204	31	subtle	subtle	ADJ
cana-2345	204	32	sentiments	sentiment	NOUN
cana-2345	204	33	3.1	3.1	NUM
cana-2345	204	34	4.6	4.6	NUM
cana-2345	204	35	6.3	6.3	NUM
cana-2345	204	36	complex	complex	ADJ
cana-2345	204	37	language	language	NOUN
cana-2345	204	38	usage	usage	NOUN
cana-2345	204	39	2.8	2.8	NUM
cana-2345	204	40	3.9	3.9	NUM
cana-2345	204	41	5.1	5.1	NUM
cana-2345	204	42	npsc	npsc	NOUN
cana-2345	204	43	demonstrated	demonstrate	VERB
cana-2345	204	44	a	a	DET
cana-2345	204	45	lower	low	ADJ
cana-2345	204	46	misclassification	misclassification	NOUN
cana-2345	204	47	rate	rate	NOUN
cana-2345	204	48	compared	compare	VERB
cana-2345	204	49	to	to	ADP
cana-2345	204	50	clm	clm	PROPN
cana-2345	205	1	[	[	X
cana-2345	205	2	16	16	NUM
cana-2345	205	3	]	]	PUNCT
cana-2345	205	4	and	and	CCONJ
cana-2345	205	5	rmdl	rmdl	NOUN
cana-2345	205	6	[	[	X
cana-2345	205	7	11	11	NUM
cana-2345	205	8	]	]	PUNCT
cana-2345	205	9	across	across	ADP
cana-2345	205	10	all	all	DET
cana-2345	205	11	error	error	NOUN
cana-2345	205	12	categories	category	NOUN
cana-2345	205	13	.	.	PUNCT
cana-2345	206	1	the	the	DET
cana-2345	206	2	model	model	NOUN
cana-2345	206	3	’s	’s	PART
cana-2345	206	4	advanced	advanced	ADJ
cana-2345	206	5	attention	attention	NOUN
cana-2345	206	6	mechanism	mechanism	NOUN
cana-2345	206	7	and	and	CCONJ
cana-2345	206	8	feature	feature	NOUN
cana-2345	206	9	fusion	fusion	NOUN
cana-2345	206	10	enabled	enable	VERB
cana-2345	206	11	it	it	PRON
cana-2345	206	12	to	to	PART
cana-2345	206	13	better	well	ADV
cana-2345	206	14	navigate	navigate	VERB
cana-2345	206	15	the	the	DET
cana-2345	206	16	complexity	complexity	NOUN
cana-2345	206	17	of	of	ADP
cana-2345	206	18	mixed	mixed	ADJ
cana-2345	206	19	sentiments	sentiment	NOUN
cana-2345	206	20	and	and	CCONJ
cana-2345	206	21	subtle	subtle	ADJ
cana-2345	206	22	language	language	NOUN
cana-2345	206	23	cues	cue	NOUN
cana-2345	206	24	.	.	PUNCT
cana-2345	207	1	however	however	ADV
cana-2345	207	2	,	,	PUNCT
cana-2345	207	3	further	far	ADV
cana-2345	207	4	fine	fine	NOUN
cana-2345	207	5	-	-	PUNCT
cana-2345	207	6	tuning	tuning	NOUN
cana-2345	207	7	of	of	ADP
cana-2345	207	8	the	the	DET
cana-2345	207	9	attention	attention	NOUN
cana-2345	207	10	weights	weight	NOUN
cana-2345	207	11	and	and	CCONJ
cana-2345	207	12	incorporating	incorporate	VERB
cana-2345	207	13	external	external	ADJ
cana-2345	207	14	sentiment	sentiment	NOUN
cana-2345	207	15	lexicons	lexicon	NOUN
cana-2345	207	16	could	could	AUX
cana-2345	207	17	potentially	potentially	ADV
cana-2345	207	18	enhance	enhance	VERB
cana-2345	207	19	npsc	npsc	PROPN
cana-2345	207	20	’s	’s	PART
cana-2345	207	21	ability	ability	NOUN
cana-2345	207	22	to	to	PART
cana-2345	207	23	handle	handle	VERB
cana-2345	207	24	these	these	DET
cana-2345	207	25	challenging	challenging	ADJ
cana-2345	207	26	cases	case	NOUN
cana-2345	207	27	even	even	ADV
cana-2345	207	28	more	more	ADV
cana-2345	207	29	effectively	effectively	ADV
cana-2345	207	30	.	.	PUNCT
cana-2345	208	1	3.2	3.2	NUM
cana-2345	208	2	result	result	NOUN
cana-2345	208	3	discussion	discussion	NOUN
cana-2345	208	4	the	the	DET
cana-2345	208	5	experimental	experimental	ADJ
cana-2345	208	6	results	result	NOUN
cana-2345	208	7	confirmed	confirm	VERB
cana-2345	208	8	that	that	SCONJ
cana-2345	208	9	npsc	npsc	VERB
cana-2345	208	10	significantly	significantly	ADV
cana-2345	208	11	outperforms	outperform	VERB
cana-2345	208	12	contemporary	contemporary	ADJ
cana-2345	208	13	models	model	NOUN
cana-2345	208	14	in	in	ADP
cana-2345	208	15	sentiment	sentiment	NOUN
cana-2345	208	16	classification	classification	NOUN
cana-2345	208	17	tasks	task	NOUN
cana-2345	208	18	.	.	PUNCT
cana-2345	209	1	the	the	DET
cana-2345	209	2	nonlinear	nonlinear	ADJ
cana-2345	209	3	approach	approach	NOUN
cana-2345	209	4	adopted	adopt	VERB
cana-2345	209	5	by	by	ADP
cana-2345	209	6	npsc	npsc	NOUN
cana-2345	209	7	,	,	PUNCT
cana-2345	209	8	integrating	integrate	VERB
cana-2345	209	9	advanced	advanced	ADJ
cana-2345	209	10	components	component	NOUN
cana-2345	209	11	like	like	ADP
cana-2345	209	12	bert	bert	PROPN
cana-2345	209	13	,	,	PUNCT
cana-2345	209	14	bilstm	bilstm	NOUN
cana-2345	209	15	,	,	PUNCT
cana-2345	209	16	cnn	cnn	PROPN
cana-2345	209	17	,	,	PUNCT
cana-2345	209	18	and	and	CCONJ
cana-2345	209	19	attention	attention	NOUN
cana-2345	209	20	,	,	PUNCT
cana-2345	209	21	provides	provide	VERB
cana-2345	209	22	a	a	DET
cana-2345	209	23	clear	clear	ADJ
cana-2345	209	24	advantage	advantage	NOUN
cana-2345	209	25	in	in	ADP
cana-2345	209	26	accurately	accurately	ADV
cana-2345	209	27	identifying	identify	VERB
cana-2345	209	28	sentiment	sentiment	NOUN
cana-2345	209	29	nuances	nuance	NOUN
cana-2345	209	30	.	.	PUNCT
cana-2345	210	1	clm	clm	PROPN
cana-2345	211	1	[	[	X
cana-2345	211	2	16	16	NUM
cana-2345	211	3	]	]	PUNCT
cana-2345	211	4	,	,	PUNCT
cana-2345	211	5	although	although	SCONJ
cana-2345	211	6	performing	perform	VERB
cana-2345	211	7	better	well	ADJ
cana-2345	211	8	than	than	ADP
cana-2345	211	9	rmdl	rmdl	NOUN
cana-2345	211	10	[	[	X
cana-2345	211	11	11	11	NUM
cana-2345	211	12	]	]	PUNCT
cana-2345	211	13	,	,	PUNCT
cana-2345	211	14	lacked	lack	VERB
cana-2345	211	15	the	the	DET
cana-2345	211	16	precision	precision	NOUN
cana-2345	211	17	and	and	CCONJ
cana-2345	211	18	adaptability	adaptability	NOUN
cana-2345	211	19	demonstrated	demonstrate	VERB
cana-2345	211	20	by	by	ADP
cana-2345	211	21	npsc	npsc	NOUN
cana-2345	211	22	,	,	PUNCT
cana-2345	211	23	highlighting	highlight	VERB
cana-2345	211	24	the	the	DET
cana-2345	211	25	importance	importance	NOUN
cana-2345	211	26	of	of	ADP
cana-2345	211	27	a	a	DET
cana-2345	211	28	multi	multi	ADJ
cana-2345	211	29	-	-	ADJ
cana-2345	211	30	layered	layered	ADJ
cana-2345	211	31	and	and	CCONJ
cana-2345	211	32	nonlinear	nonlinear	ADJ
cana-2345	211	33	feature	feature	NOUN
cana-2345	211	34	extraction	extraction	NOUN
cana-2345	211	35	strategy	strategy	NOUN
cana-2345	211	36	in	in	ADP
cana-2345	211	37	sentiment	sentiment	NOUN
cana-2345	211	38	analysis	analysis	NOUN
cana-2345	211	39	.	.	PUNCT
cana-2345	212	1	figure	figure	NOUN
cana-2345	212	2	2	2	NUM
cana-2345	212	3	:	:	PUNCT
cana-2345	212	4	performance	performance	NOUN
cana-2345	212	5	metrics	metric	NOUN
cana-2345	212	6	of	of	ADP
cana-2345	212	7	npsc	npsc	PROPN
cana-2345	212	8	vs.	vs.	ADP
cana-2345	212	9	contemporary	contemporary	ADJ
cana-2345	212	10	models	model	NOUN
cana-2345	212	11	this	this	DET
cana-2345	212	12	bar	bar	NOUN
cana-2345	212	13	graph	graph	NOUN
cana-2345	212	14	presented	present	VERB
cana-2345	212	15	in	in	ADP
cana-2345	212	16	figure	figure	NOUN
cana-2345	212	17	2	2	NUM
cana-2345	212	18	compares	compare	VERB
cana-2345	212	19	the	the	DET
cana-2345	212	20	performance	performance	NOUN
cana-2345	212	21	of	of	ADP
cana-2345	212	22	the	the	DET
cana-2345	212	23	npsc	npsc	PROPN
cana-2345	212	24	,	,	PUNCT
cana-2345	212	25	clm	clm	NOUN
cana-2345	212	26	,	,	PUNCT
cana-2345	212	27	and	and	CCONJ
cana-2345	212	28	rmdl	rmdl	NOUN
cana-2345	212	29	models	model	NOUN
cana-2345	212	30	across	across	ADP
cana-2345	212	31	four	four	NUM
cana-2345	212	32	key	key	ADJ
cana-2345	212	33	metrics	metric	NOUN
cana-2345	212	34	:	:	PUNCT
cana-2345	212	35	accuracy	accuracy	NOUN
cana-2345	212	36	,	,	PUNCT
cana-2345	212	37	precision	precision	NOUN
cana-2345	212	38	,	,	PUNCT
cana-2345	212	39	recall	recall	NOUN
cana-2345	212	40	,	,	PUNCT
cana-2345	212	41	and	and	CCONJ
cana-2345	212	42	f1	f1	NOUN
cana-2345	212	43	-	-	PUNCT
cana-2345	212	44	score	score	NOUN
cana-2345	212	45	.	.	PUNCT
cana-2345	213	1	the	the	DET
cana-2345	213	2	npsc	npsc	PROPN
cana-2345	213	3	model	model	NOUN
cana-2345	213	4	demonstrates	demonstrate	VERB
cana-2345	213	5	superior	superior	ADJ
cana-2345	213	6	performance	performance	NOUN
cana-2345	213	7	in	in	ADP
cana-2345	213	8	all	all	DET
cana-2345	213	9	metrics	metric	NOUN
cana-2345	213	10	,	,	PUNCT
cana-2345	213	11	showcasing	showcase	VERB
cana-2345	213	12	its	its	PRON
cana-2345	213	13	advanced	advanced	ADJ
cana-2345	213	14	capability	capability	NOUN
cana-2345	213	15	in	in	ADP
cana-2345	213	16	precise	precise	ADJ
cana-2345	213	17	sentiment	sentiment	NOUN
cana-2345	213	18	communications	communication	NOUN
cana-2345	213	19	on	on	ADP
cana-2345	213	20	applied	apply	VERB
cana-2345	213	21	nonlinear	nonlinear	ADJ
cana-2345	213	22	analysis	analysis	NOUN
cana-2345	213	23	issn	issn	NOUN
cana-2345	213	24	:	:	PUNCT
cana-2345	213	25	1074	1074	NUM
cana-2345	213	26	-	-	PUNCT
cana-2345	213	27	133x	133x	NUM
cana-2345	213	28	vol	vol	NOUN
cana-2345	213	29	32	32	NUM
cana-2345	213	30	no	no	NOUN
cana-2345	213	31	.	.	PUNCT
cana-2345	214	1	1s	1s	NUM
cana-2345	214	2	(	(	PUNCT
cana-2345	214	3	2025	2025	NUM
cana-2345	214	4	)	)	PUNCT
cana-2345	214	5	582	582	NUM
cana-2345	214	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2345	214	7	classification	classification	NOUN
cana-2345	214	8	.	.	PUNCT
cana-2345	215	1	clm	clm	NOUN
cana-2345	215	2	performs	perform	VERB
cana-2345	215	3	better	well	ADJ
cana-2345	215	4	than	than	ADP
cana-2345	215	5	rmdl	rmdl	NOUN
cana-2345	215	6	,	,	PUNCT
cana-2345	215	7	but	but	CCONJ
cana-2345	215	8	both	both	PRON
cana-2345	215	9	are	be	AUX
cana-2345	215	10	outperformed	outperform	VERB
cana-2345	215	11	by	by	ADP
cana-2345	215	12	npsc	npsc	PROPN
cana-2345	215	13	,	,	PUNCT
cana-2345	215	14	highlighting	highlight	VERB
cana-2345	215	15	the	the	DET
cana-2345	215	16	effectiveness	effectiveness	NOUN
cana-2345	215	17	of	of	ADP
cana-2345	215	18	the	the	DET
cana-2345	215	19	nonlinear	nonlinear	ADJ
cana-2345	215	20	approach	approach	NOUN
cana-2345	215	21	and	and	CCONJ
cana-2345	215	22	attention	attention	NOUN
cana-2345	215	23	mechanisms	mechanism	NOUN
cana-2345	215	24	in	in	ADP
cana-2345	215	25	npsc	npsc	PROPN
cana-2345	215	26	.	.	PUNCT
cana-2345	216	1	figure	figure	VERB
cana-2345	216	2	3	3	NUM
cana-2345	216	3	:	:	PUNCT
cana-2345	216	4	ablation	ablation	NOUN
cana-2345	216	5	study	study	NOUN
cana-2345	216	6	results	result	NOUN
cana-2345	216	7	of	of	ADP
cana-2345	216	8	npsc	npsc	NOUN
cana-2345	216	9	this	this	DET
cana-2345	216	10	line	line	NOUN
cana-2345	216	11	graph	graph	NOUN
cana-2345	216	12	presents	present	NOUN
cana-2345	216	13	in	in	ADP
cana-2345	216	14	figure	figure	NOUN
cana-2345	216	15	3	3	NUM
cana-2345	216	16	the	the	DET
cana-2345	216	17	results	result	NOUN
cana-2345	216	18	of	of	ADP
cana-2345	216	19	the	the	DET
cana-2345	216	20	ablation	ablation	NOUN
cana-2345	216	21	study	study	NOUN
cana-2345	216	22	conducted	conduct	VERB
cana-2345	216	23	on	on	ADP
cana-2345	216	24	the	the	DET
cana-2345	216	25	npsc	npsc	NOUN
cana-2345	216	26	model	model	NOUN
cana-2345	216	27	by	by	ADP
cana-2345	216	28	removing	remove	VERB
cana-2345	216	29	critical	critical	ADJ
cana-2345	216	30	components	component	NOUN
cana-2345	216	31	:	:	PUNCT
cana-2345	217	1	cnn	cnn	PROPN
cana-2345	217	2	,	,	PUNCT
cana-2345	217	3	attention	attention	NOUN
cana-2345	217	4	mechanism	mechanism	NOUN
cana-2345	217	5	,	,	PUNCT
cana-2345	217	6	and	and	CCONJ
cana-2345	217	7	nonlinear	nonlinear	ADJ
cana-2345	217	8	fusion	fusion	NOUN
cana-2345	217	9	layer	layer	NOUN
cana-2345	217	10	.	.	PUNCT
cana-2345	218	1	the	the	DET
cana-2345	218	2	study	study	NOUN
cana-2345	218	3	shows	show	VERB
cana-2345	218	4	a	a	DET
cana-2345	218	5	significant	significant	ADJ
cana-2345	218	6	decline	decline	NOUN
cana-2345	218	7	in	in	ADP
cana-2345	218	8	performance	performance	NOUN
cana-2345	218	9	when	when	SCONJ
cana-2345	218	10	these	these	DET
cana-2345	218	11	components	component	NOUN
cana-2345	218	12	are	be	AUX
cana-2345	218	13	excluded	exclude	VERB
cana-2345	218	14	,	,	PUNCT
cana-2345	218	15	especially	especially	ADV
cana-2345	218	16	the	the	DET
cana-2345	218	17	nonlinear	nonlinear	ADJ
cana-2345	218	18	fusion	fusion	NOUN
cana-2345	218	19	layer	layer	NOUN
cana-2345	218	20	,	,	PUNCT
cana-2345	218	21	confirming	confirm	VERB
cana-2345	218	22	their	their	PRON
cana-2345	218	23	essential	essential	ADJ
cana-2345	218	24	roles	role	NOUN
cana-2345	218	25	in	in	ADP
cana-2345	218	26	achieving	achieve	VERB
cana-2345	218	27	high	high	ADJ
cana-2345	218	28	accuracy	accuracy	NOUN
cana-2345	218	29	and	and	CCONJ
cana-2345	218	30	precision	precision	NOUN
cana-2345	218	31	in	in	ADP
cana-2345	218	32	sentiment	sentiment	NOUN
cana-2345	218	33	classification	classification	NOUN
cana-2345	218	34	.	.	PUNCT
cana-2345	219	1	figure	figure	VERB
cana-2345	219	2	4	4	NUM
cana-2345	219	3	:	:	PUNCT
cana-2345	219	4	error	error	NOUN
cana-2345	219	5	analysis	analysis	NOUN
cana-2345	219	6	of	of	ADP
cana-2345	219	7	npsc	npsc	NOUN
cana-2345	219	8	and	and	CCONJ
cana-2345	219	9	baseline	baseline	NOUN
cana-2345	219	10	models	model	NOUN
cana-2345	219	11	this	this	DET
cana-2345	219	12	bar	bar	NOUN
cana-2345	219	13	graph	graph	NOUN
cana-2345	219	14	shown	show	VERB
cana-2345	219	15	in	in	ADP
cana-2345	219	16	figure	figure	NOUN
cana-2345	219	17	4	4	NUM
cana-2345	219	18	illustrates	illustrate	VERB
cana-2345	219	19	the	the	DET
cana-2345	219	20	misclassification	misclassification	NOUN
cana-2345	219	21	rates	rate	NOUN
cana-2345	219	22	for	for	ADP
cana-2345	219	23	npsc	npsc	NOUN
cana-2345	219	24	,	,	PUNCT
cana-2345	219	25	clm	clm	X
cana-2345	220	1	[	[	X
cana-2345	220	2	16	16	NUM
cana-2345	220	3	]	]	PUNCT
cana-2345	220	4	,	,	PUNCT
cana-2345	220	5	and	and	CCONJ
cana-2345	220	6	rmdl	rmdl	NOUN
cana-2345	220	7	[	[	X
cana-2345	220	8	11	11	NUM
cana-2345	220	9	]	]	PUNCT
cana-2345	220	10	across	across	ADP
cana-2345	220	11	different	different	ADJ
cana-2345	220	12	error	error	NOUN
cana-2345	220	13	types	type	NOUN
cana-2345	220	14	,	,	PUNCT
cana-2345	220	15	including	include	VERB
cana-2345	220	16	mixed	mixed	ADJ
cana-2345	220	17	sentiments	sentiment	NOUN
cana-2345	220	18	,	,	PUNCT
cana-2345	220	19	neutral	neutral	ADJ
cana-2345	220	20	or	or	CCONJ
cana-2345	220	21	subtle	subtle	ADJ
cana-2345	220	22	sentiments	sentiment	NOUN
cana-2345	220	23	,	,	PUNCT
cana-2345	220	24	and	and	CCONJ
cana-2345	220	25	complex	complex	ADJ
cana-2345	220	26	language	language	NOUN
cana-2345	220	27	usage	usage	NOUN
cana-2345	220	28	.	.	PUNCT
cana-2345	221	1	npsc	npsc	PROPN
cana-2345	221	2	shows	show	VERB
cana-2345	221	3	the	the	DET
cana-2345	221	4	lowest	low	ADJ
cana-2345	221	5	error	error	NOUN
cana-2345	221	6	rates	rate	NOUN
cana-2345	221	7	in	in	ADP
cana-2345	221	8	all	all	DET
cana-2345	221	9	categories	category	NOUN
cana-2345	221	10	,	,	PUNCT
cana-2345	221	11	demonstrating	demonstrate	VERB
cana-2345	221	12	its	its	PRON
cana-2345	221	13	superior	superior	ADJ
cana-2345	221	14	ability	ability	NOUN
cana-2345	221	15	to	to	PART
cana-2345	221	16	handle	handle	VERB
cana-2345	221	17	ambiguous	ambiguous	ADJ
cana-2345	221	18	and	and	CCONJ
cana-2345	221	19	nuanced	nuanced	ADJ
cana-2345	221	20	text	text	NOUN
cana-2345	221	21	,	,	PUNCT
cana-2345	221	22	compared	compare	VERB
cana-2345	221	23	to	to	ADP
cana-2345	221	24	the	the	DET
cana-2345	221	25	higher	high	ADJ
cana-2345	221	26	misclassification	misclassification	NOUN
cana-2345	221	27	rates	rate	NOUN
cana-2345	221	28	observed	observe	VERB
cana-2345	221	29	in	in	ADP
cana-2345	221	30	clm	clm	NOUN
cana-2345	221	31	and	and	CCONJ
cana-2345	221	32	rmdl	rmdl	NOUN
cana-2345	221	33	.	.	PUNCT
cana-2345	222	1	the	the	DET
cana-2345	222	2	experimental	experimental	ADJ
cana-2345	222	3	study	study	NOUN
cana-2345	222	4	validates	validate	VERB
cana-2345	222	5	the	the	DET
cana-2345	222	6	effectiveness	effectiveness	NOUN
cana-2345	222	7	of	of	ADP
cana-2345	222	8	the	the	DET
cana-2345	222	9	npsc	npsc	NOUN
cana-2345	222	10	framework	framework	NOUN
cana-2345	222	11	,	,	PUNCT
cana-2345	222	12	showcasing	showcase	VERB
cana-2345	222	13	its	its	PRON
cana-2345	222	14	ability	ability	NOUN
cana-2345	222	15	to	to	PART
cana-2345	222	16	outperform	outperform	VERB
cana-2345	222	17	established	establish	VERB
cana-2345	222	18	models	model	NOUN
cana-2345	222	19	like	like	ADP
cana-2345	222	20	rmdl	rmdl	NOUN
cana-2345	222	21	and	and	CCONJ
cana-2345	222	22	clm	clm	NOUN
cana-2345	222	23	.	.	PUNCT
cana-2345	223	1	the	the	DET
cana-2345	223	2	ablation	ablation	NOUN
cana-2345	223	3	studies	study	NOUN
cana-2345	223	4	and	and	CCONJ
cana-2345	223	5	error	error	NOUN
cana-2345	223	6	analysis	analysis	NOUN
cana-2345	223	7	further	far	ADV
cana-2345	223	8	confirmed	confirm	VERB
cana-2345	223	9	the	the	DET
cana-2345	223	10	critical	critical	ADJ
cana-2345	223	11	roles	role	NOUN
cana-2345	223	12	of	of	ADP
cana-2345	223	13	each	each	DET
cana-2345	223	14	component	component	NOUN
cana-2345	223	15	within	within	ADP
cana-2345	223	16	npsc	npsc	NOUN
cana-2345	223	17	,	,	PUNCT
cana-2345	223	18	emphasizing	emphasize	VERB
cana-2345	223	19	its	its	PRON
cana-2345	223	20	superior	superior	ADJ
cana-2345	223	21	capability	capability	NOUN
cana-2345	223	22	in	in	ADP
cana-2345	223	23	precise	precise	ADJ
cana-2345	223	24	sentiment	sentiment	NOUN
cana-2345	223	25	classification	classification	NOUN
cana-2345	223	26	.	.	PUNCT
cana-2345	224	1	the	the	DET
cana-2345	224	2	findings	finding	NOUN
cana-2345	224	3	highlight	highlight	VERB
cana-2345	224	4	the	the	DET
cana-2345	224	5	strengths	strength	NOUN
cana-2345	224	6	of	of	ADP
cana-2345	224	7	a	a	DET
cana-2345	224	8	nonlinear	nonlinear	ADJ
cana-2345	224	9	approach	approach	NOUN
cana-2345	224	10	in	in	ADP
cana-2345	224	11	sentiment	sentiment	NOUN
cana-2345	224	12	classification	classification	NOUN
cana-2345	224	13	,	,	PUNCT
cana-2345	224	14	making	make	VERB
cana-2345	224	15	npsc	npsc	VERB
cana-2345	224	16	a	a	DET
cana-2345	224	17	robust	robust	ADJ
cana-2345	224	18	and	and	CCONJ
cana-2345	224	19	precise	precise	ADJ
cana-2345	224	20	tool	tool	NOUN
cana-2345	224	21	for	for	ADP
cana-2345	224	22	analyzing	analyze	VERB
cana-2345	224	23	complex	complex	ADJ
cana-2345	224	24	textual	textual	ADJ
cana-2345	224	25	data	datum	NOUN
cana-2345	224	26	.	.	PUNCT
cana-2345	225	1	communications	communication	NOUN
cana-2345	225	2	on	on	ADP
cana-2345	225	3	applied	apply	VERB
cana-2345	225	4	nonlinear	nonlinear	ADJ
cana-2345	225	5	analysis	analysis	NOUN
cana-2345	225	6	issn	issn	NOUN
cana-2345	225	7	:	:	PUNCT
cana-2345	225	8	1074	1074	NUM
cana-2345	225	9	-	-	PUNCT
cana-2345	225	10	133x	133x	NUM
cana-2345	225	11	vol	vol	NOUN
cana-2345	225	12	32	32	NUM
cana-2345	225	13	no	no	NOUN
cana-2345	225	14	.	.	PUNCT
cana-2345	226	1	1s	1s	NUM
cana-2345	226	2	(	(	PUNCT
cana-2345	226	3	2025	2025	NUM
cana-2345	226	4	)	)	PUNCT
cana-2345	226	5	583	583	NUM
cana-2345	226	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2345	226	7	4	4	NUM
cana-2345	226	8	conclusion	conclusion	NOUN
cana-2345	226	9	the	the	DET
cana-2345	226	10	study	study	NOUN
cana-2345	226	11	focused	focus	VERB
cana-2345	226	12	on	on	ADP
cana-2345	226	13	building	build	VERB
cana-2345	226	14	the	the	DET
cana-2345	226	15	nonlinear	nonlinear	ADJ
cana-2345	226	16	deep	deep	ADJ
cana-2345	226	17	learning	learning	NOUN
cana-2345	226	18	framework	framework	NOUN
cana-2345	226	19	for	for	ADP
cana-2345	226	20	precise	precise	ADJ
cana-2345	226	21	sentiment	sentiment	NOUN
cana-2345	226	22	classification	classification	NOUN
cana-2345	226	23	(	(	PUNCT
cana-2345	226	24	npsc	npsc	NOUN
cana-2345	226	25	)	)	PUNCT
cana-2345	226	26	to	to	PART
cana-2345	226	27	improve	improve	VERB
cana-2345	226	28	how	how	SCONJ
cana-2345	226	29	sentiment	sentiment	NOUN
cana-2345	226	30	is	be	AUX
cana-2345	226	31	analyzed	analyze	VERB
cana-2345	226	32	in	in	ADP
cana-2345	226	33	text	text	NOUN
cana-2345	226	34	data	datum	NOUN
cana-2345	226	35	.	.	PUNCT
cana-2345	227	1	the	the	DET
cana-2345	227	2	model	model	NOUN
cana-2345	227	3	combined	combine	VERB
cana-2345	227	4	bert	bert	PROPN
cana-2345	227	5	,	,	PUNCT
cana-2345	227	6	bilstm	bilstm	NOUN
cana-2345	227	7	,	,	PUNCT
cana-2345	227	8	cnn	cnn	PROPN
cana-2345	227	9	,	,	PUNCT
cana-2345	227	10	and	and	CCONJ
cana-2345	227	11	attention	attention	NOUN
cana-2345	227	12	layers	layer	NOUN
cana-2345	227	13	to	to	PART
cana-2345	227	14	capture	capture	VERB
cana-2345	227	15	complex	complex	ADJ
cana-2345	227	16	sentiment	sentiment	NOUN
cana-2345	227	17	patterns	pattern	NOUN
cana-2345	227	18	,	,	PUNCT
cana-2345	227	19	showing	show	VERB
cana-2345	227	20	strong	strong	ADJ
cana-2345	227	21	results	result	NOUN
cana-2345	227	22	in	in	ADP
cana-2345	227	23	key	key	ADJ
cana-2345	227	24	areas	area	NOUN
cana-2345	227	25	like	like	ADP
cana-2345	227	26	accuracy	accuracy	NOUN
cana-2345	227	27	and	and	CCONJ
cana-2345	227	28	recall	recall	NOUN
cana-2345	227	29	.	.	PUNCT
cana-2345	228	1	this	this	DET
cana-2345	228	2	combination	combination	NOUN
cana-2345	228	3	allowed	allow	VERB
cana-2345	228	4	the	the	DET
cana-2345	228	5	model	model	NOUN
cana-2345	228	6	to	to	PART
cana-2345	228	7	pick	pick	VERB
cana-2345	228	8	up	up	ADP
cana-2345	228	9	on	on	ADP
cana-2345	228	10	subtle	subtle	ADJ
cana-2345	228	11	emotions	emotion	NOUN
cana-2345	228	12	and	and	CCONJ
cana-2345	228	13	context	context	NOUN
cana-2345	228	14	within	within	ADP
cana-2345	228	15	reviews	review	NOUN
cana-2345	228	16	,	,	PUNCT
cana-2345	228	17	making	make	VERB
cana-2345	228	18	the	the	DET
cana-2345	228	19	predictions	prediction	NOUN
cana-2345	228	20	sharper	sharp	ADJ
cana-2345	228	21	and	and	CCONJ
cana-2345	228	22	more	more	ADV
cana-2345	228	23	reliable	reliable	ADJ
cana-2345	228	24	.	.	PUNCT
cana-2345	229	1	npsc	npsc	PROPN
cana-2345	229	2	’s	’s	PART
cana-2345	229	3	design	design	NOUN
cana-2345	229	4	moves	move	NOUN
cana-2345	229	5	beyond	beyond	ADP
cana-2345	229	6	simple	simple	ADJ
cana-2345	229	7	models	model	NOUN
cana-2345	229	8	by	by	ADP
cana-2345	229	9	layering	layer	VERB
cana-2345	229	10	different	different	ADJ
cana-2345	229	11	techniques	technique	NOUN
cana-2345	229	12	that	that	PRON
cana-2345	229	13	work	work	VERB
cana-2345	229	14	together	together	ADV
cana-2345	229	15	,	,	PUNCT
cana-2345	229	16	showing	show	VERB
cana-2345	229	17	that	that	SCONJ
cana-2345	229	18	blending	blend	VERB
cana-2345	229	19	multiple	multiple	ADJ
cana-2345	229	20	approaches	approach	NOUN
cana-2345	229	21	captures	capture	VERB
cana-2345	229	22	more	more	ADV
cana-2345	229	23	hidden	hide	VERB
cana-2345	229	24	sentiment	sentiment	NOUN
cana-2345	229	25	cues	cue	VERB
cana-2345	229	26	.	.	PUNCT
cana-2345	230	1	the	the	DET
cana-2345	230	2	findings	finding	NOUN
cana-2345	230	3	point	point	VERB
cana-2345	230	4	to	to	ADP
cana-2345	230	5	a	a	DET
cana-2345	230	6	clear	clear	ADJ
cana-2345	230	7	advantage	advantage	NOUN
cana-2345	230	8	of	of	ADP
cana-2345	230	9	using	use	VERB
cana-2345	230	10	nonlinear	nonlinear	ADJ
cana-2345	230	11	methods	method	NOUN
cana-2345	230	12	in	in	ADP
cana-2345	230	13	sentiment	sentiment	NOUN
cana-2345	230	14	analysis	analysis	NOUN
cana-2345	230	15	,	,	PUNCT
cana-2345	230	16	especially	especially	ADV
cana-2345	230	17	in	in	ADP
cana-2345	230	18	texts	text	NOUN
cana-2345	230	19	with	with	ADP
cana-2345	230	20	mixed	mixed	ADJ
cana-2345	230	21	or	or	CCONJ
cana-2345	230	22	unclear	unclear	ADJ
cana-2345	230	23	opinions	opinion	NOUN
cana-2345	230	24	.	.	PUNCT
cana-2345	231	1	however	however	ADV
cana-2345	231	2	,	,	PUNCT
cana-2345	231	3	the	the	DET
cana-2345	231	4	model	model	NOUN
cana-2345	231	5	was	be	AUX
cana-2345	231	6	tested	test	VERB
cana-2345	231	7	on	on	ADP
cana-2345	231	8	one	one	NUM
cana-2345	231	9	dataset	dataset	NOUN
cana-2345	231	10	,	,	PUNCT
cana-2345	231	11	and	and	CCONJ
cana-2345	231	12	its	its	PRON
cana-2345	231	13	performance	performance	NOUN
cana-2345	231	14	could	could	AUX
cana-2345	231	15	vary	vary	VERB
cana-2345	231	16	with	with	ADP
cana-2345	231	17	other	other	ADJ
cana-2345	231	18	text	text	NOUN
cana-2345	231	19	types	type	NOUN
cana-2345	231	20	.	.	PUNCT
cana-2345	232	1	future	future	ADJ
cana-2345	232	2	work	work	NOUN
cana-2345	232	3	could	could	AUX
cana-2345	232	4	look	look	VERB
cana-2345	232	5	at	at	ADP
cana-2345	232	6	testing	testing	NOUN
cana-2345	232	7	npsc	npsc	VERB
cana-2345	232	8	across	across	ADP
cana-2345	232	9	different	different	ADJ
cana-2345	232	10	datasets	dataset	NOUN
cana-2345	232	11	,	,	PUNCT
cana-2345	232	12	tweaking	tweak	VERB
cana-2345	232	13	each	each	DET
cana-2345	232	14	layer	layer	NOUN
cana-2345	232	15	,	,	PUNCT
cana-2345	232	16	or	or	CCONJ
cana-2345	232	17	adding	add	VERB
cana-2345	232	18	new	new	ADJ
cana-2345	232	19	features	feature	NOUN
cana-2345	232	20	to	to	PART
cana-2345	232	21	deepen	deepen	VERB
cana-2345	232	22	the	the	DET
cana-2345	232	23	analysis	analysis	NOUN
cana-2345	232	24	.	.	PUNCT
cana-2345	233	1	this	this	DET
cana-2345	233	2	study	study	NOUN
cana-2345	233	3	shows	show	VERB
cana-2345	233	4	that	that	SCONJ
cana-2345	233	5	npsc	npsc	NOUN
cana-2345	233	6	is	be	AUX
cana-2345	233	7	not	not	PART
cana-2345	233	8	just	just	ADV
cana-2345	233	9	another	another	DET
cana-2345	233	10	sentiment	sentiment	NOUN
cana-2345	233	11	analysis	analysis	NOUN
cana-2345	233	12	tool	tool	NOUN
cana-2345	233	13	;	;	PUNCT
cana-2345	233	14	it	it	PRON
cana-2345	233	15	’s	’	VERB
cana-2345	233	16	a	a	DET
cana-2345	233	17	step	step	NOUN
cana-2345	233	18	towards	towards	ADP
cana-2345	233	19	better	well	ADJ
cana-2345	233	20	understanding	understand	VERB
cana-2345	233	21	complex	complex	ADJ
cana-2345	233	22	emotions	emotion	NOUN
cana-2345	233	23	in	in	ADP
cana-2345	233	24	text	text	NOUN
cana-2345	233	25	.	.	PUNCT
cana-2345	234	1	the	the	DET
cana-2345	234	2	approach	approach	NOUN
cana-2345	234	3	lays	lay	VERB
cana-2345	234	4	the	the	DET
cana-2345	234	5	groundwork	groundwork	NOUN
cana-2345	234	6	for	for	ADP
cana-2345	234	7	more	more	ADV
cana-2345	234	8	precise	precise	ADJ
cana-2345	234	9	and	and	CCONJ
cana-2345	234	10	adaptable	adaptable	ADJ
cana-2345	234	11	sentiment	sentiment	NOUN
cana-2345	234	12	models	model	NOUN
cana-2345	234	13	,	,	PUNCT
cana-2345	234	14	making	make	VERB
cana-2345	234	15	it	it	PRON
cana-2345	234	16	a	a	DET
cana-2345	234	17	valuable	valuable	ADJ
cana-2345	234	18	addition	addition	NOUN
cana-2345	234	19	to	to	ADP
cana-2345	234	20	fields	field	NOUN
cana-2345	234	21	where	where	SCONJ
cana-2345	234	22	accurate	accurate	ADJ
cana-2345	234	23	emotion	emotion	NOUN
cana-2345	234	24	detection	detection	NOUN
cana-2345	234	25	matters	matter	NOUN
cana-2345	234	26	.	.	PUNCT
cana-2345	235	1	references	reference	NOUN
cana-2345	235	2	[	[	X
cana-2345	235	3	1	1	NUM
cana-2345	235	4	]	]	X
cana-2345	235	5	ain	ain	PROPN
cana-2345	235	6	,	,	PUNCT
cana-2345	235	7	qurat	qurat	PROPN
cana-2345	235	8	tul	tul	PROPN
cana-2345	235	9	,	,	PUNCT
cana-2345	235	10	mubashir	mubashir	PROPN
cana-2345	235	11	ali	ali	PROPN
cana-2345	235	12	,	,	PUNCT
cana-2345	235	13	amna	amna	PROPN
cana-2345	235	14	riaz	riaz	PROPN
cana-2345	235	15	,	,	PUNCT
cana-2345	235	16	amna	amna	PROPN
cana-2345	235	17	noureen	noureen	PROPN
cana-2345	235	18	,	,	PUNCT
cana-2345	235	19	muhammad	muhammad	PROPN
cana-2345	235	20	kamran	kamran	PROPN
cana-2345	235	21	,	,	PUNCT
cana-2345	235	22	babar	babar	PROPN
cana-2345	235	23	hayat	hayat	PROPN
cana-2345	235	24	,	,	PUNCT
cana-2345	235	25	and	and	CCONJ
cana-2345	235	26	a.	a.	NOUN
cana-2345	235	27	rehman	rehman	PROPN
cana-2345	235	28	.	.	PUNCT
cana-2345	236	1	"	"	PUNCT
cana-2345	236	2	sentiment	sentiment	NOUN
cana-2345	236	3	analysis	analysis	NOUN
cana-2345	236	4	using	use	VERB
cana-2345	236	5	deep	deep	ADJ
cana-2345	236	6	learning	learning	NOUN
cana-2345	236	7	techniques	technique	NOUN
cana-2345	236	8	:	:	PUNCT
cana-2345	236	9	a	a	DET
cana-2345	236	10	review	review	NOUN
cana-2345	236	11	.	.	PUNCT
cana-2345	236	12	"	"	PUNCT
cana-2345	237	1	international	international	ADJ
cana-2345	237	2	journal	journal	NOUN
cana-2345	237	3	of	of	ADP
cana-2345	237	4	advanced	advanced	ADJ
cana-2345	237	5	computer	computer	NOUN
cana-2345	237	6	science	science	NOUN
cana-2345	237	7	and	and	CCONJ
cana-2345	237	8	applications	application	NOUN
cana-2345	237	9	8	8	NUM
cana-2345	237	10	,	,	PUNCT
cana-2345	237	11	no	no	INTJ
cana-2345	237	12	.	.	NOUN
cana-2345	237	13	6	6	NUM
cana-2345	237	14	(	(	PUNCT
cana-2345	237	15	2017	2017	NUM
cana-2345	237	16	)	)	PUNCT
cana-2345	237	17	.	.	PUNCT
cana-2345	238	1	[	[	X
cana-2345	238	2	2	2	NUM
cana-2345	238	3	]	]	X
cana-2345	238	4	wang	wang	PROPN
cana-2345	238	5	,	,	PUNCT
cana-2345	238	6	wenliang	wenliang	PROPN
cana-2345	238	7	.	.	PUNCT
cana-2345	239	1	"	"	PUNCT
cana-2345	239	2	text	text	NOUN
cana-2345	239	3	sentiment	sentiment	NOUN
cana-2345	239	4	classification	classification	NOUN
cana-2345	239	5	method	method	NOUN
cana-2345	239	6	based	base	VERB
cana-2345	239	7	on	on	ADP
cana-2345	239	8	bilstm	bilstm	NOUN
cana-2345	239	9	.	.	PUNCT
cana-2345	239	10	"	"	PUNCT
cana-2345	240	1	highlights	highlight	NOUN
cana-2345	240	2	in	in	ADP
cana-2345	240	3	business	business	NOUN
cana-2345	240	4	,	,	PUNCT
cana-2345	240	5	economics	economic	NOUN
cana-2345	240	6	and	and	CCONJ
cana-2345	240	7	management	management	NOUN
cana-2345	240	8	21	21	NUM
cana-2345	240	9	(	(	PUNCT
cana-2345	240	10	2023	2023	NUM
cana-2345	240	11	):	):	PUNCT
cana-2345	240	12	679	679	NUM
cana-2345	240	13	-	-	SYM
cana-2345	240	14	687	687	NUM
cana-2345	240	15	.	.	PUNCT
cana-2345	241	1	[	[	X
cana-2345	241	2	3	3	NUM
cana-2345	241	3	]	]	SYM
cana-2345	241	4	wu	wu	PROPN
cana-2345	241	5	,	,	PUNCT
cana-2345	241	6	yichao	yichao	PROPN
cana-2345	241	7	,	,	PUNCT
cana-2345	241	8	zhengyu	zhengyu	PROPN
cana-2345	241	9	jin	jin	PROPN
cana-2345	241	10	,	,	PUNCT
cana-2345	241	11	chenxi	chenxi	PROPN
cana-2345	241	12	shi	shi	PROPN
cana-2345	241	13	,	,	PUNCT
cana-2345	241	14	penghao	penghao	PROPN
cana-2345	241	15	liang	liang	PROPN
cana-2345	241	16	,	,	PUNCT
cana-2345	241	17	and	and	CCONJ
cana-2345	241	18	tong	tong	PROPN
cana-2345	241	19	zhan	zhan	PROPN
cana-2345	241	20	.	.	PUNCT
cana-2345	242	1	"	"	PUNCT
cana-2345	242	2	research	research	NOUN
cana-2345	242	3	on	on	ADP
cana-2345	242	4	the	the	DET
cana-2345	242	5	application	application	NOUN
cana-2345	242	6	of	of	ADP
cana-2345	242	7	deep	deep	ADJ
cana-2345	242	8	learning	learning	NOUN
cana-2345	242	9	-	-	PUNCT
cana-2345	242	10	based	base	VERB
cana-2345	242	11	bert	bert	PROPN
cana-2345	242	12	model	model	NOUN
cana-2345	242	13	in	in	ADP
cana-2345	242	14	sentiment	sentiment	NOUN
cana-2345	242	15	analysis	analysis	NOUN
cana-2345	242	16	.	.	PUNCT
cana-2345	242	17	"	"	PUNCT
cana-2345	243	1	arxiv	arxiv	PROPN
cana-2345	243	2	preprint	preprint	VERB
cana-2345	243	3	arxiv:2403.08217	arxiv:2403.08217	NOUN
cana-2345	243	4	(	(	PUNCT
cana-2345	243	5	2024	2024	NUM
cana-2345	243	6	)	)	PUNCT
cana-2345	243	7	.	.	PUNCT
cana-2345	244	1	[	[	X
cana-2345	244	2	4	4	X
cana-2345	244	3	]	]	PUNCT
cana-2345	244	4	rose	rise	VERB
cana-2345	244	5	,	,	PUNCT
cana-2345	244	6	j.	j.	PROPN
cana-2345	244	7	dafni	dafni	PROPN
cana-2345	244	8	,	,	PUNCT
cana-2345	244	9	s.	s.	PROPN
cana-2345	244	10	bhuwaneshwaran	bhuwaneshwaran	PROPN
cana-2345	244	11	,	,	PUNCT
cana-2345	244	12	and	and	CCONJ
cana-2345	244	13	s.	s.	PROPN
cana-2345	244	14	dhiyanesh	dhiyanesh	PROPN
cana-2345	244	15	.	.	PUNCT
cana-2345	245	1	"	"	PUNCT
cana-2345	245	2	advanced	advanced	ADJ
cana-2345	245	3	novel	novel	ADJ
cana-2345	245	4	nlp	nlp	NOUN
cana-2345	245	5	approach	approach	NOUN
cana-2345	245	6	for	for	ADP
cana-2345	245	7	emotion	emotion	NOUN
cana-2345	245	8	detection	detection	NOUN
cana-2345	245	9	in	in	ADP
cana-2345	245	10	poetry	poetry	NOUN
cana-2345	245	11	.	.	PUNCT
cana-2345	245	12	"	"	PUNCT
cana-2345	246	1	in	in	ADP
cana-2345	246	2	2024	2024	NUM
cana-2345	246	3	international	international	ADJ
cana-2345	246	4	conference	conference	NOUN
cana-2345	246	5	on	on	ADP
cana-2345	246	6	advances	advance	NOUN
cana-2345	246	7	in	in	ADP
cana-2345	246	8	computing	computing	NOUN
cana-2345	246	9	,	,	PUNCT
cana-2345	246	10	communication	communication	NOUN
cana-2345	246	11	and	and	CCONJ
cana-2345	246	12	applied	apply	VERB
cana-2345	246	13	informatics	informatic	NOUN
cana-2345	246	14	(	(	PUNCT
cana-2345	246	15	accai	accai	PROPN
cana-2345	246	16	)	)	PUNCT
cana-2345	246	17	,	,	PUNCT
cana-2345	246	18	pp	pp	ADJ
cana-2345	246	19	.	.	PUNCT
cana-2345	247	1	1	1	NUM
cana-2345	247	2	-	-	SYM
cana-2345	247	3	6	6	NUM
cana-2345	247	4	.	.	PUNCT
cana-2345	247	5	ieee	ieee	NOUN
cana-2345	247	6	,	,	PUNCT
cana-2345	247	7	2024	2024	NUM
cana-2345	247	8	.	.	PUNCT
cana-2345	248	1	[	[	X
cana-2345	248	2	5	5	NUM
cana-2345	248	3	]	]	X
cana-2345	248	4	liu	liu	PROPN
cana-2345	248	5	,	,	PUNCT
cana-2345	248	6	jingyi	jingyi	PROPN
cana-2345	248	7	,	,	PUNCT
cana-2345	248	8	and	and	CCONJ
cana-2345	248	9	sheng	sheng	PROPN
cana-2345	248	10	li	li	PROPN
cana-2345	248	11	.	.	PUNCT
cana-2345	249	1	"	"	PUNCT
cana-2345	249	2	a	a	DET
cana-2345	249	3	dependency	dependency	NOUN
cana-2345	249	4	-	-	PUNCT
cana-2345	249	5	based	base	VERB
cana-2345	249	6	hybrid	hybrid	ADJ
cana-2345	249	7	deep	deep	ADJ
cana-2345	249	8	learning	learning	NOUN
cana-2345	249	9	framework	framework	NOUN
cana-2345	249	10	for	for	ADP
cana-2345	249	11	target	target	NOUN
cana-2345	249	12	-	-	PUNCT
cana-2345	249	13	dependent	dependent	ADJ
cana-2345	249	14	sentiment	sentiment	NOUN
cana-2345	249	15	classification	classification	NOUN
cana-2345	249	16	.	.	PUNCT
cana-2345	249	17	"	"	PUNCT
cana-2345	249	18	pattern	pattern	NOUN
cana-2345	249	19	recognition	recognition	NOUN
cana-2345	249	20	letters	letter	NOUN
cana-2345	249	21	176	176	NUM
cana-2345	249	22	(	(	PUNCT
cana-2345	249	23	2023	2023	NUM
cana-2345	249	24	):	):	PUNCT
cana-2345	249	25	160	160	NUM
cana-2345	249	26	-	-	SYM
cana-2345	249	27	166	166	NUM
cana-2345	249	28	.	.	PUNCT
cana-2345	250	1	[	[	X
cana-2345	250	2	6	6	NUM
cana-2345	250	3	]	]	X
cana-2345	250	4	kathiravan	kathiravan	PROPN
cana-2345	250	5	,	,	PUNCT
cana-2345	250	6	m.	m.	NOUN
cana-2345	250	7	,	,	PUNCT
cana-2345	250	8	s.	s.	PROPN
cana-2345	250	9	saravanan	saravanan	PROPN
cana-2345	250	10	,	,	PUNCT
cana-2345	250	11	m.	m.	PROPN
cana-2345	250	12	jagadeesh	jagadeesh	PROPN
cana-2345	250	13	,	,	PUNCT
cana-2345	250	14	i.	i.	NOUN
cana-2345	250	15	lakshmi	lakshmi	PROPN
cana-2345	250	16	,	,	PUNCT
cana-2345	250	17	v.	v.	PROPN
cana-2345	250	18	sathya	sathya	PROPN
cana-2345	250	19	durga	durga	PROPN
cana-2345	250	20	,	,	PUNCT
cana-2345	250	21	and	and	CCONJ
cana-2345	250	22	n.	n.	PROPN
cana-2345	250	23	bharathi	bharathi	PROPN
cana-2345	250	24	raja	raja	PROPN
cana-2345	250	25	.	.	PUNCT
cana-2345	251	1	"	"	PUNCT
cana-2345	251	2	deep	deep	ADJ
cana-2345	251	3	learningdriven	learningdriven	NOUN
cana-2345	251	4	sentiment	sentiment	NOUN
cana-2345	251	5	analysis	analysis	NOUN
cana-2345	251	6	in	in	ADP
cana-2345	251	7	textual	textual	ADJ
cana-2345	251	8	data	datum	NOUN
cana-2345	251	9	.	.	PUNCT
cana-2345	251	10	"	"	PUNCT
cana-2345	252	1	in	in	ADP
cana-2345	252	2	2024	2024	NUM
cana-2345	252	3	ieee	ieee	NOUN
cana-2345	252	4	international	international	ADJ
cana-2345	252	5	conference	conference	NOUN
cana-2345	252	6	on	on	ADP
cana-2345	252	7	computing	computing	NOUN
cana-2345	252	8	,	,	PUNCT
cana-2345	252	9	power	power	NOUN
cana-2345	252	10	and	and	CCONJ
cana-2345	252	11	communication	communication	NOUN
cana-2345	252	12	technologies	technology	NOUN
cana-2345	252	13	(	(	PUNCT
cana-2345	252	14	ic2pct	ic2pct	PROPN
cana-2345	252	15	)	)	PUNCT
cana-2345	252	16	,	,	PUNCT
cana-2345	252	17	vol	vol	NOUN
cana-2345	252	18	.	.	PROPN
cana-2345	252	19	5	5	NUM
cana-2345	252	20	,	,	PUNCT
cana-2345	252	21	pp	pp	ADJ
cana-2345	252	22	.	.	PUNCT
cana-2345	252	23	1339	1339	NUM
cana-2345	252	24	-	-	SYM
cana-2345	252	25	1342	1342	NUM
cana-2345	252	26	.	.	PUNCT
cana-2345	253	1	ieee	ieee	NOUN
cana-2345	253	2	,	,	PUNCT
cana-2345	253	3	2024	2024	NUM
cana-2345	253	4	.	.	PUNCT
cana-2345	254	1	[	[	X
cana-2345	254	2	7	7	NUM
cana-2345	254	3	]	]	X
cana-2345	254	4	radha	radha	PROPN
cana-2345	254	5	,	,	PUNCT
cana-2345	254	6	dinesh	dinesh	PROPN
cana-2345	254	7	raja	raja	PROPN
cana-2345	254	8	,	,	PUNCT
cana-2345	254	9	and	and	CCONJ
cana-2345	254	10	l.	l.	PROPN
cana-2345	254	11	mary	mary	PROPN
cana-2345	254	12	gladence	gladence	PROPN
cana-2345	254	13	.	.	PUNCT
cana-2345	255	1	"	"	PUNCT
cana-2345	255	2	text	text	VERB
cana-2345	255	3	emotion	emotion	NOUN
cana-2345	255	4	multi	multi	ADJ
cana-2345	255	5	-	-	ADJ
cana-2345	255	6	label	label	ADJ
cana-2345	255	7	classification	classification	NOUN
cana-2345	255	8	with	with	ADP
cana-2345	255	9	vectoring	vectoring	NOUN
cana-2345	255	10	and	and	CCONJ
cana-2345	255	11	deep	deep	ADJ
cana-2345	255	12	learning	learning	NOUN
cana-2345	255	13	.	.	PUNCT
cana-2345	255	14	"	"	PUNCT
cana-2345	256	1	in	in	ADP
cana-2345	256	2	2024	2024	NUM
cana-2345	256	3	international	international	ADJ
cana-2345	256	4	conference	conference	NOUN
cana-2345	256	5	on	on	ADP
cana-2345	256	6	inventive	inventive	ADJ
cana-2345	256	7	computation	computation	NOUN
cana-2345	256	8	technologies	technology	NOUN
cana-2345	256	9	(	(	PUNCT
cana-2345	256	10	icict	icict	NOUN
cana-2345	256	11	)	)	PUNCT
cana-2345	256	12	,	,	PUNCT
cana-2345	256	13	pp	pp	PROPN
cana-2345	256	14	.	.	PUNCT
cana-2345	256	15	371	371	NUM
cana-2345	256	16	-	-	SYM
cana-2345	256	17	376	376	NUM
cana-2345	256	18	.	.	PUNCT
cana-2345	256	19	ieee	ieee	NOUN
cana-2345	256	20	,	,	PUNCT
cana-2345	256	21	2024	2024	NUM
cana-2345	256	22	.	.	PUNCT
cana-2345	257	1	[	[	X
cana-2345	257	2	8	8	NUM
cana-2345	257	3	]	]	X
cana-2345	257	4	bashynska	bashynska	NOUN
cana-2345	257	5	,	,	PUNCT
cana-2345	257	6	iryna	iryna	NOUN
cana-2345	257	7	,	,	PUNCT
cana-2345	257	8	mykhailo	mykhailo	NOUN
cana-2345	257	9	sarafanov	sarafanov	NOUN
cana-2345	257	10	,	,	PUNCT
cana-2345	257	11	and	and	CCONJ
cana-2345	257	12	olga	olga	PROPN
cana-2345	257	13	manikaeva	manikaeva	PROPN
cana-2345	257	14	.	.	PUNCT
cana-2345	258	1	"	"	PUNCT
cana-2345	258	2	research	research	NOUN
cana-2345	258	3	and	and	CCONJ
cana-2345	258	4	development	development	NOUN
cana-2345	258	5	of	of	ADP
cana-2345	258	6	a	a	DET
cana-2345	258	7	modern	modern	ADJ
cana-2345	258	8	deep	deep	ADJ
cana-2345	258	9	learning	learning	NOUN
cana-2345	258	10	model	model	NOUN
cana-2345	258	11	for	for	ADP
cana-2345	258	12	emotional	emotional	ADJ
cana-2345	258	13	analysis	analysis	NOUN
cana-2345	258	14	management	management	NOUN
cana-2345	258	15	of	of	ADP
cana-2345	258	16	text	text	NOUN
cana-2345	258	17	data	datum	NOUN
cana-2345	258	18	.	.	PUNCT
cana-2345	258	19	"	"	PUNCT
cana-2345	258	20	applied	apply	VERB
cana-2345	258	21	sciences	science	NOUN
cana-2345	258	22	14	14	NUM
cana-2345	258	23	,	,	PUNCT
cana-2345	258	24	no	no	INTJ
cana-2345	258	25	.	.	NOUN
cana-2345	258	26	5	5	NUM
cana-2345	258	27	(	(	PUNCT
cana-2345	258	28	2024	2024	NUM
cana-2345	258	29	):	):	PUNCT
cana-2345	258	30	1952	1952	NUM
cana-2345	258	31	.	.	PUNCT
cana-2345	259	1	[	[	X
cana-2345	259	2	9	9	NUM
cana-2345	259	3	]	]	X
cana-2345	259	4	salman	salman	PROPN
cana-2345	259	5	al	al	PROPN
cana-2345	259	6	-	-	PUNCT
cana-2345	259	7	tameemi	tameemi	PROPN
cana-2345	259	8	,	,	PUNCT
cana-2345	259	9	israa	israa	ADJ
cana-2345	259	10	k.	k.	PROPN
cana-2345	259	11	,	,	PUNCT
cana-2345	259	12	mohammad	mohammad	PROPN
cana-2345	259	13	-	-	PUNCT
cana-2345	259	14	reza	reza	PROPN
cana-2345	259	15	feizi	feizi	NOUN
cana-2345	259	16	-	-	PUNCT
cana-2345	259	17	derakhshi	derakhshi	PROPN
cana-2345	259	18	,	,	PUNCT
cana-2345	259	19	saeed	saeed	NOUN
cana-2345	259	20	pashazadeh	pashazadeh	PROPN
cana-2345	259	21	,	,	PUNCT
cana-2345	259	22	and	and	CCONJ
cana-2345	259	23	mohammad	mohammad	PROPN
cana-2345	259	24	asadpour	asadpour	PROPN
cana-2345	259	25	.	.	PUNCT
cana-2345	260	1	"	"	PUNCT
cana-2345	260	2	an	an	DET
cana-2345	260	3	efficient	efficient	ADJ
cana-2345	260	4	sentiment	sentiment	NOUN
cana-2345	260	5	classification	classification	NOUN
cana-2345	260	6	method	method	NOUN
cana-2345	260	7	with	with	ADP
cana-2345	260	8	the	the	DET
cana-2345	260	9	help	help	NOUN
cana-2345	260	10	of	of	ADP
cana-2345	260	11	neighbors	neighbor	NOUN
cana-2345	260	12	and	and	CCONJ
cana-2345	260	13	a	a	DET
cana-2345	260	14	hybrid	hybrid	NOUN
cana-2345	260	15	of	of	ADP
cana-2345	260	16	rnn	rnn	NOUN
cana-2345	260	17	models	model	NOUN
cana-2345	260	18	.	.	PUNCT
cana-2345	260	19	"	"	PUNCT
cana-2345	261	1	complexity	complexity	NOUN
cana-2345	261	2	2023	2023	NUM
cana-2345	261	3	,	,	PUNCT
cana-2345	261	4	no	no	INTJ
cana-2345	261	5	.	.	NOUN
cana-2345	261	6	1	1	NUM
cana-2345	261	7	(	(	PUNCT
cana-2345	261	8	2023	2023	NUM
cana-2345	261	9	):	):	PUNCT
cana-2345	261	10	1896556	1896556	NUM
cana-2345	261	11	.	.	PUNCT
cana-2345	262	1	[	[	X
cana-2345	262	2	10	10	NUM
cana-2345	262	3	]	]	X
cana-2345	262	4	zhao	zhao	PROPN
cana-2345	262	5	,	,	PUNCT
cana-2345	262	6	yuxia	yuxia	PROPN
cana-2345	262	7	,	,	PUNCT
cana-2345	262	8	mahpirat	mahpirat	NOUN
cana-2345	262	9	mamat	mamat	PROPN
cana-2345	262	10	,	,	PUNCT
cana-2345	262	11	alimjan	alimjan	NOUN
cana-2345	262	12	aysa	aysa	NOUN
cana-2345	262	13	,	,	PUNCT
cana-2345	262	14	and	and	CCONJ
cana-2345	262	15	kurban	kurban	PROPN
cana-2345	262	16	ubul	ubul	PROPN
cana-2345	262	17	.	.	PUNCT
cana-2345	263	1	"	"	PUNCT
cana-2345	263	2	a	a	DET
cana-2345	263	3	dynamic	dynamic	ADJ
cana-2345	263	4	graph	graph	NOUN
cana-2345	263	5	structural	structural	ADJ
cana-2345	263	6	framework	framework	NOUN
cana-2345	263	7	for	for	ADP
cana-2345	263	8	implicit	implicit	ADJ
cana-2345	263	9	sentiment	sentiment	NOUN
cana-2345	263	10	identification	identification	NOUN
cana-2345	263	11	based	base	VERB
cana-2345	263	12	on	on	ADP
cana-2345	263	13	complementary	complementary	ADJ
cana-2345	263	14	semantic	semantic	ADJ
cana-2345	263	15	and	and	CCONJ
cana-2345	263	16	structural	structural	ADJ
cana-2345	263	17	information	information	NOUN
cana-2345	263	18	.	.	PUNCT
cana-2345	263	19	"	"	PUNCT
cana-2345	264	1	scientific	scientific	ADJ
cana-2345	264	2	reports	report	NOUN
cana-2345	264	3	14	14	NUM
cana-2345	264	4	,	,	PUNCT
cana-2345	264	5	no	no	INTJ
cana-2345	264	6	.	.	NOUN
cana-2345	264	7	1	1	NUM
cana-2345	264	8	(	(	PUNCT
cana-2345	264	9	2024	2024	NUM
cana-2345	264	10	):	):	PUNCT
cana-2345	264	11	16563	16563	NUM
cana-2345	264	12	.	.	PUNCT
cana-2345	265	1	[	[	X
cana-2345	265	2	11	11	NUM
cana-2345	265	3	]	]	PUNCT
cana-2345	265	4	adilakshmi	adilakshmi	PROPN
cana-2345	265	5	,	,	PUNCT
cana-2345	265	6	konda	konda	PROPN
cana-2345	265	7	,	,	PUNCT
cana-2345	265	8	malladi	malladi	PROPN
cana-2345	265	9	srinivas	srinivas	PROPN
cana-2345	265	10	,	,	PUNCT
cana-2345	265	11	anuradha	anuradha	PROPN
cana-2345	265	12	kodali	kodali	PROPN
cana-2345	265	13	,	,	PUNCT
cana-2345	265	14	and	and	CCONJ
cana-2345	265	15	v.	v.	ADP
cana-2345	265	16	srilakshmi	srilakshmi	NOUN
cana-2345	265	17	.	.	PUNCT
cana-2345	266	1	"	"	PUNCT
cana-2345	266	2	optimized	optimize	VERB
cana-2345	266	3	rmdl	rmdl	NOUN
cana-2345	266	4	with	with	ADP
cana-2345	266	5	transfer	transfer	NOUN
cana-2345	266	6	learning	learning	NOUN
cana-2345	266	7	for	for	ADP
cana-2345	266	8	sentiment	sentiment	NOUN
cana-2345	266	9	classification	classification	NOUN
cana-2345	266	10	in	in	ADP
cana-2345	266	11	the	the	DET
cana-2345	266	12	mapreduce	mapreduce	NOUN
cana-2345	266	13	framework	framework	NOUN
cana-2345	266	14	.	.	PUNCT
cana-2345	266	15	"	"	PUNCT
cana-2345	267	1	journal	journal	PROPN
cana-2345	267	2	of	of	ADP
cana-2345	267	3	web	web	NOUN
cana-2345	267	4	engineering	engineering	NOUN
cana-2345	267	5	22	22	NUM
cana-2345	267	6	,	,	PUNCT
cana-2345	267	7	no	no	INTJ
cana-2345	267	8	.	.	NOUN
cana-2345	267	9	8	8	NUM
cana-2345	267	10	(	(	PUNCT
cana-2345	267	11	2023	2023	NUM
cana-2345	267	12	):	):	PUNCT
cana-2345	267	13	1101	1101	NUM
cana-2345	267	14	-	-	SYM
cana-2345	267	15	1132	1132	NUM
cana-2345	267	16	.	.	PUNCT
cana-2345	268	1	communications	communication	NOUN
cana-2345	268	2	on	on	ADP
cana-2345	268	3	applied	apply	VERB
cana-2345	268	4	nonlinear	nonlinear	ADJ
cana-2345	268	5	analysis	analysis	NOUN
cana-2345	268	6	issn	issn	NOUN
cana-2345	268	7	:	:	PUNCT
cana-2345	268	8	1074	1074	NUM
cana-2345	268	9	-	-	PUNCT
cana-2345	268	10	133x	133x	NUM
cana-2345	268	11	vol	vol	NOUN
cana-2345	268	12	32	32	NUM
cana-2345	268	13	no	no	NOUN
cana-2345	268	14	.	.	PUNCT
cana-2345	269	1	1s	1s	NUM
cana-2345	269	2	(	(	PUNCT
cana-2345	269	3	2025	2025	NUM
cana-2345	269	4	)	)	PUNCT
cana-2345	269	5	584	584	NUM
cana-2345	269	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2345	270	1	[	[	X
cana-2345	270	2	12	12	NUM
cana-2345	270	3	]	]	X
cana-2345	270	4	sachin	sachin	PROPN
cana-2345	270	5	sambhaji	sambhaji	PROPN
cana-2345	270	6	patil	patil	PROPN
cana-2345	270	7	,	,	PUNCT
cana-2345	270	8	anthon	anthon	PROPN
cana-2345	270	9	rodrigues	rodrigues	PROPN
cana-2345	270	10	,	,	PUNCT
cana-2345	270	11	rahul	rahul	PROPN
cana-2345	270	12	telangi	telangi	PROPN
cana-2345	270	13	,	,	PUNCT
cana-2345	270	14	vishwajeet	vishwajeet	NOUN
cana-2345	270	15	chavan	chavan	NOUN
cana-2345	270	16	.	.	PUNCT
cana-2345	271	1	"	"	PUNCT
cana-2345	271	2	convolutional	convolutional	ADJ
cana-2345	271	3	neural	neural	ADJ
cana-2345	271	4	networks	network	NOUN
cana-2345	271	5	for	for	ADP
cana-2345	271	6	text	text	NOUN
cana-2345	271	7	classification	classification	NOUN
cana-2345	271	8	:	:	PUNCT
cana-2345	271	9	a	a	DET
cana-2345	271	10	comprehensive	comprehensive	ADJ
cana-2345	271	11	analysis	analysis	NOUN
cana-2345	271	12	"	"	PUNCT
cana-2345	271	13	,	,	PUNCT
cana-2345	271	14	international	international	ADJ
cana-2345	271	15	journal	journal	NOUN
cana-2345	271	16	for	for	ADP
cana-2345	271	17	research	research	NOUN
cana-2345	271	18	in	in	ADP
cana-2345	271	19	applied	apply	VERB
cana-2345	271	20	science	science	PROPN
cana-2345	271	21	&	&	CCONJ
cana-2345	271	22	engineering	engineering	PROPN
cana-2345	271	23	technology	technology	NOUN
cana-2345	271	24	(	(	PUNCT
cana-2345	271	25	ijraset	ijraset	PROPN
cana-2345	271	26	)	)	PUNCT
cana-2345	271	27	,	,	PUNCT
cana-2345	271	28	issn	issn	PROPN
cana-2345	271	29	:	:	PUNCT
cana-2345	271	30	2321	2321	NUM
cana-2345	271	31	-	-	SYM
cana-2345	271	32	9653	9653	NUM
cana-2345	271	33	;	;	PUNCT
cana-2345	271	34	pp	pp	NUM
cana-2345	271	35	.	.	PUNCT
cana-2345	272	1	6039	6039	NUM
cana-2345	272	2	-	-	SYM
cana-2345	272	3	6047	6047	NUM
cana-2345	272	4	,	,	PUNCT
cana-2345	272	5	volume	volume	NOUN
cana-2345	272	6	11	11	NUM
cana-2345	272	7	issue	issue	NOUN
cana-2345	272	8	v	v	NOUN
cana-2345	272	9	may	may	AUX
cana-2345	272	10	2023	2023	NUM
cana-2345	272	11	[	[	PUNCT
cana-2345	272	12	13	13	NUM
cana-2345	272	13	]	]	SYM
cana-2345	272	14	nguyen	nguyen	NOUN
cana-2345	272	15	,	,	PUNCT
cana-2345	272	16	cuong	cuong	PROPN
cana-2345	272	17	v.	v.	ADV
cana-2345	272	18	,	,	PUNCT
cana-2345	272	19	khiem	khiem	PROPN
cana-2345	272	20	h.	h.	PROPN
cana-2345	272	21	le	le	PROPN
cana-2345	272	22	,	,	PUNCT
cana-2345	272	23	anh	anh	PROPN
cana-2345	272	24	m.	m.	NOUN
cana-2345	272	25	tran	tran	PROPN
cana-2345	272	26	,	,	PUNCT
cana-2345	272	27	quang	quang	PROPN
cana-2345	272	28	h.	h.	PROPN
cana-2345	272	29	pham	pham	PROPN
cana-2345	272	30	,	,	PUNCT
cana-2345	272	31	and	and	CCONJ
cana-2345	272	32	binh	binh	PROPN
cana-2345	272	33	t.	t.	PROPN
cana-2345	272	34	nguyen	nguyen	PROPN
cana-2345	272	35	.	.	PUNCT
cana-2345	273	1	"	"	PUNCT
cana-2345	273	2	learning	learn	VERB
cana-2345	273	3	for	for	ADP
cana-2345	273	4	amalgamation	amalgamation	NOUN
cana-2345	273	5	:	:	PUNCT
cana-2345	273	6	a	a	DET
cana-2345	273	7	multi	multi	ADJ
cana-2345	273	8	-	-	ADJ
cana-2345	273	9	source	source	NOUN
cana-2345	273	10	transfer	transfer	NOUN
cana-2345	273	11	learning	learn	VERB
cana-2345	273	12	framework	framework	NOUN
cana-2345	273	13	for	for	ADP
cana-2345	273	14	sentiment	sentiment	NOUN
cana-2345	273	15	classification	classification	NOUN
cana-2345	273	16	.	.	PUNCT
cana-2345	273	17	"	"	PUNCT
cana-2345	274	1	information	information	NOUN
cana-2345	274	2	sciences	science	NOUN
cana-2345	274	3	590	590	NUM
cana-2345	274	4	(	(	PUNCT
cana-2345	274	5	2022	2022	NUM
cana-2345	274	6	):	):	PUNCT
cana-2345	274	7	1	1	NUM
cana-2345	274	8	-	-	SYM
cana-2345	274	9	14	14	NUM
cana-2345	274	10	.	.	PUNCT
cana-2345	275	1	[	[	X
cana-2345	275	2	14	14	NUM
cana-2345	275	3	]	]	X
cana-2345	275	4	suganya	suganya	ADJ
cana-2345	275	5	,	,	PUNCT
cana-2345	275	6	v.	v.	ADV
cana-2345	275	7	,	,	PUNCT
cana-2345	275	8	and	and	CCONJ
cana-2345	275	9	m.	m.	PROPN
cana-2345	275	10	rajeev	rajeev	PROPN
cana-2345	275	11	kumar	kumar	PROPN
cana-2345	275	12	.	.	PUNCT
cana-2345	276	1	"	"	PUNCT
cana-2345	276	2	advancing	advance	VERB
cana-2345	276	3	sentiment	sentiment	NOUN
cana-2345	276	4	analysis	analysis	NOUN
cana-2345	276	5	precision	precision	NOUN
cana-2345	276	6	through	through	ADP
cana-2345	276	7	convnet	convnet	NOUN
cana-2345	276	8	-	-	PUNCT
cana-2345	276	9	svmbovw	svmbovw	ADJ
cana-2345	276	10	mode	mode	NOUN
cana-2345	276	11	in	in	ADP
cana-2345	276	12	a	a	DET
cana-2345	276	13	hybrid	hybrid	ADJ
cana-2345	276	14	context	context	NOUN
cana-2345	276	15	-	-	PUNCT
cana-2345	276	16	enhanced	enhance	VERB
cana-2345	276	17	deep	deep	ADJ
cana-2345	276	18	learning	learning	NOUN
cana-2345	276	19	paradigm	paradigm	NOUN
cana-2345	276	20	.	.	PUNCT
cana-2345	276	21	"	"	PUNCT
cana-2345	277	1	in	in	ADP
cana-2345	277	2	2024	2024	NUM
cana-2345	277	3	3rd	3rd	ADJ
cana-2345	277	4	international	international	ADJ
cana-2345	277	5	conference	conference	NOUN
cana-2345	277	6	on	on	ADP
cana-2345	277	7	applied	apply	VERB
cana-2345	277	8	artificial	artificial	ADJ
cana-2345	277	9	intelligence	intelligence	NOUN
cana-2345	277	10	and	and	CCONJ
cana-2345	277	11	computing	computing	NOUN
cana-2345	277	12	(	(	PUNCT
cana-2345	277	13	icaaic	icaaic	ADJ
cana-2345	277	14	)	)	PUNCT
cana-2345	277	15	,	,	PUNCT
cana-2345	277	16	pp	pp	PROPN
cana-2345	277	17	.	.	PUNCT
cana-2345	278	1	836	836	NUM
cana-2345	278	2	-	-	SYM
cana-2345	278	3	843	843	NUM
cana-2345	278	4	.	.	PUNCT
cana-2345	278	5	ieee	ieee	NOUN
cana-2345	278	6	,	,	PUNCT
cana-2345	278	7	2024	2024	NUM
cana-2345	278	8	.	.	PUNCT
cana-2345	279	1	[	[	X
cana-2345	279	2	15	15	NUM
cana-2345	279	3	]	]	X
cana-2345	279	4	priyanka	priyanka	PROPN
cana-2345	279	5	,	,	PUNCT
cana-2345	279	6	madineni	madineni	PROPN
cana-2345	279	7	,	,	PUNCT
cana-2345	279	8	pushadapu	pushadapu	PROPN
cana-2345	279	9	karthik	karthik	PROPN
cana-2345	279	10	,	,	PUNCT
cana-2345	279	11	and	and	CCONJ
cana-2345	279	12	prasanth	prasanth	PROPN
cana-2345	279	13	yalla	yalla	PROPN
cana-2345	279	14	.	.	PUNCT
cana-2345	280	1	"	"	PUNCT
cana-2345	280	2	text	text	VERB
cana-2345	280	3	classification	classification	NOUN
cana-2345	280	4	into	into	ADP
cana-2345	280	5	emotional	emotional	ADJ
cana-2345	280	6	states	state	NOUN
cana-2345	280	7	using	use	VERB
cana-2345	280	8	deep	deep	ADJ
cana-2345	280	9	learning	learning	NOUN
cana-2345	280	10	based	base	VERB
cana-2345	280	11	bert	bert	PROPN
cana-2345	280	12	technique	technique	NOUN
cana-2345	280	13	.	.	PUNCT
cana-2345	280	14	"	"	PUNCT
cana-2345	281	1	in	in	ADP
cana-2345	281	2	2023	2023	NUM
cana-2345	281	3	8th	8th	ADJ
cana-2345	281	4	international	international	ADJ
cana-2345	281	5	conference	conference	NOUN
cana-2345	281	6	on	on	ADP
cana-2345	281	7	communication	communication	NOUN
cana-2345	281	8	and	and	CCONJ
cana-2345	281	9	electronics	electronic	NOUN
cana-2345	281	10	systems	system	NOUN
cana-2345	281	11	(	(	PUNCT
cana-2345	281	12	icces	icce	NOUN
cana-2345	281	13	)	)	PUNCT
cana-2345	281	14	,	,	PUNCT
cana-2345	281	15	pp	pp	ADJ
cana-2345	281	16	.	.	PUNCT
cana-2345	281	17	1005	1005	NUM
cana-2345	281	18	-	-	SYM
cana-2345	281	19	1014	1014	NUM
cana-2345	281	20	.	.	PUNCT
cana-2345	281	21	ieee	ieee	NOUN
cana-2345	281	22	,	,	PUNCT
cana-2345	281	23	2023	2023	NUM
cana-2345	281	24	.	.	PUNCT
cana-2345	282	1	[	[	X
cana-2345	282	2	16	16	NUM
cana-2345	282	3	]	]	X
cana-2345	282	4	chen	chen	PROPN
cana-2345	282	5	,	,	PUNCT
cana-2345	282	6	long	long	ADV
cana-2345	282	7	,	,	PUNCT
cana-2345	282	8	zhilin	zhilin	PROPN
cana-2345	282	9	ren	ren	PROPN
cana-2345	282	10	,	,	PUNCT
cana-2345	282	11	shaoshuai	shaoshuai	PROPN
cana-2345	282	12	lu	lu	PROPN
cana-2345	282	13	,	,	PUNCT
cana-2345	282	14	xiaohua	xiaohua	PROPN
cana-2345	282	15	huang	huang	PROPN
cana-2345	282	16	,	,	PUNCT
cana-2345	282	17	wenjing	wenje	VERB
cana-2345	282	18	wang	wang	PROPN
cana-2345	282	19	,	,	PUNCT
cana-2345	282	20	cai	cai	PROPN
cana-2345	282	21	xu	xu	PROPN
cana-2345	282	22	,	,	PUNCT
cana-2345	282	23	wei	wei	PROPN
cana-2345	282	24	zhao	zhao	PROPN
cana-2345	282	25	,	,	PUNCT
cana-2345	282	26	and	and	CCONJ
cana-2345	282	27	ziyu	ziyu	PROPN
cana-2345	282	28	guan	guan	PROPN
cana-2345	282	29	.	.	PUNCT
cana-2345	283	1	"	"	PUNCT
cana-2345	283	2	a	a	DET
cana-2345	283	3	simple	simple	ADJ
cana-2345	283	4	weakly	weakly	ADV
cana-2345	283	5	-	-	PUNCT
cana-2345	283	6	supervised	supervise	VERB
cana-2345	283	7	contrastive	contrastive	ADJ
cana-2345	283	8	learning	learning	NOUN
cana-2345	283	9	framework	framework	NOUN
cana-2345	283	10	for	for	ADP
cana-2345	283	11	few	few	ADJ
cana-2345	283	12	-	-	PUNCT
cana-2345	283	13	shot	shot	NOUN
cana-2345	283	14	sentiment	sentiment	NOUN
cana-2345	283	15	classification	classification	NOUN
cana-2345	283	16	.	.	PUNCT
cana-2345	283	17	"	"	PUNCT
cana-2345	284	1	(	(	PUNCT
cana-2345	284	2	2023	2023	NUM
cana-2345	284	3	)	)	PUNCT
cana-2345	284	4	.	.	PUNCT
cana-2345	285	1	[	[	X
cana-2345	285	2	17	17	NUM
cana-2345	285	3	]	]	X
cana-2345	285	4	jadon	jadon	PROPN
cana-2345	285	5	,	,	PUNCT
cana-2345	285	6	adarsh	adarsh	PROPN
cana-2345	285	7	singh	singh	PROPN
cana-2345	285	8	,	,	PUNCT
cana-2345	285	9	mahesh	mahesh	PROPN
cana-2345	285	10	parmar	parmar	PROPN
cana-2345	285	11	,	,	PUNCT
cana-2345	285	12	and	and	CCONJ
cana-2345	285	13	rohit	rohit	PROPN
cana-2345	285	14	agrawal	agrawal	PROPN
cana-2345	285	15	.	.	PUNCT
cana-2345	286	1	"	"	PUNCT
cana-2345	286	2	hinglish	hinglish	VERB
cana-2345	286	3	sentiment	sentiment	NOUN
cana-2345	286	4	analysis	analysis	NOUN
cana-2345	286	5	:	:	PUNCT
cana-2345	286	6	deep	deep	ADJ
cana-2345	286	7	learning	learning	NOUN
cana-2345	286	8	models	model	NOUN
cana-2345	286	9	for	for	ADP
cana-2345	286	10	nuanced	nuanced	ADJ
cana-2345	286	11	sentiment	sentiment	NOUN
cana-2345	286	12	classification	classification	NOUN
cana-2345	286	13	in	in	ADP
cana-2345	286	14	multilingual	multilingual	ADJ
cana-2345	286	15	digital	digital	ADJ
cana-2345	286	16	communication	communication	NOUN
cana-2345	286	17	.	.	PUNCT
cana-2345	286	18	"	"	PUNCT
cana-2345	287	1	in	in	ADP
cana-2345	287	2	2024	2024	NUM
cana-2345	287	3	2nd	2nd	ADJ
cana-2345	287	4	international	international	ADJ
cana-2345	287	5	conference	conference	NOUN
cana-2345	287	6	on	on	ADP
cana-2345	287	7	device	device	NOUN
cana-2345	287	8	intelligence	intelligence	NOUN
cana-2345	287	9	,	,	PUNCT
cana-2345	287	10	computing	computing	NOUN
cana-2345	287	11	and	and	CCONJ
cana-2345	287	12	communication	communication	NOUN
cana-2345	287	13	technologies	technology	NOUN
cana-2345	287	14	(	(	PUNCT
cana-2345	287	15	dicct	dicct	NOUN
cana-2345	287	16	)	)	PUNCT
cana-2345	287	17	,	,	PUNCT
cana-2345	287	18	pp	pp	PROPN
cana-2345	287	19	.	.	PUNCT
cana-2345	287	20	318	318	NUM
cana-2345	287	21	-	-	SYM
cana-2345	287	22	323	323	NUM
cana-2345	287	23	.	.	PUNCT
cana-2345	287	24	ieee	ieee	NOUN
cana-2345	287	25	,	,	PUNCT
cana-2345	287	26	2024	2024	NUM
cana-2345	287	27	.	.	PUNCT
cana-2345	288	1	[	[	X
cana-2345	288	2	18	18	NUM
cana-2345	288	3	]	]	X
cana-2345	288	4	gamal	gamal	PROPN
cana-2345	288	5	,	,	PUNCT
cana-2345	288	6	donia	donia	PROPN
cana-2345	288	7	,	,	PUNCT
cana-2345	288	8	marco	marco	PROPN
cana-2345	288	9	alfonse	alfonse	NOUN
cana-2345	288	10	,	,	PUNCT
cana-2345	288	11	salud	salud	PROPN
cana-2345	288	12	maría	maría	PROPN
cana-2345	288	13	jiménez	jiménez	PROPN
cana-2345	288	14	-	-	PUNCT
cana-2345	288	15	zafra	zafra	PROPN
cana-2345	288	16	,	,	PUNCT
cana-2345	288	17	and	and	CCONJ
cana-2345	288	18	mostafa	mostafa	PROPN
cana-2345	288	19	aref	aref	PROPN
cana-2345	288	20	.	.	PUNCT
cana-2345	289	1	"	"	PUNCT
cana-2345	289	2	arabic	arabic	ADJ
cana-2345	289	3	sentiment	sentiment	NOUN
cana-2345	289	4	classification	classification	NOUN
cana-2345	289	5	on	on	ADP
cana-2345	289	6	twitter	twitter	NOUN
cana-2345	289	7	using	use	VERB
cana-2345	289	8	deep	deep	ADJ
cana-2345	289	9	learning	learning	NOUN
cana-2345	289	10	techniques	technique	NOUN
cana-2345	289	11	.	.	PUNCT
cana-2345	289	12	"	"	PUNCT
cana-2345	290	1	in	in	ADP
cana-2345	290	2	the	the	DET
cana-2345	290	3	international	international	ADJ
cana-2345	290	4	symposium	symposium	NOUN
cana-2345	290	5	on	on	ADP
cana-2345	290	6	computer	computer	NOUN
cana-2345	290	7	science	science	NOUN
cana-2345	290	8	,	,	PUNCT
cana-2345	290	9	digital	digital	ADJ
cana-2345	290	10	economy	economy	NOUN
cana-2345	290	11	and	and	CCONJ
cana-2345	290	12	intelligent	intelligent	ADJ
cana-2345	290	13	systems	system	NOUN
cana-2345	290	14	,	,	PUNCT
cana-2345	290	15	pp	pp	ADJ
cana-2345	290	16	.	.	PUNCT
cana-2345	290	17	236	236	NUM
cana-2345	290	18	-	-	SYM
cana-2345	290	19	251	251	NUM
cana-2345	290	20	.	.	PUNCT
cana-2345	290	21	cham	cham	PROPN
cana-2345	290	22	:	:	PUNCT
cana-2345	290	23	springer	springer	NOUN
cana-2345	290	24	nature	nature	PROPN
cana-2345	290	25	switzerland	switzerland	PROPN
cana-2345	290	26	,	,	PUNCT
cana-2345	290	27	2022	2022	NUM
cana-2345	290	28	.	.	PUNCT
cana-2345	291	1	[	[	X
cana-2345	291	2	19	19	NUM
cana-2345	291	3	]	]	X
cana-2345	291	4	garg	garg	NOUN
cana-2345	291	5	,	,	PUNCT
cana-2345	291	6	sarita	sarita	PROPN
cana-2345	291	7	bansal	bansal	NOUN
cana-2345	291	8	,	,	PUNCT
cana-2345	291	9	and	and	CCONJ
cana-2345	291	10	v.	v.	ADP
cana-2345	291	11	v.	v.	CCONJ
cana-2345	291	12	subrahmanyam	subrahmanyam	NOUN
cana-2345	291	13	.	.	PUNCT
cana-2345	292	1	"	"	PUNCT
cana-2345	292	2	super	super	ADJ
cana-2345	292	3	deep	deep	ADJ
cana-2345	292	4	learning	learn	VERB
cana-2345	292	5	ensemble	ensemble	ADJ
cana-2345	292	6	model	model	NOUN
cana-2345	292	7	for	for	ADP
cana-2345	292	8	sentiment	sentiment	NOUN
cana-2345	292	9	analysis	analysis	NOUN
cana-2345	292	10	.	.	PUNCT
cana-2345	292	11	"	"	PUNCT
cana-2345	293	1	in	in	ADP
cana-2345	293	2	2023	2023	NUM
cana-2345	293	3	international	international	ADJ
cana-2345	293	4	conference	conference	NOUN
cana-2345	293	5	on	on	ADP
cana-2345	293	6	computing	computing	NOUN
cana-2345	293	7	,	,	PUNCT
cana-2345	293	8	communication	communication	NOUN
cana-2345	293	9	,	,	PUNCT
cana-2345	293	10	and	and	CCONJ
cana-2345	293	11	intelligent	intelligent	ADJ
cana-2345	293	12	systems	system	NOUN
cana-2345	293	13	(	(	PUNCT
cana-2345	293	14	icccis	icccis	NOUN
cana-2345	293	15	)	)	PUNCT
cana-2345	293	16	,	,	PUNCT
cana-2345	293	17	pp	pp	ADP
cana-2345	293	18	.	.	PUNCT
cana-2345	294	1	341	341	NUM
cana-2345	294	2	-	-	SYM
cana-2345	294	3	346	346	NUM
cana-2345	294	4	.	.	PUNCT
cana-2345	294	5	ieee	ieee	NOUN
cana-2345	294	6	,	,	PUNCT
cana-2345	294	7	2023	2023	NUM
cana-2345	294	8	.	.	PUNCT
cana-2345	295	1	[	[	X
cana-2345	295	2	20	20	NUM
cana-2345	295	3	]	]	X
cana-2345	295	4	raja	raja	PROPN
cana-2345	295	5	,	,	PUNCT
cana-2345	295	6	m.	m.	PROPN
cana-2345	295	7	ramesh	ramesh	PROPN
cana-2345	295	8	,	,	PUNCT
cana-2345	295	9	and	and	CCONJ
cana-2345	295	10	j.	j.	PROPN
cana-2345	295	11	arunadevi	arunadevi	PROPN
cana-2345	295	12	.	.	PUNCT
cana-2345	296	1	"	"	PUNCT
cana-2345	296	2	deep	deep	ADJ
cana-2345	296	3	active	active	ADJ
cana-2345	296	4	learning	learn	VERB
cana-2345	296	5	multiclass	multiclass	ADJ
cana-2345	296	6	classifier	classifier	NOUN
cana-2345	296	7	for	for	ADP
cana-2345	296	8	the	the	DET
cana-2345	296	9	sentimental	sentimental	ADJ
cana-2345	296	10	analysis	analysis	NOUN
cana-2345	296	11	in	in	ADP
cana-2345	296	12	imbalanced	imbalanced	ADJ
cana-2345	296	13	unstructured	unstructured	ADJ
cana-2345	296	14	text	text	NOUN
cana-2345	296	15	data	datum	NOUN
cana-2345	296	16	.	.	PUNCT
cana-2345	296	17	"	"	PUNCT
cana-2345	297	1	in	in	ADP
cana-2345	297	2	2023	2023	NUM
cana-2345	297	3	international	international	ADJ
cana-2345	297	4	conference	conference	NOUN
cana-2345	297	5	on	on	ADP
cana-2345	297	6	data	datum	NOUN
cana-2345	297	7	science	science	NOUN
cana-2345	297	8	,	,	PUNCT
cana-2345	297	9	agents	agent	NOUN
cana-2345	297	10	&	&	CCONJ
cana-2345	297	11	artificial	artificial	ADJ
cana-2345	297	12	intelligence	intelligence	NOUN
cana-2345	297	13	(	(	PUNCT
cana-2345	297	14	icdsaai	icdsaai	PROPN
cana-2345	297	15	)	)	PUNCT
cana-2345	297	16	,	,	PUNCT
cana-2345	297	17	pp	pp	ADJ
cana-2345	297	18	.	.	PUNCT
cana-2345	298	1	1	1	NUM
cana-2345	298	2	-	-	SYM
cana-2345	298	3	6	6	NUM
cana-2345	298	4	.	.	PUNCT
cana-2345	298	5	ieee	ieee	NOUN
cana-2345	298	6	,	,	PUNCT
cana-2345	298	7	2023	2023	NUM
cana-2345	298	8	.	.	PUNCT
cana-2345	299	1	[	[	X
cana-2345	299	2	21	21	NUM
cana-2345	299	3	]	]	X
cana-2345	299	4	kiran	kiran	PROPN
cana-2345	299	5	,	,	PUNCT
cana-2345	299	6	k.	k.	PROPN
cana-2345	299	7	,	,	PUNCT
cana-2345	299	8	abhishek	abhishek	PROPN
cana-2345	299	9	appaji	appaji	PROPN
cana-2345	299	10	,	,	PUNCT
cana-2345	299	11	and	and	CCONJ
cana-2345	299	12	p.	p.	PROPN
cana-2345	299	13	deepa	deepa	PROPN
cana-2345	299	14	shenoy	shenoy	VERB
cana-2345	299	15	.	.	PUNCT
cana-2345	300	1	"	"	PUNCT
cana-2345	300	2	sentiment	sentiment	NOUN
cana-2345	300	3	analysis	analysis	NOUN
cana-2345	300	4	of	of	ADP
cana-2345	300	5	textual	textual	ADJ
cana-2345	300	6	data	datum	NOUN
cana-2345	300	7	using	use	VERB
cana-2345	300	8	word	word	NOUN
cana-2345	300	9	embedding	embed	VERB
cana-2345	300	10	and	and	CCONJ
cana-2345	300	11	deep	deep	ADJ
cana-2345	300	12	learning	learning	NOUN
cana-2345	300	13	approaches	approach	NOUN
cana-2345	300	14	.	.	PUNCT
cana-2345	300	15	"	"	PUNCT
cana-2345	301	1	in	in	ADP
cana-2345	301	2	2023	2023	NUM
cana-2345	301	3	ieee	ieee	NOUN
cana-2345	301	4	north	north	PROPN
cana-2345	301	5	karnataka	karnataka	PROPN
cana-2345	301	6	subsection	subsection	NOUN
cana-2345	301	7	flagship	flagship	NOUN
cana-2345	301	8	international	international	ADJ
cana-2345	301	9	conference	conference	NOUN
cana-2345	301	10	(	(	PUNCT
cana-2345	301	11	nkcon	nkcon	NOUN
cana-2345	301	12	)	)	PUNCT
cana-2345	301	13	,	,	PUNCT
cana-2345	301	14	pp	pp	PROPN
cana-2345	301	15	.	.	PUNCT
cana-2345	302	1	1	1	NUM
cana-2345	302	2	-	-	SYM
cana-2345	302	3	6	6	NUM
cana-2345	302	4	.	.	PUNCT
cana-2345	302	5	ieee	ieee	NOUN
cana-2345	302	6	,	,	PUNCT
cana-2345	302	7	2023	2023	NUM
cana-2345	302	8	.	.	PUNCT
cana-2345	303	1	[	[	X
cana-2345	303	2	22	22	NUM
cana-2345	303	3	]	]	PUNCT
cana-2345	303	4	rangarajan	rangarajan	PROPN
cana-2345	303	5	,	,	PUNCT
cana-2345	303	6	prasanna	prasanna	PROPN
cana-2345	303	7	kumar	kumar	PROPN
cana-2345	303	8	,	,	PUNCT
cana-2345	303	9	bharathi	bharathi	PROPN
cana-2345	303	10	mohan	mohan	PROPN
cana-2345	303	11	gurusamy	gurusamy	PROPN
cana-2345	303	12	,	,	PUNCT
cana-2345	303	13	gayathri	gayathri	PROPN
cana-2345	303	14	muthurasu	muthurasu	PROPN
cana-2345	303	15	,	,	PUNCT
cana-2345	303	16	rithani	rithani	PROPN
cana-2345	303	17	mohan	mohan	PROPN
cana-2345	303	18	,	,	PUNCT
cana-2345	303	19	gundala	gundala	PROPN
cana-2345	303	20	pallavi	pallavi	NOUN
cana-2345	303	21	,	,	PUNCT
cana-2345	303	22	sulochana	sulochana	NOUN
cana-2345	303	23	vijayakumar	vijayakumar	PROPN
cana-2345	303	24	,	,	PUNCT
cana-2345	303	25	and	and	CCONJ
cana-2345	303	26	ali	ali	PROPN
cana-2345	303	27	altalbe	altalbe	NOUN
cana-2345	303	28	.	.	PUNCT
cana-2345	304	1	"	"	PUNCT
cana-2345	304	2	the	the	DET
cana-2345	304	3	social	social	ADJ
cana-2345	304	4	media	medium	NOUN
cana-2345	304	5	sentiment	sentiment	NOUN
cana-2345	304	6	analysis	analysis	NOUN
cana-2345	304	7	framework	framework	NOUN
cana-2345	304	8	:	:	PUNCT
cana-2345	304	9	deep	deep	ADJ
cana-2345	304	10	learning	learn	VERB
cana-2345	304	11	for	for	ADP
cana-2345	304	12	sentiment	sentiment	NOUN
cana-2345	304	13	analysis	analysis	NOUN
cana-2345	304	14	on	on	ADP
cana-2345	304	15	social	social	ADJ
cana-2345	304	16	media	medium	NOUN
cana-2345	304	17	.	.	PUNCT
cana-2345	304	18	"	"	PUNCT
cana-2345	305	1	international	international	ADJ
cana-2345	305	2	journal	journal	NOUN
cana-2345	305	3	of	of	ADP
cana-2345	305	4	electrical	electrical	ADJ
cana-2345	305	5	and	and	CCONJ
cana-2345	305	6	computer	computer	NOUN
cana-2345	305	7	engineering	engineering	NOUN
cana-2345	305	8	(	(	PUNCT
cana-2345	305	9	ijece	ijece	PROPN
cana-2345	305	10	)	)	PUNCT
cana-2345	305	11	14	14	NUM
cana-2345	305	12	,	,	PUNCT
cana-2345	305	13	no	no	INTJ
cana-2345	305	14	.	.	NOUN
cana-2345	305	15	3	3	NUM
cana-2345	305	16	(	(	PUNCT
cana-2345	305	17	2024	2024	NUM
cana-2345	305	18	):	):	PUNCT
cana-2345	305	19	1	1	NUM
cana-2345	305	20	-	-	SYM
cana-2345	305	21	1x	1x	NUM
cana-2345	305	22	.	.	PUNCT
cana-2345	306	1	[	[	X
cana-2345	306	2	23	23	NUM
cana-2345	306	3	]	]	X
cana-2345	306	4	hernández	hernández	PROPN
cana-2345	306	5	,	,	PUNCT
cana-2345	306	6	nayeli	nayeli	PROPN
cana-2345	306	7	,	,	PUNCT
cana-2345	306	8	ildar	ildar	PROPN
cana-2345	306	9	batyrshin	batyrshin	PROPN
cana-2345	306	10	,	,	PUNCT
cana-2345	306	11	and	and	CCONJ
cana-2345	306	12	grigori	grigori	PROPN
cana-2345	306	13	sidorov	sidorov	PROPN
cana-2345	306	14	.	.	PUNCT
cana-2345	307	1	"	"	PUNCT
cana-2345	307	2	evaluation	evaluation	NOUN
cana-2345	307	3	of	of	ADP
cana-2345	307	4	deep	deep	ADJ
cana-2345	307	5	learning	learning	NOUN
cana-2345	307	6	models	model	NOUN
cana-2345	307	7	for	for	ADP
cana-2345	307	8	sentiment	sentiment	NOUN
cana-2345	307	9	analysis	analysis	NOUN
cana-2345	307	10	.	.	PUNCT
cana-2345	307	11	"	"	PUNCT
cana-2345	308	1	journal	journal	NOUN
cana-2345	308	2	of	of	ADP
cana-2345	308	3	intelligent	intelligent	ADJ
cana-2345	308	4	&	&	CCONJ
cana-2345	308	5	fuzzy	fuzzy	ADJ
cana-2345	308	6	systems	system	NOUN
cana-2345	308	7	43	43	NUM
cana-2345	308	8	,	,	PUNCT
cana-2345	308	9	no	no	INTJ
cana-2345	308	10	.	.	NOUN
cana-2345	308	11	6	6	NUM
cana-2345	308	12	(	(	PUNCT
cana-2345	308	13	2022	2022	NUM
cana-2345	308	14	):	):	PUNCT
cana-2345	308	15	6953	6953	NUM
cana-2345	308	16	-	-	SYM
cana-2345	308	17	6963	6963	NUM
cana-2345	308	18	.	.	PUNCT