id	sid	tid	token	lemma	pos
cana-2597	1	1	communications	communication	NOUN
cana-2597	1	2	on	on	ADP
cana-2597	1	3	applied	apply	VERB
cana-2597	1	4	nonlinear	nonlinear	ADJ
cana-2597	1	5	analysis	analysis	NOUN
cana-2597	1	6	issn	issn	NOUN
cana-2597	1	7	:	:	PUNCT
cana-2597	1	8	1074	1074	NUM
cana-2597	1	9	-	-	PUNCT
cana-2597	1	10	133x	133x	NUM
cana-2597	1	11	vol	vol	NOUN
cana-2597	1	12	32	32	NUM
cana-2597	1	13	no	no	NOUN
cana-2597	1	14	.	.	PUNCT
cana-2597	2	1	3s	3s	NUM
cana-2597	2	2	(	(	PUNCT
cana-2597	2	3	2025	2025	NUM
cana-2597	2	4	)	)	PUNCT
cana-2597	2	5	198	198	NUM
cana-2597	2	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2597	2	7	ai	ai	ADJ
cana-2597	2	8	-	-	PUNCT
cana-2597	2	9	powered	power	VERB
cana-2597	2	10	multimodal	multimodal	NOUN
cana-2597	2	11	disaster	disaster	NOUN
cana-2597	2	12	response	response	NOUN
cana-2597	2	13	enhancement	enhancement	NOUN
cana-2597	2	14	using	use	VERB
cana-2597	2	15	social	social	ADJ
cana-2597	2	16	media	medium	NOUN
cana-2597	2	17	streams	stream	NOUN
cana-2597	2	18	rohini	rohini	PROPN
cana-2597	2	19	temkar1	temkar1	PROPN
cana-2597	2	20	,	,	PUNCT
cana-2597	2	21	gresha	gresha	PROPN
cana-2597	2	22	bhatia2	bhatia2	PROPN
cana-2597	2	23	,	,	PUNCT
cana-2597	2	24	sai	sai	X
cana-2597	2	25	thikekar3	thikekar3	PROPN
cana-2597	2	26	,	,	PUNCT
cana-2597	2	27	arya	arya	PROPN
cana-2597	2	28	banavali4	banavali4	PROPN
cana-2597	2	29	,	,	PUNCT
cana-2597	2	30	yash	yash	PROPN
cana-2597	2	31	chhaproo5	chhaproo5	NUM
cana-2597	2	32	,	,	PUNCT
cana-2597	2	33	aradhya	aradhya	NOUN
cana-2597	2	34	ingle6	ingle6	NOUN
cana-2597	3	1	1,2,3,4,5,6department	1,2,3,4,5,6department	NOUN
cana-2597	3	2	of	of	ADP
cana-2597	3	3	computer	computer	NOUN
cana-2597	3	4	engineering	engineering	NOUN
cana-2597	3	5	,	,	PUNCT
cana-2597	3	6	vivekanand	vivekanand	NOUN
cana-2597	3	7	education	education	PROPN
cana-2597	3	8	society	society	PROPN
cana-2597	3	9	’s	’s	PART
cana-2597	3	10	institute	institute	PROPN
cana-2597	3	11	of	of	ADP
cana-2597	3	12	technology	technology	PROPN
cana-2597	3	13	,	,	PUNCT
cana-2597	3	14	mumbai	mumbai	PROPN
cana-2597	3	15	,	,	PUNCT
cana-2597	3	16	india	india	PROPN
cana-2597	3	17	article	article	PROPN
cana-2597	3	18	history	history	NOUN
cana-2597	3	19	:	:	PUNCT
cana-2597	3	20	received	receive	VERB
cana-2597	3	21	:	:	PUNCT
cana-2597	3	22	26	26	NUM
cana-2597	3	23	-	-	SYM
cana-2597	3	24	09	09	NUM
cana-2597	3	25	-	-	PUNCT
cana-2597	3	26	2024	2024	NUM
cana-2597	3	27	revised	revise	VERB
cana-2597	3	28	:	:	PUNCT
cana-2597	3	29	04	04	NUM
cana-2597	3	30	-	-	SYM
cana-2597	3	31	11	11	NUM
cana-2597	3	32	-	-	PUNCT
cana-2597	3	33	2024	2024	NUM
cana-2597	3	34	accepted	accept	VERB
cana-2597	3	35	:	:	PUNCT
cana-2597	3	36	19	19	NUM
cana-2597	3	37	-	-	SYM
cana-2597	3	38	11	11	NUM
cana-2597	3	39	-	-	PUNCT
cana-2597	3	40	2024	2024	NUM
cana-2597	3	41	abstract	abstract	NOUN
cana-2597	3	42	:	:	PUNCT
cana-2597	3	43	social	social	ADJ
cana-2597	3	44	media	media	NOUN
cana-2597	3	45	serves	serve	VERB
cana-2597	3	46	as	as	ADP
cana-2597	3	47	a	a	DET
cana-2597	3	48	valuable	valuable	ADJ
cana-2597	3	49	source	source	NOUN
cana-2597	3	50	of	of	ADP
cana-2597	3	51	real	real	ADJ
cana-2597	3	52	-	-	PUNCT
cana-2597	3	53	time	time	NOUN
cana-2597	3	54	disaster	disaster	NOUN
cana-2597	3	55	data	datum	NOUN
cana-2597	3	56	,	,	PUNCT
cana-2597	3	57	providing	provide	VERB
cana-2597	3	58	timely	timely	ADJ
cana-2597	3	59	updates	update	NOUN
cana-2597	3	60	that	that	PRON
cana-2597	3	61	are	be	AUX
cana-2597	3	62	critical	critical	ADJ
cana-2597	3	63	for	for	ADP
cana-2597	3	64	effective	effective	ADJ
cana-2597	3	65	disaster	disaster	NOUN
cana-2597	3	66	monitoring	monitoring	NOUN
cana-2597	3	67	and	and	CCONJ
cana-2597	3	68	response	response	NOUN
cana-2597	3	69	.	.	PUNCT
cana-2597	4	1	it	it	PRON
cana-2597	4	2	plays	play	VERB
cana-2597	4	3	an	an	DET
cana-2597	4	4	essential	essential	ADJ
cana-2597	4	5	role	role	NOUN
cana-2597	4	6	in	in	ADP
cana-2597	4	7	raising	raise	VERB
cana-2597	4	8	awareness	awareness	NOUN
cana-2597	4	9	,	,	PUNCT
cana-2597	4	10	enhancing	enhance	VERB
cana-2597	4	11	preparedness	preparedness	NOUN
cana-2597	4	12	,	,	PUNCT
cana-2597	4	13	and	and	CCONJ
cana-2597	4	14	facilitating	facilitate	VERB
cana-2597	4	15	prompt	prompt	ADJ
cana-2597	4	16	action	action	NOUN
cana-2597	4	17	during	during	ADP
cana-2597	4	18	emergencies	emergency	NOUN
cana-2597	4	19	.	.	PUNCT
cana-2597	5	1	while	while	SCONJ
cana-2597	5	2	much	much	ADJ
cana-2597	5	3	of	of	ADP
cana-2597	5	4	the	the	DET
cana-2597	5	5	current	current	ADJ
cana-2597	5	6	research	research	NOUN
cana-2597	5	7	focuses	focus	VERB
cana-2597	5	8	on	on	ADP
cana-2597	5	9	disaster	disaster	NOUN
cana-2597	5	10	prediction	prediction	NOUN
cana-2597	5	11	,	,	PUNCT
cana-2597	5	12	real	real	ADJ
cana-2597	5	13	-	-	PUNCT
cana-2597	5	14	time	time	NOUN
cana-2597	5	15	updates	update	NOUN
cana-2597	5	16	,	,	PUNCT
cana-2597	5	17	which	which	PRON
cana-2597	5	18	are	be	AUX
cana-2597	5	19	key	key	ADJ
cana-2597	5	20	to	to	ADP
cana-2597	5	21	saving	save	VERB
cana-2597	5	22	lives	life	NOUN
cana-2597	5	23	and	and	CCONJ
cana-2597	5	24	coordinating	coordinate	VERB
cana-2597	5	25	response	response	NOUN
cana-2597	5	26	efforts	effort	NOUN
cana-2597	5	27	,	,	PUNCT
cana-2597	5	28	are	be	AUX
cana-2597	5	29	often	often	ADV
cana-2597	5	30	overlooked	overlook	VERB
cana-2597	5	31	.	.	PUNCT
cana-2597	6	1	our	our	PRON
cana-2597	6	2	project	project	NOUN
cana-2597	6	3	addresses	address	VERB
cana-2597	6	4	this	this	DET
cana-2597	6	5	gap	gap	NOUN
cana-2597	6	6	by	by	ADP
cana-2597	6	7	focusing	focus	VERB
cana-2597	6	8	on	on	ADP
cana-2597	6	9	the	the	DET
cana-2597	6	10	collection	collection	NOUN
cana-2597	6	11	and	and	CCONJ
cana-2597	6	12	analysis	analysis	NOUN
cana-2597	6	13	of	of	ADP
cana-2597	6	14	real	real	ADJ
cana-2597	6	15	-	-	PUNCT
cana-2597	6	16	time	time	NOUN
cana-2597	6	17	data	datum	NOUN
cana-2597	6	18	from	from	ADP
cana-2597	6	19	various	various	ADJ
cana-2597	6	20	social	social	ADJ
cana-2597	6	21	media	medium	NOUN
cana-2597	6	22	platforms	platform	NOUN
cana-2597	6	23	,	,	PUNCT
cana-2597	6	24	including	include	VERB
cana-2597	6	25	disaster	disaster	NOUN
cana-2597	6	26	-	-	PUNCT
cana-2597	6	27	related	relate	VERB
cana-2597	6	28	images	image	NOUN
cana-2597	6	29	and	and	CCONJ
cana-2597	6	30	textual	textual	ADJ
cana-2597	6	31	data	datum	NOUN
cana-2597	6	32	like	like	SCONJ
cana-2597	6	33	tweets.we	tweets.we	X
cana-2597	6	34	use	use	VERB
cana-2597	6	35	multiple	multiple	ADJ
cana-2597	6	36	apis	apis	ADJ
cana-2597	6	37	to	to	PART
cana-2597	6	38	gather	gather	VERB
cana-2597	6	39	real	real	ADJ
cana-2597	6	40	-	-	PUNCT
cana-2597	6	41	time	time	NOUN
cana-2597	6	42	data	datum	NOUN
cana-2597	6	43	from	from	ADP
cana-2597	6	44	platforms	platform	NOUN
cana-2597	6	45	such	such	ADJ
cana-2597	6	46	as	as	ADP
cana-2597	6	47	instagram	instagram	PROPN
cana-2597	6	48	,	,	PUNCT
cana-2597	6	49	facebook	facebook	NOUN
cana-2597	6	50	,	,	PUNCT
cana-2597	6	51	and	and	CCONJ
cana-2597	6	52	twitter	twitter	NOUN
cana-2597	6	53	.	.	PUNCT
cana-2597	7	1	to	to	PART
cana-2597	7	2	analyze	analyze	VERB
cana-2597	7	3	visual	visual	ADJ
cana-2597	7	4	data	datum	NOUN
cana-2597	7	5	,	,	PUNCT
cana-2597	7	6	we	we	PRON
cana-2597	7	7	employ	employ	VERB
cana-2597	7	8	state	state	NOUN
cana-2597	7	9	-	-	PUNCT
cana-2597	7	10	of	of	ADP
cana-2597	7	11	-	-	PUNCT
cana-2597	7	12	the	the	DET
cana-2597	7	13	-	-	PUNCT
cana-2597	7	14	art	art	NOUN
cana-2597	7	15	multimodal	multimodal	NOUN
cana-2597	7	16	models	model	NOUN
cana-2597	7	17	like	like	ADP
cana-2597	7	18	resnet50	resnet50	NOUN
cana-2597	7	19	and	and	CCONJ
cana-2597	7	20	efficientnet	efficientnet	NOUN
cana-2597	7	21	for	for	ADP
cana-2597	7	22	image	image	NOUN
cana-2597	7	23	classification	classification	NOUN
cana-2597	7	24	.	.	PUNCT
cana-2597	8	1	for	for	ADP
cana-2597	8	2	the	the	DET
cana-2597	8	3	textual	textual	ADJ
cana-2597	8	4	data	datum	NOUN
cana-2597	8	5	,	,	PUNCT
cana-2597	8	6	advanced	advanced	ADJ
cana-2597	8	7	natural	natural	ADJ
cana-2597	8	8	language	language	NOUN
cana-2597	8	9	processing	processing	NOUN
cana-2597	8	10	models	model	NOUN
cana-2597	8	11	like	like	ADP
cana-2597	8	12	bert	bert	PROPN
cana-2597	8	13	and	and	CCONJ
cana-2597	8	14	xlnet	xlnet	PROPN
cana-2597	8	15	are	be	AUX
cana-2597	8	16	used	use	VERB
cana-2597	8	17	for	for	ADP
cana-2597	8	18	sentiment	sentiment	NOUN
cana-2597	8	19	analysis	analysis	NOUN
cana-2597	8	20	,	,	PUNCT
cana-2597	8	21	event	event	NOUN
cana-2597	8	22	detection	detection	NOUN
cana-2597	8	23	,	,	PUNCT
cana-2597	8	24	and	and	CCONJ
cana-2597	8	25	severity	severity	NOUN
cana-2597	8	26	assessment	assessment	NOUN
cana-2597	8	27	.	.	PUNCT
cana-2597	9	1	by	by	ADP
cana-2597	9	2	integrating	integrate	VERB
cana-2597	9	3	and	and	CCONJ
cana-2597	9	4	fusing	fuse	VERB
cana-2597	9	5	the	the	DET
cana-2597	9	6	image	image	NOUN
cana-2597	9	7	and	and	CCONJ
cana-2597	9	8	text	text	NOUN
cana-2597	9	9	data	datum	NOUN
cana-2597	9	10	,	,	PUNCT
cana-2597	9	11	our	our	PRON
cana-2597	9	12	system	system	NOUN
cana-2597	9	13	is	be	AUX
cana-2597	9	14	able	able	ADJ
cana-2597	9	15	to	to	PART
cana-2597	9	16	identify	identify	VERB
cana-2597	9	17	the	the	DET
cana-2597	9	18	type	type	NOUN
cana-2597	9	19	of	of	ADP
cana-2597	9	20	disaster	disaster	NOUN
cana-2597	9	21	,	,	PUNCT
cana-2597	9	22	estimate	estimate	VERB
cana-2597	9	23	its	its	PRON
cana-2597	9	24	severity	severity	NOUN
cana-2597	9	25	,	,	PUNCT
cana-2597	9	26	and	and	CCONJ
cana-2597	9	27	provide	provide	VERB
cana-2597	9	28	actionable	actionable	ADJ
cana-2597	9	29	insights	insight	NOUN
cana-2597	9	30	.	.	PUNCT
cana-2597	10	1	keywords	keyword	NOUN
cana-2597	10	2	:	:	PUNCT
cana-2597	10	3	real	real	ADJ
cana-2597	10	4	-	-	PUNCT
cana-2597	10	5	time	time	NOUN
cana-2597	10	6	disaster	disaster	NOUN
cana-2597	10	7	data	datum	NOUN
cana-2597	10	8	,	,	PUNCT
cana-2597	10	9	apis	apis	ADJ
cana-2597	10	10	,	,	PUNCT
cana-2597	10	11	multimodal	multimodal	NOUN
cana-2597	10	12	,	,	PUNCT
cana-2597	10	13	resnet50	resnet50	NOUN
cana-2597	10	14	,	,	PUNCT
cana-2597	10	15	efficientnet	efficientnet	NOUN
cana-2597	10	16	,	,	PUNCT
cana-2597	10	17	bert	bert	PROPN
cana-2597	10	18	,	,	PUNCT
cana-2597	10	19	xlnet	xlnet	PROPN
cana-2597	10	20	,	,	PUNCT
cana-2597	10	21	disaster	disaster	NOUN
cana-2597	10	22	severity	severity	NOUN
cana-2597	10	23	assessment	assessment	NOUN
cana-2597	10	24	.	.	PUNCT
cana-2597	11	1	i.	i.	PROPN
cana-2597	11	2	introduction	introduction	NOUN
cana-2597	11	3	natural	natural	ADJ
cana-2597	11	4	disasters	disaster	NOUN
cana-2597	11	5	,	,	PUNCT
cana-2597	11	6	including	include	VERB
cana-2597	11	7	tsunamis	tsunamis	NOUN
cana-2597	11	8	,	,	PUNCT
cana-2597	11	9	floods	flood	NOUN
cana-2597	11	10	,	,	PUNCT
cana-2597	11	11	and	and	CCONJ
cana-2597	11	12	earthquakes	earthquake	NOUN
cana-2597	11	13	,	,	PUNCT
cana-2597	11	14	are	be	AUX
cana-2597	11	15	destructive	destructive	ADJ
cana-2597	11	16	and	and	CCONJ
cana-2597	11	17	often	often	ADV
cana-2597	11	18	beyond	beyond	ADP
cana-2597	11	19	human	human	ADJ
cana-2597	11	20	control	control	NOUN
cana-2597	11	21	.	.	PUNCT
cana-2597	12	1	however	however	ADV
cana-2597	12	2	,	,	PUNCT
cana-2597	12	3	disaster	disaster	NOUN
cana-2597	12	4	management	management	NOUN
cana-2597	12	5	offers	offer	VERB
cana-2597	12	6	a	a	DET
cana-2597	12	7	structured	structured	ADJ
cana-2597	12	8	approach	approach	NOUN
cana-2597	12	9	to	to	PART
cana-2597	12	10	mitigate	mitigate	VERB
cana-2597	12	11	the	the	DET
cana-2597	12	12	devastating	devastating	ADJ
cana-2597	12	13	impacts	impact	NOUN
cana-2597	12	14	of	of	ADP
cana-2597	12	15	such	such	ADJ
cana-2597	12	16	events	event	NOUN
cana-2597	12	17	through	through	ADP
cana-2597	12	18	preparedness	preparedness	NOUN
cana-2597	12	19	,	,	PUNCT
cana-2597	12	20	response	response	NOUN
cana-2597	12	21	,	,	PUNCT
cana-2597	12	22	and	and	CCONJ
cana-2597	12	23	recovery	recovery	NOUN
cana-2597	12	24	strategies	strategy	NOUN
cana-2597	12	25	.	.	PUNCT
cana-2597	13	1	india	india	PROPN
cana-2597	13	2	,	,	PUNCT
cana-2597	13	3	in	in	ADP
cana-2597	13	4	particular	particular	ADJ
cana-2597	13	5	,	,	PUNCT
cana-2597	13	6	ranks	rank	VERB
cana-2597	13	7	third	third	ADV
cana-2597	13	8	globally	globally	ADV
cana-2597	13	9	in	in	ADP
cana-2597	13	10	human	human	ADJ
cana-2597	13	11	mortality	mortality	NOUN
cana-2597	13	12	from	from	ADP
cana-2597	13	13	extreme	extreme	ADJ
cana-2597	13	14	weather	weather	NOUN
cana-2597	13	15	events	event	NOUN
cana-2597	13	16	between	between	ADP
cana-2597	13	17	2000	2000	NUM
cana-2597	13	18	and	and	CCONJ
cana-2597	13	19	2019	2019	NUM
cana-2597	13	20	.	.	PUNCT
cana-2597	14	1	over	over	ADP
cana-2597	14	2	the	the	DET
cana-2597	14	3	period	period	NOUN
cana-2597	14	4	from	from	ADP
cana-2597	14	5	1983	1983	NUM
cana-2597	14	6	to	to	ADP
cana-2597	14	7	2011	2011	NUM
cana-2597	14	8	,	,	PUNCT
cana-2597	14	9	the	the	DET
cana-2597	14	10	country	country	NOUN
cana-2597	14	11	faced	face	VERB
cana-2597	14	12	190	190	NUM
cana-2597	14	13	floods	flood	NOUN
cana-2597	14	14	and	and	CCONJ
cana-2597	14	15	54	54	NUM
cana-2597	14	16	cyclones	cyclone	NOUN
cana-2597	14	17	,	,	PUNCT
cana-2597	14	18	leading	lead	VERB
cana-2597	14	19	to	to	ADP
cana-2597	14	20	severe	severe	ADJ
cana-2597	14	21	human	human	ADJ
cana-2597	14	22	and	and	CCONJ
cana-2597	14	23	economic	economic	ADJ
cana-2597	14	24	losses	loss	NOUN
cana-2597	14	25	,	,	PUNCT
cana-2597	14	26	with	with	ADP
cana-2597	14	27	approximately	approximately	ADV
cana-2597	14	28	60,919	60,919	NUM
cana-2597	14	29	deaths	death	NOUN
cana-2597	14	30	and	and	CCONJ
cana-2597	14	31	1.22	1.22	NUM
cana-2597	14	32	billion	billion	NUM
cana-2597	14	33	individuals	individual	NOUN
cana-2597	14	34	affected	affect	VERB
cana-2597	14	35	by	by	ADP
cana-2597	14	36	floods	flood	NOUN
cana-2597	14	37	,	,	PUNCT
cana-2597	14	38	and	and	CCONJ
cana-2597	14	39	20,360	20,360	NUM
cana-2597	14	40	deaths	death	NOUN
cana-2597	14	41	and	and	CCONJ
cana-2597	14	42	56	56	NUM
cana-2597	14	43	million	million	NUM
cana-2597	14	44	people	people	NOUN
cana-2597	14	45	impacted	impact	VERB
cana-2597	14	46	by	by	ADP
cana-2597	14	47	cyclones	cyclone	NOUN
cana-2597	14	48	.	.	PUNCT
cana-2597	15	1	these	these	DET
cana-2597	15	2	statistics	statistic	NOUN
cana-2597	15	3	emphasize	emphasize	VERB
cana-2597	15	4	the	the	DET
cana-2597	15	5	critical	critical	ADJ
cana-2597	15	6	vulnerability	vulnerability	NOUN
cana-2597	15	7	of	of	ADP
cana-2597	15	8	india	india	PROPN
cana-2597	15	9	to	to	ADP
cana-2597	15	10	natural	natural	ADJ
cana-2597	15	11	disasters	disaster	NOUN
cana-2597	15	12	and	and	CCONJ
cana-2597	15	13	the	the	DET
cana-2597	15	14	urgent	urgent	ADJ
cana-2597	15	15	need	need	NOUN
cana-2597	15	16	for	for	ADP
cana-2597	15	17	more	more	ADV
cana-2597	15	18	effective	effective	ADJ
cana-2597	15	19	disaster	disaster	NOUN
cana-2597	15	20	management	management	NOUN
cana-2597	15	21	and	and	CCONJ
cana-2597	15	22	resilient	resilient	ADJ
cana-2597	15	23	infrastructure	infrastructure	NOUN
cana-2597	15	24	systems	system	NOUN
cana-2597	15	25	to	to	PART
cana-2597	15	26	reduce	reduce	VERB
cana-2597	15	27	future	future	ADJ
cana-2597	15	28	risks	risk	NOUN
cana-2597	15	29	[	[	X
cana-2597	15	30	1	1	NUM
cana-2597	15	31	]	]	PUNCT
cana-2597	15	32	.	.	PUNCT
cana-2597	16	1	in	in	ADP
cana-2597	16	2	recent	recent	ADJ
cana-2597	16	3	years	year	NOUN
cana-2597	16	4	,	,	PUNCT
cana-2597	16	5	social	social	ADJ
cana-2597	16	6	media	medium	NOUN
cana-2597	16	7	platforms	platform	NOUN
cana-2597	16	8	,	,	PUNCT
cana-2597	16	9	such	such	ADJ
cana-2597	16	10	as	as	ADP
cana-2597	16	11	twitter	twitter	NOUN
cana-2597	16	12	,	,	PUNCT
cana-2597	16	13	instagram	instagram	PROPN
cana-2597	16	14	,	,	PUNCT
cana-2597	16	15	and	and	CCONJ
cana-2597	16	16	facebook	facebook	NOUN
cana-2597	16	17	,	,	PUNCT
cana-2597	16	18	have	have	AUX
cana-2597	16	19	played	play	VERB
cana-2597	16	20	an	an	DET
cana-2597	16	21	increasingly	increasingly	ADV
cana-2597	16	22	important	important	ADJ
cana-2597	16	23	role	role	NOUN
cana-2597	16	24	in	in	ADP
cana-2597	16	25	disaster	disaster	NOUN
cana-2597	16	26	management	management	NOUN
cana-2597	16	27	by	by	ADP
cana-2597	16	28	providing	provide	VERB
cana-2597	16	29	real	real	ADJ
cana-2597	16	30	-	-	PUNCT
cana-2597	16	31	time	time	NOUN
cana-2597	16	32	data	datum	NOUN
cana-2597	16	33	that	that	PRON
cana-2597	16	34	enhances	enhance	VERB
cana-2597	16	35	situational	situational	ADJ
cana-2597	16	36	awareness	awareness	NOUN
cana-2597	16	37	and	and	CCONJ
cana-2597	16	38	facilitates	facilitate	VERB
cana-2597	16	39	rapid	rapid	ADJ
cana-2597	16	40	emergency	emergency	NOUN
cana-2597	16	41	response	response	NOUN
cana-2597	16	42	.	.	PUNCT
cana-2597	17	1	the	the	DET
cana-2597	17	2	integration	integration	NOUN
cana-2597	17	3	of	of	ADP
cana-2597	17	4	social	social	ADJ
cana-2597	17	5	media	medium	NOUN
cana-2597	17	6	analytics	analytic	NOUN
cana-2597	17	7	into	into	ADP
cana-2597	17	8	disaster	disaster	NOUN
cana-2597	17	9	management	management	NOUN
cana-2597	17	10	frameworks	framework	NOUN
cana-2597	17	11	has	have	AUX
cana-2597	17	12	revolutionized	revolutionize	VERB
cana-2597	17	13	how	how	SCONJ
cana-2597	17	14	information	information	NOUN
cana-2597	17	15	is	be	AUX
cana-2597	17	16	collected	collect	VERB
cana-2597	17	17	,	,	PUNCT
cana-2597	17	18	processed	process	VERB
cana-2597	17	19	,	,	PUNCT
cana-2597	17	20	and	and	CCONJ
cana-2597	17	21	acted	act	VERB
cana-2597	17	22	upon	upon	SCONJ
cana-2597	17	23	during	during	ADP
cana-2597	17	24	crises	crisis	NOUN
cana-2597	17	25	.	.	PUNCT
cana-2597	18	1	social	social	ADJ
cana-2597	18	2	media	medium	NOUN
cana-2597	18	3	's	's	PART
cana-2597	18	4	ability	ability	NOUN
cana-2597	18	5	to	to	PART
cana-2597	18	6	provide	provide	VERB
cana-2597	18	7	up	up	ADP
cana-2597	18	8	-	-	PUNCT
cana-2597	18	9	to	to	ADP
cana-2597	18	10	-	-	PUNCT
cana-2597	18	11	date	date	NOUN
cana-2597	18	12	,	,	PUNCT
cana-2597	18	13	location	location	NOUN
cana-2597	18	14	-	-	PUNCT
cana-2597	18	15	specific	specific	ADJ
cana-2597	18	16	data	datum	NOUN
cana-2597	18	17	enables	enable	VERB
cana-2597	18	18	authorities	authority	NOUN
cana-2597	18	19	and	and	CCONJ
cana-2597	18	20	responders	responder	NOUN
cana-2597	18	21	to	to	PART
cana-2597	18	22	monitor	monitor	VERB
cana-2597	18	23	dynamic	dynamic	ADJ
cana-2597	18	24	situations	situation	NOUN
cana-2597	18	25	and	and	CCONJ
cana-2597	18	26	coordinate	coordinate	VERB
cana-2597	18	27	more	more	ADV
cana-2597	18	28	efficient	efficient	ADJ
cana-2597	18	29	responses	response	NOUN
cana-2597	18	30	[	[	X
cana-2597	18	31	2	2	NUM
cana-2597	18	32	]	]	PUNCT
cana-2597	18	33	.	.	PUNCT
cana-2597	19	1	communications	communication	NOUN
cana-2597	19	2	on	on	ADP
cana-2597	19	3	applied	apply	VERB
cana-2597	19	4	nonlinear	nonlinear	ADJ
cana-2597	19	5	analysis	analysis	NOUN
cana-2597	19	6	issn	issn	NOUN
cana-2597	19	7	:	:	PUNCT
cana-2597	19	8	1074	1074	NUM
cana-2597	19	9	-	-	PUNCT
cana-2597	19	10	133x	133x	NUM
cana-2597	19	11	vol	vol	NOUN
cana-2597	19	12	32	32	NUM
cana-2597	19	13	no	no	NOUN
cana-2597	19	14	.	.	PUNCT
cana-2597	20	1	3s	3s	NUM
cana-2597	20	2	(	(	PUNCT
cana-2597	20	3	2025	2025	NUM
cana-2597	20	4	)	)	PUNCT
cana-2597	20	5	199	199	NUM
cana-2597	20	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2597	20	7	further	far	ADV
cana-2597	20	8	,	,	PUNCT
cana-2597	20	9	multimodal	multimodal	ADJ
cana-2597	20	10	data	datum	NOUN
cana-2597	20	11	—	—	PUNCT
cana-2597	20	12	text	text	NOUN
cana-2597	20	13	,	,	PUNCT
cana-2597	20	14	images	image	NOUN
cana-2597	20	15	,	,	PUNCT
cana-2597	20	16	and	and	CCONJ
cana-2597	20	17	videos	video	NOUN
cana-2597	20	18	—	—	PUNCT
cana-2597	20	19	has	have	AUX
cana-2597	20	20	proven	prove	VERB
cana-2597	20	21	essential	essential	ADJ
cana-2597	20	22	for	for	ADP
cana-2597	20	23	enhancing	enhance	VERB
cana-2597	20	24	the	the	DET
cana-2597	20	25	precision	precision	NOUN
cana-2597	20	26	of	of	ADP
cana-2597	20	27	urban	urban	ADJ
cana-2597	20	28	event	event	NOUN
cana-2597	20	29	monitoring	monitoring	NOUN
cana-2597	20	30	.	.	PUNCT
cana-2597	21	1	for	for	ADP
cana-2597	21	2	instance	instance	NOUN
cana-2597	21	3	,	,	PUNCT
cana-2597	21	4	[	[	X
cana-2597	21	5	2	2	NUM
cana-2597	21	6	]	]	PUNCT
cana-2597	21	7	and	and	CCONJ
cana-2597	21	8	related	related	ADJ
cana-2597	21	9	studies	study	NOUN
cana-2597	21	10	demonstrate	demonstrate	VERB
cana-2597	21	11	that	that	SCONJ
cana-2597	21	12	combining	combine	VERB
cana-2597	21	13	various	various	ADJ
cana-2597	21	14	data	datum	NOUN
cana-2597	21	15	types	type	NOUN
cana-2597	21	16	significantly	significantly	ADV
cana-2597	21	17	improves	improve	VERB
cana-2597	21	18	the	the	DET
cana-2597	21	19	spatial	spatial	ADJ
cana-2597	21	20	accuracy	accuracy	NOUN
cana-2597	21	21	of	of	ADP
cana-2597	21	22	event	event	NOUN
cana-2597	21	23	detection	detection	NOUN
cana-2597	21	24	,	,	PUNCT
cana-2597	21	25	such	such	ADJ
cana-2597	21	26	as	as	ADP
cana-2597	21	27	urban	urban	ADJ
cana-2597	21	28	flooding	flooding	NOUN
cana-2597	21	29	,	,	PUNCT
cana-2597	21	30	achieving	achieve	VERB
cana-2597	21	31	up	up	ADP
cana-2597	21	32	to	to	PART
cana-2597	21	33	100	100	NUM
cana-2597	21	34	%	%	NOUN
cana-2597	21	35	precision	precision	NOUN
cana-2597	21	36	in	in	ADP
cana-2597	21	37	specific	specific	ADJ
cana-2597	21	38	instances	instance	NOUN
cana-2597	21	39	(	(	PUNCT
cana-2597	21	40	"	"	PUNCT
cana-2597	21	41	a	a	DET
cana-2597	21	42	spacel	spacel	PROPN
cana-2597	21	43	information	information	NOUN
cana-2597	21	44	extraction	extraction	NOUN
cana-2597	21	45	method	method	NOUN
cana-2597	21	46	based	base	VERB
cana-2597	21	47	on	on	ADP
cana-2597	21	48	multimodal	multimodal	ADJ
cana-2597	21	49	social	social	ADJ
cana-2597	21	50	media	medium	NOUN
cana-2597	21	51	data	datum	NOUN
cana-2597	21	52	:	:	PUNCT
cana-2597	21	53	a	a	DET
cana-2597	21	54	case	case	NOUN
cana-2597	21	55	study	study	NOUN
cana-2597	21	56	on	on	ADP
cana-2597	21	57	urban	urban	ADJ
cana-2597	21	58	inundation	inundation	NOUN
cana-2597	21	59	,	,	PUNCT
cana-2597	21	60	"	"	PUNCT
cana-2597	21	61	2023	2023	NUM
cana-2597	21	62	)	)	PUNCT
cana-2597	21	63	.	.	PUNCT
cana-2597	22	1	moreover	moreover	ADV
cana-2597	22	2	,	,	PUNCT
cana-2597	22	3	chatterjee	chatterjee	PROPN
cana-2597	22	4	[	[	X
cana-2597	22	5	3	3	X
cana-2597	22	6	]	]	X
cana-2597	22	7	propose	propose	VERB
cana-2597	22	8	an	an	DET
cana-2597	22	9	innovative	innovative	ADJ
cana-2597	22	10	approach	approach	NOUN
cana-2597	22	11	for	for	ADP
cana-2597	22	12	monitoring	monitor	VERB
cana-2597	22	13	mental	mental	ADJ
cana-2597	22	14	health	health	NOUN
cana-2597	22	15	through	through	ADP
cana-2597	22	16	social	social	ADJ
cana-2597	22	17	media	medium	NOUN
cana-2597	22	18	posts	post	NOUN
cana-2597	22	19	,	,	PUNCT
cana-2597	22	20	using	use	VERB
cana-2597	22	21	sentiment	sentiment	NOUN
cana-2597	22	22	analysis	analysis	NOUN
cana-2597	22	23	and	and	CCONJ
cana-2597	22	24	multimodal	multimodal	NOUN
cana-2597	22	25	features	feature	NOUN
cana-2597	22	26	to	to	PART
cana-2597	22	27	construct	construct	VERB
cana-2597	22	28	a	a	DET
cana-2597	22	29	mental	mental	ADJ
cana-2597	22	30	health	health	NOUN
cana-2597	22	31	index	index	NOUN
cana-2597	22	32	,	,	PUNCT
cana-2597	22	33	achieving	achieve	VERB
cana-2597	22	34	89	89	NUM
cana-2597	22	35	%	%	NOUN
cana-2597	22	36	accuracy	accuracy	NOUN
cana-2597	22	37	with	with	ADP
cana-2597	22	38	support	support	NOUN
cana-2597	22	39	vector	vector	NOUN
cana-2597	22	40	machines	machine	NOUN
cana-2597	22	41	(	(	PUNCT
cana-2597	22	42	svm	svm	PROPN
cana-2597	22	43	)	)	PUNCT
cana-2597	22	44	.	.	PUNCT
cana-2597	23	1	this	this	DET
cana-2597	23	2	real	real	ADJ
cana-2597	23	3	-	-	PUNCT
cana-2597	23	4	time	time	NOUN
cana-2597	23	5	analysis	analysis	NOUN
cana-2597	23	6	has	have	VERB
cana-2597	23	7	profound	profound	ADJ
cana-2597	23	8	implications	implication	NOUN
cana-2597	23	9	for	for	ADP
cana-2597	23	10	disaster	disaster	NOUN
cana-2597	23	11	response	response	NOUN
cana-2597	23	12	,	,	PUNCT
cana-2597	23	13	as	as	SCONJ
cana-2597	23	14	psychological	psychological	ADJ
cana-2597	23	15	well	well	ADV
cana-2597	23	16	-	-	PUNCT
cana-2597	23	17	being	being	NOUN
cana-2597	23	18	is	be	AUX
cana-2597	23	19	a	a	DET
cana-2597	23	20	critical	critical	ADJ
cana-2597	23	21	aspect	aspect	NOUN
cana-2597	23	22	of	of	ADP
cana-2597	23	23	managing	manage	VERB
cana-2597	23	24	the	the	DET
cana-2597	23	25	human	human	ADJ
cana-2597	23	26	impact	impact	NOUN
cana-2597	23	27	of	of	ADP
cana-2597	23	28	disasters	disaster	NOUN
cana-2597	23	29	.	.	PUNCT
cana-2597	24	1	disaster	disaster	NOUN
cana-2597	24	2	response	response	NOUN
cana-2597	24	3	can	can	AUX
cana-2597	24	4	also	also	ADV
cana-2597	24	5	be	be	AUX
cana-2597	24	6	enhanced	enhance	VERB
cana-2597	24	7	by	by	ADP
cana-2597	24	8	employing	employ	VERB
cana-2597	24	9	advanced	advanced	ADJ
cana-2597	24	10	natural	natural	ADJ
cana-2597	24	11	language	language	NOUN
cana-2597	24	12	processing	processing	NOUN
cana-2597	24	13	models	model	NOUN
cana-2597	24	14	like	like	ADP
cana-2597	24	15	late	late	ADJ
cana-2597	24	16	dirichlet	dirichlet	PROPN
cana-2597	24	17	allocation	allocation	NOUN
cana-2597	24	18	and	and	CCONJ
cana-2597	24	19	bert	bert	NOUN
cana-2597	24	20	.	.	PUNCT
cana-2597	25	1	[	[	X
cana-2597	25	2	2	2	X
cana-2597	25	3	]	]	AUX
cana-2597	25	4	show	show	VERB
cana-2597	25	5	that	that	SCONJ
cana-2597	25	6	real	real	ADJ
cana-2597	25	7	-	-	PUNCT
cana-2597	25	8	time	time	NOUN
cana-2597	25	9	classification	classification	NOUN
cana-2597	25	10	of	of	ADP
cana-2597	25	11	disaster	disaster	NOUN
cana-2597	25	12	-	-	PUNCT
cana-2597	25	13	related	relate	VERB
cana-2597	25	14	topics	topic	NOUN
cana-2597	25	15	using	use	VERB
cana-2597	25	16	such	such	ADJ
cana-2597	25	17	models	model	NOUN
cana-2597	25	18	improves	improve	VERB
cana-2597	25	19	accuracy	accuracy	NOUN
cana-2597	25	20	by	by	ADP
cana-2597	25	21	12	12	NUM
cana-2597	25	22	%	%	NOUN
cana-2597	25	23	compared	compare	VERB
cana-2597	25	24	to	to	ADP
cana-2597	25	25	earlier	early	ADJ
cana-2597	25	26	techniques	technique	NOUN
cana-2597	25	27	.	.	PUNCT
cana-2597	26	1	despite	despite	SCONJ
cana-2597	26	2	these	these	DET
cana-2597	26	3	advancements	advancement	NOUN
cana-2597	26	4	,	,	PUNCT
cana-2597	26	5	challenges	challenge	NOUN
cana-2597	26	6	remain	remain	VERB
cana-2597	26	7	,	,	PUNCT
cana-2597	26	8	especially	especially	ADV
cana-2597	26	9	in	in	ADP
cana-2597	26	10	managing	manage	VERB
cana-2597	26	11	the	the	DET
cana-2597	26	12	noisy	noisy	ADJ
cana-2597	26	13	and	and	CCONJ
cana-2597	26	14	unstructured	unstructured	ADJ
cana-2597	26	15	nature	nature	NOUN
cana-2597	26	16	of	of	ADP
cana-2597	26	17	social	social	ADJ
cana-2597	26	18	media	medium	NOUN
cana-2597	26	19	data	datum	NOUN
cana-2597	26	20	,	,	PUNCT
cana-2597	26	21	which	which	PRON
cana-2597	26	22	can	can	AUX
cana-2597	26	23	affect	affect	VERB
cana-2597	26	24	the	the	DET
cana-2597	26	25	reliability	reliability	NOUN
cana-2597	26	26	of	of	ADP
cana-2597	26	27	insights	insight	NOUN
cana-2597	26	28	.	.	PUNCT
cana-2597	27	1	ensuring	ensure	VERB
cana-2597	27	2	a	a	DET
cana-2597	27	3	balance	balance	NOUN
cana-2597	27	4	between	between	ADP
cana-2597	27	5	high	high	ADJ
cana-2597	27	6	accuracy	accuracy	NOUN
cana-2597	27	7	and	and	CCONJ
cana-2597	27	8	the	the	DET
cana-2597	27	9	unpredictability	unpredictability	NOUN
cana-2597	27	10	of	of	ADP
cana-2597	27	11	social	social	ADJ
cana-2597	27	12	media	medium	NOUN
cana-2597	27	13	remains	remain	VERB
cana-2597	27	14	a	a	DET
cana-2597	27	15	vital	vital	ADJ
cana-2597	27	16	area	area	NOUN
cana-2597	27	17	for	for	ADP
cana-2597	27	18	ongoing	ongoing	ADJ
cana-2597	27	19	research	research	NOUN
cana-2597	27	20	.	.	PUNCT
cana-2597	28	1	ii	ii	X
cana-2597	28	2	.	.	PROPN
cana-2597	28	3	related	relate	VERB
cana-2597	28	4	work	work	NOUN
cana-2597	28	5	efficient	efficient	ADJ
cana-2597	28	6	data	datum	NOUN
cana-2597	28	7	collection	collection	NOUN
cana-2597	28	8	and	and	CCONJ
cana-2597	28	9	aggregation	aggregation	NOUN
cana-2597	28	10	from	from	ADP
cana-2597	28	11	diverse	diverse	ADJ
cana-2597	28	12	sources	source	NOUN
cana-2597	28	13	,	,	PUNCT
cana-2597	28	14	such	such	ADJ
cana-2597	28	15	as	as	ADP
cana-2597	28	16	social	social	ADJ
cana-2597	28	17	media	medium	NOUN
cana-2597	28	18	and	and	CCONJ
cana-2597	28	19	news	news	NOUN
cana-2597	28	20	outlets	outlet	NOUN
cana-2597	28	21	,	,	PUNCT
cana-2597	28	22	play	play	VERB
cana-2597	28	23	a	a	DET
cana-2597	28	24	crucial	crucial	ADJ
cana-2597	28	25	role	role	NOUN
cana-2597	28	26	in	in	ADP
cana-2597	28	27	real	real	ADJ
cana-2597	28	28	-	-	PUNCT
cana-2597	28	29	time	time	NOUN
cana-2597	28	30	disaster	disaster	NOUN
cana-2597	28	31	management	management	NOUN
cana-2597	28	32	systems	system	NOUN
cana-2597	28	33	.	.	PUNCT
cana-2597	29	1	in	in	ADP
cana-2597	29	2	recent	recent	ADJ
cana-2597	29	3	work	work	NOUN
cana-2597	29	4	,	,	PUNCT
cana-2597	29	5	deepscraper	deepscraper	NOUN
cana-2597	29	6	by	by	ADP
cana-2597	29	7	jaebeom	jaebeom	PROPN
cana-2597	30	1	[	[	X
cana-2597	30	2	4	4	X
cana-2597	30	3	]	]	PUNCT
cana-2597	30	4	demonstrated	demonstrate	VERB
cana-2597	30	5	a	a	DET
cana-2597	30	6	novel	novel	ADJ
cana-2597	30	7	approach	approach	NOUN
cana-2597	30	8	to	to	ADP
cana-2597	30	9	tweet	tweet	NOUN
cana-2597	30	10	scraping	scrape	VERB
cana-2597	30	11	using	use	VERB
cana-2597	30	12	authenticated	authenticate	VERB
cana-2597	30	13	multiprocessing	multiprocessing	NOUN
cana-2597	30	14	,	,	PUNCT
cana-2597	30	15	significantly	significantly	ADV
cana-2597	30	16	improving	improve	VERB
cana-2597	30	17	the	the	DET
cana-2597	30	18	collection	collection	NOUN
cana-2597	30	19	speed	speed	NOUN
cana-2597	30	20	and	and	CCONJ
cana-2597	30	21	volume	volume	NOUN
cana-2597	30	22	of	of	ADP
cana-2597	30	23	social	social	ADJ
cana-2597	30	24	media	medium	NOUN
cana-2597	30	25	data	datum	NOUN
cana-2597	30	26	.	.	PUNCT
cana-2597	31	1	this	this	DET
cana-2597	31	2	method	method	NOUN
cana-2597	31	3	successfully	successfully	ADV
cana-2597	31	4	scraped	scrape	VERB
cana-2597	31	5	over	over	ADP
cana-2597	31	6	5	5	NUM
cana-2597	31	7	million	million	NUM
cana-2597	31	8	tweets	tweet	NOUN
cana-2597	31	9	,	,	PUNCT
cana-2597	31	10	outperforming	outperform	VERB
cana-2597	31	11	the	the	DET
cana-2597	31	12	standard	standard	ADJ
cana-2597	31	13	twitter	twitter	NOUN
cana-2597	31	14	api	api	NOUN
cana-2597	31	15	by	by	ADP
cana-2597	31	16	23.7	23.7	NUM
cana-2597	31	17	times	time	NOUN
cana-2597	31	18	,	,	PUNCT
cana-2597	31	19	making	make	VERB
cana-2597	31	20	it	it	PRON
cana-2597	31	21	highly	highly	ADV
cana-2597	31	22	applicable	applicable	ADJ
cana-2597	31	23	for	for	ADP
cana-2597	31	24	the	the	DET
cana-2597	31	25	social	social	ADJ
cana-2597	31	26	media	medium	NOUN
cana-2597	31	27	data	datum	NOUN
cana-2597	31	28	collection	collection	NOUN
cana-2597	31	29	component	component	NOUN
cana-2597	31	30	of	of	ADP
cana-2597	31	31	disaster	disaster	NOUN
cana-2597	31	32	response	response	NOUN
cana-2597	31	33	systems	system	NOUN
cana-2597	31	34	.	.	PUNCT
cana-2597	32	1	to	to	PART
cana-2597	32	2	complement	complement	VERB
cana-2597	32	3	social	social	ADJ
cana-2597	32	4	media	medium	NOUN
cana-2597	32	5	data	datum	NOUN
cana-2597	32	6	collection	collection	NOUN
cana-2597	32	7	,	,	PUNCT
cana-2597	32	8	the	the	DET
cana-2597	32	9	aggregation	aggregation	NOUN
cana-2597	32	10	of	of	ADP
cana-2597	32	11	disaster	disaster	NOUN
cana-2597	32	12	-	-	PUNCT
cana-2597	32	13	related	relate	VERB
cana-2597	32	14	news	news	NOUN
cana-2597	32	15	is	be	AUX
cana-2597	32	16	essential	essential	ADJ
cana-2597	32	17	for	for	ADP
cana-2597	32	18	a	a	DET
cana-2597	32	19	comprehensive	comprehensive	ADJ
cana-2597	32	20	crisis	crisis	NOUN
cana-2597	32	21	management	management	NOUN
cana-2597	32	22	system	system	NOUN
cana-2597	32	23	.	.	PUNCT
cana-2597	33	1	domala	domala	NOUN
cana-2597	34	1	[	[	X
cana-2597	34	2	5	5	NUM
cana-2597	34	3	]	]	PUNCT
cana-2597	34	4	proposed	propose	VERB
cana-2597	34	5	a	a	DET
cana-2597	34	6	system	system	NOUN
cana-2597	34	7	that	that	PRON
cana-2597	34	8	automates	automate	VERB
cana-2597	34	9	the	the	DET
cana-2597	34	10	collection	collection	NOUN
cana-2597	34	11	of	of	ADP
cana-2597	34	12	disaster	disaster	NOUN
cana-2597	34	13	news	news	NOUN
cana-2597	34	14	using	use	VERB
cana-2597	34	15	web	web	NOUN
cana-2597	34	16	scraping	scrape	VERB
cana-2597	34	17	techniques	technique	NOUN
cana-2597	34	18	and	and	CCONJ
cana-2597	34	19	classifies	classify	VERB
cana-2597	34	20	the	the	DET
cana-2597	34	21	gathered	gather	VERB
cana-2597	34	22	data	datum	NOUN
cana-2597	34	23	through	through	ADP
cana-2597	34	24	machine	machine	NOUN
cana-2597	34	25	learning	learning	NOUN
cana-2597	34	26	models	model	NOUN
cana-2597	34	27	such	such	ADJ
cana-2597	34	28	as	as	ADP
cana-2597	34	29	support	support	NOUN
cana-2597	34	30	vector	vector	NOUN
cana-2597	34	31	machines	machine	NOUN
cana-2597	34	32	(	(	PUNCT
cana-2597	34	33	svm	svm	PROPN
cana-2597	34	34	)	)	PUNCT
cana-2597	34	35	and	and	CCONJ
cana-2597	34	36	logistic	logistic	ADJ
cana-2597	34	37	regression	regression	NOUN
cana-2597	34	38	.	.	PUNCT
cana-2597	35	1	this	this	DET
cana-2597	35	2	system	system	NOUN
cana-2597	35	3	also	also	ADV
cana-2597	35	4	incorporates	incorporate	VERB
cana-2597	35	5	natural	natural	ADJ
cana-2597	35	6	language	language	NOUN
cana-2597	35	7	processing	processing	NOUN
cana-2597	35	8	(	(	PUNCT
cana-2597	35	9	nlp	nlp	NOUN
cana-2597	35	10	)	)	PUNCT
cana-2597	35	11	to	to	PART
cana-2597	35	12	preprocess	preprocess	VERB
cana-2597	35	13	the	the	DET
cana-2597	35	14	news	news	NOUN
cana-2597	35	15	data	datum	NOUN
cana-2597	35	16	and	and	CCONJ
cana-2597	35	17	geoparses	geoparse	VERB
cana-2597	35	18	relevant	relevant	ADJ
cana-2597	35	19	disaster	disaster	NOUN
cana-2597	35	20	locations	location	NOUN
cana-2597	35	21	.	.	PUNCT
cana-2597	36	1	with	with	ADP
cana-2597	36	2	an	an	DET
cana-2597	36	3	evaluation	evaluation	NOUN
cana-2597	36	4	based	base	VERB
cana-2597	36	5	on	on	ADP
cana-2597	36	6	precision	precision	NOUN
cana-2597	36	7	,	,	PUNCT
cana-2597	36	8	recall	recall	NOUN
cana-2597	36	9	,	,	PUNCT
cana-2597	36	10	f1	f1	NOUN
cana-2597	36	11	score	score	NOUN
cana-2597	36	12	,	,	PUNCT
cana-2597	36	13	and	and	CCONJ
cana-2597	36	14	a	a	DET
cana-2597	36	15	confusion	confusion	NOUN
cana-2597	36	16	matrix	matrix	NOUN
cana-2597	36	17	,	,	PUNCT
cana-2597	36	18	the	the	DET
cana-2597	36	19	approach	approach	NOUN
cana-2597	36	20	ensures	ensure	VERB
cana-2597	36	21	that	that	SCONJ
cana-2597	36	22	only	only	ADV
cana-2597	36	23	relevant	relevant	ADJ
cana-2597	36	24	news	news	NOUN
cana-2597	36	25	is	be	AUX
cana-2597	36	26	included	include	VERB
cana-2597	36	27	in	in	ADP
cana-2597	36	28	crisis	crisis	NOUN
cana-2597	36	29	management	management	NOUN
cana-2597	36	30	workflows	workflow	NOUN
cana-2597	36	31	.	.	PUNCT
cana-2597	37	1	accurate	accurate	ADJ
cana-2597	37	2	data	datum	NOUN
cana-2597	37	3	validation	validation	NOUN
cana-2597	37	4	is	be	AUX
cana-2597	37	5	critical	critical	ADJ
cana-2597	37	6	in	in	ADP
cana-2597	37	7	disaster	disaster	NOUN
cana-2597	37	8	management	management	NOUN
cana-2597	37	9	to	to	PART
cana-2597	37	10	prevent	prevent	VERB
cana-2597	37	11	the	the	DET
cana-2597	37	12	spread	spread	NOUN
cana-2597	37	13	of	of	ADP
cana-2597	37	14	misinformation	misinformation	NOUN
cana-2597	37	15	,	,	PUNCT
cana-2597	37	16	especially	especially	ADV
cana-2597	37	17	on	on	ADP
cana-2597	37	18	social	social	ADJ
cana-2597	37	19	media	medium	NOUN
cana-2597	37	20	platforms	platform	NOUN
cana-2597	37	21	.	.	PUNCT
cana-2597	38	1	mishra	mishra	PROPN
cana-2597	39	1	[	[	X
cana-2597	39	2	6	6	NUM
cana-2597	39	3	]	]	PUNCT
cana-2597	39	4	addressed	address	VERB
cana-2597	39	5	this	this	DET
cana-2597	39	6	challenge	challenge	NOUN
cana-2597	39	7	through	through	ADP
cana-2597	39	8	the	the	DET
cana-2597	39	9	disaster	disaster	NOUN
cana-2597	39	10	information	information	NOUN
cana-2597	39	11	verification	verification	NOUN
cana-2597	39	12	and	and	CCONJ
cana-2597	39	13	validation	validation	NOUN
cana-2597	39	14	application	application	NOUN
cana-2597	39	15	(	(	PUNCT
cana-2597	39	16	divva	divva	NOUN
cana-2597	39	17	)	)	PUNCT
cana-2597	39	18	,	,	PUNCT
cana-2597	39	19	which	which	PRON
cana-2597	39	20	uses	use	VERB
cana-2597	39	21	machine	machine	NOUN
cana-2597	39	22	learning	learn	VERB
cana-2597	39	23	to	to	PART
cana-2597	39	24	verify	verify	VERB
cana-2597	39	25	disaster	disaster	NOUN
cana-2597	39	26	-	-	PUNCT
cana-2597	39	27	related	relate	VERB
cana-2597	39	28	information	information	NOUN
cana-2597	39	29	by	by	ADP
cana-2597	39	30	cross	cros	VERB
cana-2597	39	31	-	-	ADJ
cana-2597	39	32	referencing	reference	VERB
cana-2597	39	33	it	it	PRON
cana-2597	39	34	with	with	ADP
cana-2597	39	35	official	official	ADJ
cana-2597	39	36	government	government	NOUN
cana-2597	39	37	sources	source	NOUN
cana-2597	39	38	.	.	PUNCT
cana-2597	40	1	the	the	DET
cana-2597	40	2	system	system	NOUN
cana-2597	40	3	utilizes	utilize	VERB
cana-2597	40	4	a	a	DET
cana-2597	40	5	bidirectional	bidirectional	ADJ
cana-2597	40	6	lstm	lstm	NOUN
cana-2597	40	7	model	model	NOUN
cana-2597	40	8	to	to	PART
cana-2597	40	9	classify	classify	VERB
cana-2597	40	10	tweets	tweet	NOUN
cana-2597	40	11	as	as	ADP
cana-2597	40	12	real	real	ADJ
cana-2597	40	13	or	or	CCONJ
cana-2597	40	14	fake	fake	ADJ
cana-2597	40	15	based	base	VERB
cana-2597	40	16	on	on	ADP
cana-2597	40	17	textual	textual	ADJ
cana-2597	40	18	input	input	NOUN
cana-2597	40	19	analysis	analysis	NOUN
cana-2597	40	20	,	,	PUNCT
cana-2597	40	21	achieving	achieve	VERB
cana-2597	40	22	an	an	DET
cana-2597	40	23	accuracy	accuracy	NOUN
cana-2597	40	24	of	of	ADP
cana-2597	40	25	84	84	NUM
cana-2597	40	26	%	%	NOUN
cana-2597	40	27	.	.	PUNCT
cana-2597	41	1	divva	divva	PROPN
cana-2597	41	2	’s	’s	PART
cana-2597	41	3	approach	approach	NOUN
cana-2597	41	4	to	to	ADP
cana-2597	41	5	ensuring	ensure	VERB
cana-2597	41	6	the	the	DET
cana-2597	41	7	reliability	reliability	NOUN
cana-2597	41	8	of	of	ADP
cana-2597	41	9	disaster	disaster	NOUN
cana-2597	41	10	information	information	NOUN
cana-2597	41	11	is	be	AUX
cana-2597	41	12	highly	highly	ADV
cana-2597	41	13	relevant	relevant	ADJ
cana-2597	41	14	to	to	ADP
cana-2597	41	15	our	our	PRON
cana-2597	41	16	project	project	NOUN
cana-2597	41	17	,	,	PUNCT
cana-2597	41	18	which	which	PRON
cana-2597	41	19	focuses	focus	VERB
cana-2597	41	20	on	on	ADP
cana-2597	41	21	aggregating	aggregate	VERB
cana-2597	41	22	and	and	CCONJ
cana-2597	41	23	validating	validate	VERB
cana-2597	41	24	data	datum	NOUN
cana-2597	41	25	from	from	ADP
cana-2597	41	26	social	social	ADJ
cana-2597	41	27	media	medium	NOUN
cana-2597	41	28	and	and	CCONJ
cana-2597	41	29	news	news	NOUN
cana-2597	41	30	,	,	PUNCT
cana-2597	41	31	ensuring	ensure	VERB
cana-2597	41	32	only	only	ADV
cana-2597	41	33	verified	verified	ADJ
cana-2597	41	34	and	and	CCONJ
cana-2597	41	35	actionable	actionable	ADJ
cana-2597	41	36	information	information	NOUN
cana-2597	41	37	is	be	AUX
cana-2597	41	38	included	include	VERB
cana-2597	41	39	in	in	ADP
cana-2597	41	40	disaster	disaster	NOUN
cana-2597	41	41	response	response	NOUN
cana-2597	41	42	efforts	effort	NOUN
cana-2597	41	43	.	.	PUNCT
cana-2597	42	1	beyond	beyond	ADP
cana-2597	42	2	data	data	NOUN
cana-2597	42	3	collection	collection	NOUN
cana-2597	42	4	and	and	CCONJ
cana-2597	42	5	validation	validation	NOUN
cana-2597	42	6	,	,	PUNCT
cana-2597	42	7	effective	effective	ADJ
cana-2597	42	8	categorization	categorization	NOUN
cana-2597	42	9	of	of	ADP
cana-2597	42	10	disaster	disaster	NOUN
cana-2597	42	11	-	-	PUNCT
cana-2597	42	12	related	relate	VERB
cana-2597	42	13	information	information	NOUN
cana-2597	42	14	is	be	AUX
cana-2597	42	15	essential	essential	ADJ
cana-2597	42	16	for	for	ADP
cana-2597	42	17	timely	timely	ADJ
cana-2597	42	18	response	response	NOUN
cana-2597	42	19	.	.	PUNCT
cana-2597	43	1	the	the	DET
cana-2597	43	2	work	work	NOUN
cana-2597	43	3	by	by	ADP
cana-2597	43	4	mishra	mishra	PROPN
cana-2597	43	5	[	[	X
cana-2597	43	6	6	6	NUM
cana-2597	43	7	]	]	PUNCT
cana-2597	43	8	introduced	introduce	VERB
cana-2597	43	9	a	a	DET
cana-2597	43	10	deep	deep	ADJ
cana-2597	43	11	multimodal	multimodal	NOUN
cana-2597	43	12	learning	learn	VERB
cana-2597	43	13	approach	approach	NOUN
cana-2597	43	14	communications	communication	NOUN
cana-2597	43	15	on	on	ADP
cana-2597	43	16	applied	apply	VERB
cana-2597	43	17	nonlinear	nonlinear	ADJ
cana-2597	43	18	analysis	analysis	NOUN
cana-2597	43	19	issn	issn	NOUN
cana-2597	43	20	:	:	PUNCT
cana-2597	43	21	1074	1074	NUM
cana-2597	43	22	-	-	PUNCT
cana-2597	43	23	133x	133x	NUM
cana-2597	43	24	vol	vol	NOUN
cana-2597	43	25	32	32	NUM
cana-2597	43	26	no	no	NOUN
cana-2597	43	27	.	.	PUNCT
cana-2597	44	1	3s	3s	NUM
cana-2597	44	2	(	(	PUNCT
cana-2597	44	3	2025	2025	NUM
cana-2597	44	4	)	)	PUNCT
cana-2597	44	5	200	200	NUM
cana-2597	44	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2597	44	7	that	that	PRON
cana-2597	44	8	integrates	integrate	VERB
cana-2597	44	9	visual	visual	ADJ
cana-2597	44	10	and	and	CCONJ
cana-2597	44	11	textual	textual	ADJ
cana-2597	44	12	features	feature	NOUN
cana-2597	44	13	to	to	PART
cana-2597	44	14	classify	classify	VERB
cana-2597	44	15	disaster	disaster	NOUN
cana-2597	44	16	types	type	NOUN
cana-2597	44	17	from	from	ADP
cana-2597	44	18	social	social	ADJ
cana-2597	44	19	media	medium	NOUN
cana-2597	44	20	posts	post	NOUN
cana-2597	44	21	.	.	PUNCT
cana-2597	45	1	the	the	DET
cana-2597	45	2	system	system	NOUN
cana-2597	45	3	employs	employ	VERB
cana-2597	45	4	resnet50	resnet50	NOUN
cana-2597	45	5	for	for	ADP
cana-2597	45	6	visual	visual	ADJ
cana-2597	45	7	feature	feature	NOUN
cana-2597	45	8	extraction	extraction	NOUN
cana-2597	45	9	and	and	CCONJ
cana-2597	45	10	a	a	DET
cana-2597	45	11	bilstm	bilstm	NOUN
cana-2597	45	12	model	model	NOUN
cana-2597	45	13	with	with	ADP
cana-2597	45	14	an	an	DET
cana-2597	45	15	attention	attention	NOUN
cana-2597	45	16	mechanism	mechanism	NOUN
cana-2597	45	17	for	for	ADP
cana-2597	45	18	textual	textual	ADJ
cana-2597	45	19	analysis	analysis	NOUN
cana-2597	45	20	,	,	PUNCT
cana-2597	45	21	showing	show	VERB
cana-2597	45	22	a	a	DET
cana-2597	45	23	significant	significant	ADJ
cana-2597	45	24	performance	performance	NOUN
cana-2597	45	25	boost	boost	NOUN
cana-2597	45	26	over	over	ADP
cana-2597	45	27	unimodal	unimodal	ADJ
cana-2597	45	28	models	model	NOUN
cana-2597	45	29	and	and	CCONJ
cana-2597	45	30	even	even	ADV
cana-2597	45	31	improving	improve	VERB
cana-2597	45	32	by	by	ADP
cana-2597	45	33	7	7	NUM
cana-2597	45	34	%	%	NOUN
cana-2597	45	35	compared	compare	VERB
cana-2597	45	36	to	to	ADP
cana-2597	45	37	other	other	ADJ
cana-2597	45	38	multimodal	multimodal	NOUN
cana-2597	45	39	approaches	approach	NOUN
cana-2597	45	40	.	.	PUNCT
cana-2597	46	1	this	this	DET
cana-2597	46	2	method	method	NOUN
cana-2597	46	3	addresses	address	VERB
cana-2597	46	4	the	the	DET
cana-2597	46	5	challenge	challenge	NOUN
cana-2597	46	6	of	of	ADP
cana-2597	46	7	processing	process	VERB
cana-2597	46	8	vast	vast	ADJ
cana-2597	46	9	amounts	amount	NOUN
cana-2597	46	10	of	of	ADP
cana-2597	46	11	social	social	ADJ
cana-2597	46	12	media	medium	NOUN
cana-2597	46	13	data	datum	NOUN
cana-2597	46	14	,	,	PUNCT
cana-2597	46	15	ensuring	ensure	VERB
cana-2597	46	16	more	more	ADV
cana-2597	46	17	accurate	accurate	ADJ
cana-2597	46	18	disaster	disaster	NOUN
cana-2597	46	19	categorization	categorization	NOUN
cana-2597	46	20	.	.	PUNCT
cana-2597	47	1	incorporating	incorporate	VERB
cana-2597	47	2	such	such	DET
cana-2597	47	3	a	a	DET
cana-2597	47	4	multimodal	multimodal	ADJ
cana-2597	47	5	approach	approach	NOUN
cana-2597	47	6	into	into	ADP
cana-2597	47	7	our	our	PRON
cana-2597	47	8	project	project	NOUN
cana-2597	47	9	can	can	AUX
cana-2597	47	10	enhance	enhance	VERB
cana-2597	47	11	the	the	DET
cana-2597	47	12	classification	classification	NOUN
cana-2597	47	13	of	of	ADP
cana-2597	47	14	disaster	disaster	NOUN
cana-2597	47	15	data	datum	NOUN
cana-2597	47	16	,	,	PUNCT
cana-2597	47	17	making	make	VERB
cana-2597	47	18	the	the	DET
cana-2597	47	19	system	system	NOUN
cana-2597	47	20	more	more	ADV
cana-2597	47	21	robust	robust	ADJ
cana-2597	47	22	in	in	ADP
cana-2597	47	23	identifying	identify	VERB
cana-2597	47	24	and	and	CCONJ
cana-2597	47	25	organizing	organize	VERB
cana-2597	47	26	real	real	ADJ
cana-2597	47	27	-	-	PUNCT
cana-2597	47	28	time	time	NOUN
cana-2597	47	29	information	information	NOUN
cana-2597	47	30	.	.	PUNCT
cana-2597	48	1	effective	effective	ADJ
cana-2597	48	2	classification	classification	NOUN
cana-2597	48	3	of	of	ADP
cana-2597	48	4	disaster	disaster	NOUN
cana-2597	48	5	-	-	PUNCT
cana-2597	48	6	related	relate	VERB
cana-2597	48	7	tweets	tweet	NOUN
cana-2597	48	8	is	be	AUX
cana-2597	48	9	crucial	crucial	ADJ
cana-2597	48	10	for	for	ADP
cana-2597	48	11	timely	timely	ADJ
cana-2597	48	12	response	response	NOUN
cana-2597	48	13	and	and	CCONJ
cana-2597	48	14	rescue	rescue	NOUN
cana-2597	48	15	efforts	effort	NOUN
cana-2597	48	16	.	.	PUNCT
cana-2597	49	1	ningsih	ningsih	PROPN
cana-2597	50	1	[	[	X
cana-2597	50	2	7	7	X
cana-2597	50	3	]	]	PUNCT
cana-2597	50	4	explored	explore	VERB
cana-2597	50	5	the	the	DET
cana-2597	50	6	use	use	NOUN
cana-2597	50	7	of	of	ADP
cana-2597	50	8	the	the	DET
cana-2597	50	9	bidirectional	bidirectional	ADJ
cana-2597	50	10	encoder	encoder	NOUN
cana-2597	50	11	representations	representation	VERB
cana-2597	50	12	from	from	ADP
cana-2597	50	13	transformers	transformer	NOUN
cana-2597	50	14	(	(	PUNCT
cana-2597	50	15	bert	bert	PROPN
cana-2597	50	16	)	)	PUNCT
cana-2597	50	17	model	model	NOUN
cana-2597	50	18	to	to	PART
cana-2597	50	19	enhance	enhance	VERB
cana-2597	50	20	the	the	DET
cana-2597	50	21	identification	identification	NOUN
cana-2597	50	22	of	of	ADP
cana-2597	50	23	disaster	disaster	NOUN
cana-2597	50	24	-	-	PUNCT
cana-2597	50	25	related	relate	VERB
cana-2597	50	26	tweets	tweet	NOUN
cana-2597	50	27	,	,	PUNCT
cana-2597	50	28	particularly	particularly	ADV
cana-2597	50	29	focusing	focus	VERB
cana-2597	50	30	on	on	ADP
cana-2597	50	31	sentiment	sentiment	NOUN
cana-2597	50	32	analysis	analysis	NOUN
cana-2597	50	33	.	.	PUNCT
cana-2597	51	1	this	this	DET
cana-2597	51	2	approach	approach	NOUN
cana-2597	51	3	helps	help	VERB
cana-2597	51	4	differentiate	differentiate	VERB
cana-2597	51	5	critical	critical	ADJ
cana-2597	51	6	tweets	tweet	NOUN
cana-2597	51	7	,	,	PUNCT
cana-2597	51	8	such	such	ADJ
cana-2597	51	9	as	as	ADP
cana-2597	51	10	rescue	rescue	NOUN
cana-2597	51	11	requests	request	NOUN
cana-2597	51	12	,	,	PUNCT
cana-2597	51	13	from	from	ADP
cana-2597	51	14	general	general	ADJ
cana-2597	51	15	disaster	disaster	NOUN
cana-2597	51	16	discussions	discussion	NOUN
cana-2597	51	17	,	,	PUNCT
cana-2597	51	18	addressing	address	VERB
cana-2597	51	19	the	the	DET
cana-2597	51	20	challenge	challenge	NOUN
cana-2597	51	21	of	of	ADP
cana-2597	51	22	understanding	understand	VERB
cana-2597	51	23	complex	complex	ADJ
cana-2597	51	24	language	language	NOUN
cana-2597	51	25	structures	structure	NOUN
cana-2597	51	26	during	during	ADP
cana-2597	51	27	emergencies	emergency	NOUN
cana-2597	51	28	.	.	PUNCT
cana-2597	52	1	bert	bert	PROPN
cana-2597	52	2	's	's	PART
cana-2597	52	3	ability	ability	NOUN
cana-2597	52	4	to	to	PART
cana-2597	52	5	manage	manage	VERB
cana-2597	52	6	context	context	NOUN
cana-2597	52	7	and	and	CCONJ
cana-2597	52	8	sentiment	sentiment	NOUN
cana-2597	52	9	uncertainty	uncertainty	NOUN
cana-2597	52	10	proves	prove	VERB
cana-2597	52	11	beneficial	beneficial	ADJ
cana-2597	52	12	in	in	ADP
cana-2597	52	13	improving	improve	VERB
cana-2597	52	14	tweet	tweet	NOUN
cana-2597	52	15	classification	classification	NOUN
cana-2597	52	16	,	,	PUNCT
cana-2597	52	17	thereby	thereby	ADV
cana-2597	52	18	aiding	aid	VERB
cana-2597	52	19	disaster	disaster	NOUN
cana-2597	52	20	response	response	NOUN
cana-2597	52	21	teams	team	NOUN
cana-2597	52	22	in	in	ADP
cana-2597	52	23	prioritizing	prioritize	VERB
cana-2597	52	24	rescue	rescue	NOUN
cana-2597	52	25	operations	operation	NOUN
cana-2597	52	26	.	.	PUNCT
cana-2597	53	1	integrating	integrate	VERB
cana-2597	53	2	bert	bert	NOUN
cana-2597	53	3	-	-	PUNCT
cana-2597	53	4	like	like	ADJ
cana-2597	53	5	models	model	NOUN
cana-2597	53	6	into	into	ADP
cana-2597	53	7	our	our	PRON
cana-2597	53	8	project	project	NOUN
cana-2597	53	9	can	can	AUX
cana-2597	53	10	significantly	significantly	ADV
cana-2597	53	11	enhance	enhance	VERB
cana-2597	53	12	the	the	DET
cana-2597	53	13	accuracy	accuracy	NOUN
cana-2597	53	14	and	and	CCONJ
cana-2597	53	15	relevance	relevance	NOUN
cana-2597	53	16	of	of	ADP
cana-2597	53	17	disaster	disaster	NOUN
cana-2597	53	18	data	datum	NOUN
cana-2597	53	19	classification	classification	NOUN
cana-2597	53	20	.	.	PUNCT
cana-2597	54	1	iii	iii	X
cana-2597	54	2	.	.	PROPN
cana-2597	54	3	proposed	propose	VERB
cana-2597	54	4	methodology	methodology	NOUN
cana-2597	54	5	in	in	ADP
cana-2597	54	6	natural	natural	ADJ
cana-2597	54	7	disasters	disaster	NOUN
cana-2597	54	8	,	,	PUNCT
cana-2597	54	9	platforms	platform	NOUN
cana-2597	54	10	like	like	ADP
cana-2597	54	11	twitter	twitter	NOUN
cana-2597	54	12	become	become	VERB
cana-2597	54	13	critical	critical	ADJ
cana-2597	54	14	for	for	ADP
cana-2597	54	15	real	real	ADJ
cana-2597	54	16	-	-	PUNCT
cana-2597	54	17	time	time	NOUN
cana-2597	54	18	updates	update	NOUN
cana-2597	54	19	.	.	PUNCT
cana-2597	55	1	the	the	DET
cana-2597	55	2	proposed	propose	VERB
cana-2597	55	3	solution	solution	NOUN
cana-2597	55	4	addresses	address	VERB
cana-2597	55	5	the	the	DET
cana-2597	55	6	challenge	challenge	NOUN
cana-2597	55	7	of	of	ADP
cana-2597	55	8	collecting	collecting	NOUN
cana-2597	55	9	,	,	PUNCT
cana-2597	55	10	processing	processing	NOUN
cana-2597	55	11	,	,	PUNCT
cana-2597	55	12	and	and	CCONJ
cana-2597	55	13	classifying	classify	VERB
cana-2597	55	14	disaster	disaster	NOUN
cana-2597	55	15	-	-	PUNCT
cana-2597	55	16	related	relate	VERB
cana-2597	55	17	tweets	tweet	NOUN
cana-2597	55	18	,	,	PUNCT
cana-2597	55	19	overcoming	overcome	VERB
cana-2597	55	20	twitter	twitter	NOUN
cana-2597	55	21	's	's	PART
cana-2597	55	22	api	api	NOUN
cana-2597	55	23	restrictions	restriction	NOUN
cana-2597	55	24	and	and	CCONJ
cana-2597	55	25	large	large	ADJ
cana-2597	55	26	data	datum	NOUN
cana-2597	55	27	volumes	volume	NOUN
cana-2597	55	28	.	.	PUNCT
cana-2597	56	1	the	the	DET
cana-2597	56	2	system	system	NOUN
cana-2597	56	3	integrates	integrate	VERB
cana-2597	56	4	web	web	NOUN
cana-2597	56	5	scraping	scraping	NOUN
cana-2597	56	6	,	,	PUNCT
cana-2597	56	7	real	real	ADJ
cana-2597	56	8	-	-	PUNCT
cana-2597	56	9	time	time	NOUN
cana-2597	56	10	streaming	streaming	NOUN
cana-2597	56	11	,	,	PUNCT
cana-2597	56	12	and	and	CCONJ
cana-2597	56	13	machine	machine	NOUN
cana-2597	56	14	learning	learn	VERB
cana-2597	56	15	to	to	PART
cana-2597	56	16	efficiently	efficiently	ADV
cana-2597	56	17	filter	filter	VERB
cana-2597	56	18	and	and	CCONJ
cana-2597	56	19	classify	classify	VERB
cana-2597	56	20	tweets	tweet	NOUN
cana-2597	56	21	,	,	PUNCT
cana-2597	56	22	thereby	thereby	ADV
cana-2597	56	23	improving	improve	VERB
cana-2597	56	24	disaster	disaster	NOUN
cana-2597	56	25	response	response	NOUN
cana-2597	56	26	.	.	PUNCT
cana-2597	57	1	3.1	3.1	NUM
cana-2597	57	2	.	.	PUNCT
cana-2597	57	3	data	datum	NOUN
cana-2597	57	4	collection	collection	NOUN
cana-2597	57	5	via	via	ADP
cana-2597	57	6	web	web	NOUN
cana-2597	57	7	scraping	scraping	NOUN
cana-2597	57	8	due	due	ADP
cana-2597	57	9	to	to	ADP
cana-2597	57	10	twitter	twitter	NOUN
cana-2597	57	11	’s	’s	PART
cana-2597	57	12	recent	recent	ADJ
cana-2597	57	13	api	api	NOUN
cana-2597	57	14	restrictions	restriction	NOUN
cana-2597	57	15	,	,	PUNCT
cana-2597	57	16	web	web	NOUN
cana-2597	57	17	scraping	scraping	NOUN
cana-2597	57	18	has	have	AUX
cana-2597	57	19	become	become	VERB
cana-2597	57	20	an	an	DET
cana-2597	57	21	essential	essential	ADJ
cana-2597	57	22	alternative	alternative	NOUN
cana-2597	57	23	for	for	ADP
cana-2597	57	24	collecting	collect	VERB
cana-2597	57	25	tweets	tweet	NOUN
cana-2597	57	26	.	.	PUNCT
cana-2597	58	1	this	this	DET
cana-2597	58	2	solution	solution	NOUN
cana-2597	58	3	leverages	leverage	VERB
cana-2597	58	4	selenium	selenium	NOUN
cana-2597	58	5	and	and	CCONJ
cana-2597	58	6	beautifulsoup	beautifulsoup	NOUN
cana-2597	58	7	to	to	PART
cana-2597	58	8	scrape	scrape	VERB
cana-2597	58	9	tweet	tweet	NOUN
cana-2597	58	10	content	content	NOUN
cana-2597	58	11	,	,	PUNCT
cana-2597	58	12	metadata	metadata	NOUN
cana-2597	58	13	,	,	PUNCT
cana-2597	58	14	and	and	CCONJ
cana-2597	58	15	user	user	NOUN
cana-2597	58	16	information	information	NOUN
cana-2597	58	17	directly	directly	ADV
cana-2597	58	18	from	from	ADP
cana-2597	58	19	twitter	twitter	NOUN
cana-2597	58	20	's	's	PART
cana-2597	58	21	web	web	NOUN
cana-2597	58	22	interface	interface	NOUN
cana-2597	58	23	.	.	PUNCT
cana-2597	59	1	to	to	PART
cana-2597	59	2	handle	handle	VERB
cana-2597	59	3	twitter	twitter	NOUN
cana-2597	59	4	’s	’s	PART
cana-2597	59	5	rate	rate	NOUN
cana-2597	59	6	limits	limit	NOUN
cana-2597	59	7	and	and	CCONJ
cana-2597	59	8	prevent	prevent	VERB
cana-2597	59	9	detection	detection	NOUN
cana-2597	59	10	,	,	PUNCT
cana-2597	59	11	multiple	multiple	ADJ
cana-2597	59	12	techniques	technique	NOUN
cana-2597	59	13	are	be	AUX
cana-2597	59	14	employed	employ	VERB
cana-2597	59	15	.	.	PUNCT
cana-2597	60	1	one	one	NUM
cana-2597	60	2	such	such	ADJ
cana-2597	60	3	technique	technique	NOUN
cana-2597	60	4	is	be	AUX
cana-2597	60	5	proxy	proxy	ADJ
cana-2597	60	6	rotation	rotation	NOUN
cana-2597	60	7	,	,	PUNCT
cana-2597	60	8	which	which	PRON
cana-2597	60	9	uses	use	VERB
cana-2597	60	10	a	a	DET
cana-2597	60	11	rotating	rotate	VERB
cana-2597	60	12	pool	pool	NOUN
cana-2597	60	13	of	of	ADP
cana-2597	60	14	proxies	proxy	NOUN
cana-2597	60	15	to	to	PART
cana-2597	60	16	prevent	prevent	VERB
cana-2597	60	17	ip	ip	PRON
cana-2597	60	18	bans	ban	NOUN
cana-2597	60	19	and	and	CCONJ
cana-2597	60	20	ensure	ensure	VERB
cana-2597	60	21	consistent	consistent	ADJ
cana-2597	60	22	scraping	scraping	NOUN
cana-2597	60	23	throughout	throughout	ADP
cana-2597	60	24	the	the	DET
cana-2597	60	25	process	process	NOUN
cana-2597	60	26	[	[	X
cana-2597	60	27	8	8	NUM
cana-2597	60	28	]	]	PUNCT
cana-2597	60	29	.	.	PUNCT
cana-2597	61	1	additionally	additionally	ADV
cana-2597	61	2	,	,	PUNCT
cana-2597	61	3	automated	automated	ADJ
cana-2597	61	4	scrolling	scrolling	NOUN
cana-2597	61	5	is	be	AUX
cana-2597	61	6	simulated	simulate	VERB
cana-2597	61	7	using	use	VERB
cana-2597	61	8	selenium	selenium	NOUN
cana-2597	61	9	to	to	PART
cana-2597	61	10	load	load	VERB
cana-2597	61	11	and	and	CCONJ
cana-2597	61	12	extract	extract	VERB
cana-2597	61	13	older	old	ADJ
cana-2597	61	14	tweets	tweet	NOUN
cana-2597	61	15	that	that	PRON
cana-2597	61	16	are	be	AUX
cana-2597	61	17	beyond	beyond	ADP
cana-2597	61	18	the	the	DET
cana-2597	61	19	initial	initial	ADJ
cana-2597	61	20	page	page	NOUN
cana-2597	61	21	view	view	NOUN
cana-2597	62	1	[	[	X
cana-2597	62	2	9	9	NUM
cana-2597	62	3	]	]	PUNCT
cana-2597	62	4	.	.	PUNCT
cana-2597	63	1	furthermore	furthermore	ADV
cana-2597	63	2	,	,	PUNCT
cana-2597	63	3	throttling	throttling	NOUN
cana-2597	63	4	and	and	CCONJ
cana-2597	63	5	user	user	NOUN
cana-2597	63	6	-	-	PUNCT
cana-2597	63	7	agent	agent	NOUN
cana-2597	63	8	spoofing	spoofing	NOUN
cana-2597	63	9	are	be	AUX
cana-2597	63	10	implemented	implement	VERB
cana-2597	63	11	to	to	PART
cana-2597	63	12	mimic	mimic	VERB
cana-2597	63	13	human	human	ADJ
cana-2597	63	14	behavior	behavior	NOUN
cana-2597	63	15	and	and	CCONJ
cana-2597	63	16	reduce	reduce	VERB
cana-2597	63	17	the	the	DET
cana-2597	63	18	risk	risk	NOUN
cana-2597	63	19	of	of	ADP
cana-2597	63	20	being	be	AUX
cana-2597	63	21	detected	detect	VERB
cana-2597	63	22	as	as	ADP
cana-2597	63	23	a	a	DET
cana-2597	63	24	bot	bot	NOUN
cana-2597	63	25	.	.	PUNCT
cana-2597	64	1	by	by	ADP
cana-2597	64	2	introducing	introduce	VERB
cana-2597	64	3	randomized	randomized	ADJ
cana-2597	64	4	delays	delay	NOUN
cana-2597	64	5	between	between	ADP
cana-2597	64	6	requests	request	NOUN
cana-2597	64	7	and	and	CCONJ
cana-2597	64	8	changing	change	VERB
cana-2597	64	9	user	user	NOUN
cana-2597	64	10	-	-	PUNCT
cana-2597	64	11	agents	agent	NOUN
cana-2597	64	12	at	at	ADP
cana-2597	64	13	regular	regular	ADJ
cana-2597	64	14	intervals	interval	NOUN
cana-2597	64	15	,	,	PUNCT
cana-2597	64	16	the	the	DET
cana-2597	64	17	system	system	NOUN
cana-2597	64	18	closely	closely	ADV
cana-2597	64	19	emulates	emulate	VERB
cana-2597	64	20	typical	typical	ADJ
cana-2597	64	21	browsing	browse	VERB
cana-2597	64	22	patterns	pattern	NOUN
cana-2597	64	23	,	,	PUNCT
cana-2597	64	24	allowing	allow	VERB
cana-2597	64	25	it	it	PRON
cana-2597	64	26	to	to	PART
cana-2597	64	27	bypass	bypass	NOUN
cana-2597	64	28	twitter	twitter	NOUN
cana-2597	64	29	’s	’s	PART
cana-2597	64	30	anti	anti	ADJ
cana-2597	64	31	-	-	ADJ
cana-2597	64	32	bot	bot	ADJ
cana-2597	64	33	mechanisms	mechanism	NOUN
cana-2597	64	34	[	[	X
cana-2597	64	35	10	10	NUM
cana-2597	64	36	]	]	PUNCT
cana-2597	64	37	.	.	PUNCT
cana-2597	65	1	these	these	DET
cana-2597	65	2	combined	combined	ADJ
cana-2597	65	3	methods	method	NOUN
cana-2597	65	4	ensure	ensure	VERB
cana-2597	65	5	scalable	scalable	ADJ
cana-2597	65	6	and	and	CCONJ
cana-2597	65	7	efficient	efficient	ADJ
cana-2597	65	8	tweet	tweet	NOUN
cana-2597	65	9	collection	collection	NOUN
cana-2597	65	10	despite	despite	SCONJ
cana-2597	65	11	the	the	DET
cana-2597	65	12	recent	recent	ADJ
cana-2597	65	13	restrictions	restriction	NOUN
cana-2597	65	14	imposed	impose	VERB
cana-2597	65	15	by	by	ADP
cana-2597	65	16	twitter	twitter	NOUN
cana-2597	65	17	’s	’s	PART
cana-2597	65	18	api	api	NOUN
cana-2597	65	19	,	,	PUNCT
cana-2597	65	20	enabling	enable	VERB
cana-2597	65	21	a	a	DET
cana-2597	65	22	continuous	continuous	ADJ
cana-2597	65	23	stream	stream	NOUN
cana-2597	65	24	of	of	ADP
cana-2597	65	25	relevant	relevant	ADJ
cana-2597	65	26	data	datum	NOUN
cana-2597	65	27	for	for	ADP
cana-2597	65	28	disaster	disaster	NOUN
cana-2597	65	29	-	-	PUNCT
cana-2597	65	30	related	relate	VERB
cana-2597	65	31	analysis	analysis	NOUN
cana-2597	65	32	.	.	PUNCT
cana-2597	66	1	communications	communication	NOUN
cana-2597	66	2	on	on	ADP
cana-2597	66	3	applied	apply	VERB
cana-2597	66	4	nonlinear	nonlinear	ADJ
cana-2597	66	5	analysis	analysis	NOUN
cana-2597	66	6	issn	issn	NOUN
cana-2597	66	7	:	:	PUNCT
cana-2597	66	8	1074	1074	NUM
cana-2597	66	9	-	-	PUNCT
cana-2597	66	10	133x	133x	NUM
cana-2597	66	11	vol	vol	NOUN
cana-2597	66	12	32	32	NUM
cana-2597	66	13	no	no	NOUN
cana-2597	66	14	.	.	PUNCT
cana-2597	67	1	3s	3s	NUM
cana-2597	67	2	(	(	PUNCT
cana-2597	67	3	2025	2025	NUM
cana-2597	67	4	)	)	PUNCT
cana-2597	67	5	201	201	NUM
cana-2597	67	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2597	67	7	fig	fig	NOUN
cana-2597	67	8	1	1	NUM
cana-2597	67	9	.	.	PUNCT
cana-2597	68	1	methodolgy	methodolgy	ADJ
cana-2597	68	2	figure	figure	NOUN
cana-2597	68	3	1	1	NUM
cana-2597	68	4	illustrates	illustrate	VERB
cana-2597	68	5	the	the	DET
cana-2597	68	6	process	process	NOUN
cana-2597	68	7	flow	flow	NOUN
cana-2597	68	8	of	of	ADP
cana-2597	68	9	collecting	collect	VERB
cana-2597	68	10	data	datum	NOUN
cana-2597	68	11	via	via	ADP
cana-2597	68	12	web	web	NOUN
cana-2597	68	13	scraping	scraping	NOUN
cana-2597	68	14	and	and	CCONJ
cana-2597	68	15	real	real	ADJ
cana-2597	68	16	-	-	PUNCT
cana-2597	68	17	time	time	NOUN
cana-2597	68	18	streaming	streaming	NOUN
cana-2597	68	19	,	,	PUNCT
cana-2597	68	20	followed	follow	VERB
cana-2597	68	21	by	by	ADP
cana-2597	68	22	categorizing	categorize	VERB
cana-2597	68	23	tweets	tweet	NOUN
cana-2597	68	24	using	use	VERB
cana-2597	68	25	machine	machine	NOUN
cana-2597	68	26	learning	learning	NOUN
cana-2597	68	27	models	model	NOUN
cana-2597	68	28	.	.	PUNCT
cana-2597	69	1	this	this	DET
cana-2597	69	2	flow	flow	NOUN
cana-2597	69	3	starts	start	VERB
cana-2597	69	4	from	from	ADP
cana-2597	69	5	the	the	DET
cana-2597	69	6	data	datum	NOUN
cana-2597	69	7	scraping	scrape	VERB
cana-2597	69	8	process	process	NOUN
cana-2597	69	9	,	,	PUNCT
cana-2597	69	10	streaming	streaming	NOUN
cana-2597	69	11	to	to	ADP
cana-2597	69	12	kafka	kafka	PROPN
cana-2597	69	13	topics	topic	NOUN
cana-2597	69	14	,	,	PUNCT
cana-2597	69	15	followed	follow	VERB
cana-2597	69	16	by	by	ADP
cana-2597	69	17	nlp	nlp	NOUN
cana-2597	69	18	processing	processing	NOUN
cana-2597	69	19	,	,	PUNCT
cana-2597	69	20	classification	classification	NOUN
cana-2597	69	21	,	,	PUNCT
cana-2597	69	22	and	and	CCONJ
cana-2597	69	23	storage	storage	NOUN
cana-2597	69	24	.	.	PUNCT
cana-2597	70	1	3.2	3.2	NUM
cana-2597	70	2	.	.	PUNCT
cana-2597	70	3	real	real	ADJ
cana-2597	70	4	-	-	PUNCT
cana-2597	70	5	time	time	NOUN
cana-2597	70	6	data	datum	NOUN
cana-2597	70	7	ingestion	ingestion	NOUN
cana-2597	70	8	and	and	CCONJ
cana-2597	70	9	processing	processing	NOUN
cana-2597	70	10	pipeline	pipeline	NOUN
cana-2597	70	11	to	to	PART
cana-2597	70	12	handle	handle	VERB
cana-2597	70	13	the	the	DET
cana-2597	70	14	influx	influx	NOUN
cana-2597	70	15	of	of	ADP
cana-2597	70	16	disaster	disaster	NOUN
cana-2597	70	17	-	-	PUNCT
cana-2597	70	18	related	relate	VERB
cana-2597	70	19	tweets	tweet	NOUN
cana-2597	70	20	,	,	PUNCT
cana-2597	70	21	a	a	DET
cana-2597	70	22	robust	robust	ADJ
cana-2597	70	23	real	real	ADJ
cana-2597	70	24	-	-	PUNCT
cana-2597	70	25	time	time	NOUN
cana-2597	70	26	data	datum	NOUN
cana-2597	70	27	ingestion	ingestion	NOUN
cana-2597	70	28	and	and	CCONJ
cana-2597	70	29	processing	processing	NOUN
cana-2597	70	30	pipeline	pipeline	NOUN
cana-2597	70	31	is	be	AUX
cana-2597	70	32	established	establish	VERB
cana-2597	70	33	using	use	VERB
cana-2597	70	34	apache	apache	PROPN
cana-2597	70	35	kafka	kafka	PROPN
cana-2597	70	36	and	and	CCONJ
cana-2597	70	37	apache	apache	NOUN
cana-2597	70	38	spark	spark	NOUN
cana-2597	70	39	.	.	PUNCT
cana-2597	71	1	this	this	DET
cana-2597	71	2	pipeline	pipeline	NOUN
cana-2597	71	3	is	be	AUX
cana-2597	71	4	critical	critical	ADJ
cana-2597	71	5	for	for	ADP
cana-2597	71	6	ensuring	ensure	VERB
cana-2597	71	7	that	that	SCONJ
cana-2597	71	8	the	the	DET
cana-2597	71	9	system	system	NOUN
cana-2597	71	10	can	can	AUX
cana-2597	71	11	manage	manage	VERB
cana-2597	71	12	the	the	DET
cana-2597	71	13	large	large	ADJ
cana-2597	71	14	volume	volume	NOUN
cana-2597	71	15	of	of	ADP
cana-2597	71	16	tweets	tweet	NOUN
cana-2597	71	17	in	in	ADP
cana-2597	71	18	real	real	ADJ
cana-2597	71	19	-	-	PUNCT
cana-2597	71	20	time	time	NOUN
cana-2597	71	21	,	,	PUNCT
cana-2597	71	22	particularly	particularly	ADV
cana-2597	71	23	during	during	ADP
cana-2597	71	24	high	high	ADJ
cana-2597	71	25	-	-	PUNCT
cana-2597	71	26	traffic	traffic	NOUN
cana-2597	71	27	disaster	disaster	NOUN
cana-2597	71	28	periods	period	NOUN
cana-2597	71	29	.	.	PUNCT
cana-2597	72	1	3.3	3.3	NUM
cana-2597	72	2	kafka	kafka	PROPN
cana-2597	72	3	-	-	PUNCT
cana-2597	72	4	based	base	VERB
cana-2597	72	5	data	data	NOUN
cana-2597	72	6	ingestion	ingestion	NOUN
cana-2597	72	7	tweets	tweet	NOUN
cana-2597	72	8	are	be	AUX
cana-2597	72	9	streamed	stream	VERB
cana-2597	72	10	into	into	ADP
cana-2597	72	11	kafka	kafka	PROPN
cana-2597	72	12	topics	topic	NOUN
cana-2597	72	13	,	,	PUNCT
cana-2597	72	14	where	where	SCONJ
cana-2597	72	15	they	they	PRON
cana-2597	72	16	are	be	AUX
cana-2597	72	17	tagged	tag	VERB
cana-2597	72	18	with	with	ADP
cana-2597	72	19	disaster	disaster	NOUN
cana-2597	72	20	-	-	PUNCT
cana-2597	72	21	related	relate	VERB
cana-2597	72	22	metadata	metadata	NOUN
cana-2597	72	23	such	such	ADJ
cana-2597	72	24	as	as	ADP
cana-2597	72	25	"	"	PUNCT
cana-2597	72	26	earthquake	earthquake	NOUN
cana-2597	72	27	"	"	PUNCT
cana-2597	72	28	or	or	CCONJ
cana-2597	72	29	"	"	PUNCT
cana-2597	72	30	flood	flood	NOUN
cana-2597	72	31	.	.	PUNCT
cana-2597	72	32	"	"	PUNCT
cana-2597	73	1	kafka	kafka	PROPN
cana-2597	73	2	’s	’s	PART
cana-2597	73	3	partitioning	partitioning	NOUN
cana-2597	73	4	capability	capability	NOUN
cana-2597	73	5	plays	play	VERB
cana-2597	73	6	a	a	DET
cana-2597	73	7	significant	significant	ADJ
cana-2597	73	8	role	role	NOUN
cana-2597	73	9	in	in	ADP
cana-2597	73	10	this	this	DET
cana-2597	73	11	process	process	NOUN
cana-2597	73	12	,	,	PUNCT
cana-2597	73	13	allowing	allow	VERB
cana-2597	73	14	the	the	DET
cana-2597	73	15	system	system	NOUN
cana-2597	73	16	to	to	PART
cana-2597	73	17	divide	divide	VERB
cana-2597	73	18	and	and	CCONJ
cana-2597	73	19	handle	handle	VERB
cana-2597	73	20	tweets	tweet	NOUN
cana-2597	73	21	in	in	ADP
cana-2597	73	22	parallel	parallel	NOUN
cana-2597	73	23	based	base	VERB
cana-2597	73	24	on	on	ADP
cana-2597	73	25	metadata	metadata	NOUN
cana-2597	73	26	,	,	PUNCT
cana-2597	73	27	such	such	ADJ
cana-2597	73	28	as	as	ADP
cana-2597	73	29	disaster	disaster	NOUN
cana-2597	73	30	type	type	NOUN
cana-2597	73	31	or	or	CCONJ
cana-2597	73	32	geographic	geographic	ADJ
cana-2597	73	33	location	location	NOUN
cana-2597	73	34	.	.	PUNCT
cana-2597	74	1	this	this	PRON
cana-2597	74	2	enables	enable	VERB
cana-2597	74	3	the	the	DET
cana-2597	74	4	system	system	NOUN
cana-2597	74	5	to	to	PART
cana-2597	74	6	efficiently	efficiently	ADV
cana-2597	74	7	manage	manage	VERB
cana-2597	74	8	multiple	multiple	ADJ
cana-2597	74	9	streams	stream	NOUN
cana-2597	74	10	of	of	ADP
cana-2597	74	11	disaster	disaster	NOUN
cana-2597	74	12	-	-	PUNCT
cana-2597	74	13	related	relate	VERB
cana-2597	74	14	data	datum	NOUN
cana-2597	74	15	without	without	ADP
cana-2597	74	16	bottlenecks	bottleneck	NOUN
cana-2597	74	17	[	[	X
cana-2597	74	18	11	11	NUM
cana-2597	74	19	]	]	PUNCT
cana-2597	74	20	.	.	PUNCT
cana-2597	75	1	in	in	ADP
cana-2597	75	2	practice	practice	NOUN
cana-2597	75	3	,	,	PUNCT
cana-2597	75	4	the	the	DET
cana-2597	75	5	partitioning	partitioning	NOUN
cana-2597	75	6	ensures	ensure	VERB
cana-2597	75	7	that	that	SCONJ
cana-2597	75	8	different	different	ADJ
cana-2597	75	9	disaster	disaster	NOUN
cana-2597	75	10	events	event	NOUN
cana-2597	75	11	are	be	AUX
cana-2597	75	12	processed	process	VERB
cana-2597	75	13	simultaneously	simultaneously	ADV
cana-2597	75	14	,	,	PUNCT
cana-2597	75	15	offering	offer	VERB
cana-2597	75	16	real	real	ADJ
cana-2597	75	17	-	-	PUNCT
cana-2597	75	18	time	time	NOUN
cana-2597	75	19	insights	insight	NOUN
cana-2597	75	20	on	on	ADP
cana-2597	75	21	various	various	ADJ
cana-2597	75	22	ongoing	ongoing	ADJ
cana-2597	75	23	disasters	disaster	NOUN
cana-2597	75	24	.	.	PUNCT
cana-2597	76	1	each	each	DET
cana-2597	76	2	kafka	kafka	PROPN
cana-2597	76	3	topic	topic	NOUN
cana-2597	76	4	serves	serve	VERB
cana-2597	76	5	as	as	ADP
cana-2597	76	6	a	a	DET
cana-2597	76	7	queue	queue	NOUN
cana-2597	76	8	for	for	ADP
cana-2597	76	9	a	a	DET
cana-2597	76	10	particular	particular	ADJ
cana-2597	76	11	category	category	NOUN
cana-2597	76	12	,	,	PUNCT
cana-2597	76	13	which	which	PRON
cana-2597	76	14	guarantees	guarantee	VERB
cana-2597	76	15	that	that	SCONJ
cana-2597	76	16	no	no	DET
cana-2597	76	17	tweet	tweet	NOUN
cana-2597	76	18	is	be	AUX
cana-2597	76	19	lost	lose	VERB
cana-2597	76	20	or	or	CCONJ
cana-2597	76	21	delayed	delay	VERB
cana-2597	76	22	,	,	PUNCT
cana-2597	76	23	even	even	ADV
cana-2597	76	24	when	when	SCONJ
cana-2597	76	25	traffic	traffic	NOUN
cana-2597	76	26	spikes	spike	VERB
cana-2597	76	27	.	.	PUNCT
cana-2597	77	1	the	the	DET
cana-2597	77	2	combination	combination	NOUN
cana-2597	77	3	of	of	ADP
cana-2597	77	4	kafka	kafka	PROPN
cana-2597	77	5	's	's	PART
cana-2597	77	6	durability	durability	NOUN
cana-2597	77	7	and	and	CCONJ
cana-2597	77	8	parallelism	parallelism	NOUN
cana-2597	77	9	ensures	ensure	VERB
cana-2597	77	10	a	a	DET
cana-2597	77	11	continuous	continuous	ADJ
cana-2597	77	12	flow	flow	NOUN
cana-2597	77	13	of	of	ADP
cana-2597	77	14	data	datum	NOUN
cana-2597	77	15	into	into	ADP
cana-2597	77	16	the	the	DET
cana-2597	77	17	processing	processing	NOUN
cana-2597	77	18	system	system	NOUN
cana-2597	77	19	,	,	PUNCT
cana-2597	77	20	ready	ready	ADJ
cana-2597	77	21	for	for	ADP
cana-2597	77	22	real	real	ADJ
cana-2597	77	23	-	-	PUNCT
cana-2597	77	24	time	time	NOUN
cana-2597	77	25	analytics	analytic	NOUN
cana-2597	77	26	.	.	PUNCT
cana-2597	78	1	to	to	PART
cana-2597	78	2	avoid	avoid	VERB
cana-2597	78	3	reliance	reliance	NOUN
cana-2597	78	4	on	on	ADP
cana-2597	78	5	twitter	twitter	NOUN
cana-2597	78	6	's	's	PART
cana-2597	78	7	paid	pay	VERB
cana-2597	78	8	api	api	NOUN
cana-2597	78	9	,	,	PUNCT
cana-2597	78	10	we	we	PRON
cana-2597	78	11	leveraged	leveraged	ADJ
cana-2597	78	12	twscrape	twscrape	NOUN
cana-2597	78	13	,	,	PUNCT
cana-2597	78	14	a	a	DET
cana-2597	78	15	web	web	NOUN
cana-2597	78	16	-	-	PUNCT
cana-2597	78	17	scraping	scrape	VERB
cana-2597	78	18	tool	tool	NOUN
cana-2597	78	19	that	that	PRON
cana-2597	78	20	efficiently	efficiently	ADV
cana-2597	78	21	extracts	extract	VERB
cana-2597	78	22	tweets	tweet	NOUN
cana-2597	78	23	based	base	VERB
cana-2597	78	24	on	on	ADP
cana-2597	78	25	specific	specific	ADJ
cana-2597	78	26	disaster	disaster	NOUN
cana-2597	78	27	-	-	PUNCT
cana-2597	78	28	related	relate	VERB
cana-2597	78	29	queries	query	NOUN
cana-2597	78	30	.	.	PUNCT
cana-2597	79	1	the	the	DET
cana-2597	79	2	extracted	extract	VERB
cana-2597	79	3	tweets	tweet	NOUN
cana-2597	79	4	are	be	AUX
cana-2597	79	5	then	then	ADV
cana-2597	79	6	streamed	stream	VERB
cana-2597	79	7	into	into	ADP
cana-2597	79	8	apache	apache	PROPN
cana-2597	79	9	kafka	kafka	PROPN
cana-2597	79	10	in	in	ADP
cana-2597	79	11	real	real	ADJ
cana-2597	79	12	time	time	NOUN
cana-2597	79	13	,	,	PUNCT
cana-2597	79	14	where	where	SCONJ
cana-2597	79	15	each	each	DET
cana-2597	79	16	tweet	tweet	NOUN
cana-2597	79	17	is	be	AUX
cana-2597	79	18	stored	store	VERB
cana-2597	79	19	in	in	ADP
cana-2597	79	20	a	a	DET
cana-2597	79	21	structured	structured	ADJ
cana-2597	79	22	json	json	NOUN
cana-2597	79	23	format	format	NOUN
cana-2597	79	24	.	.	PUNCT
cana-2597	80	1	the	the	DET
cana-2597	80	2	key	key	ADJ
cana-2597	80	3	code	code	NOUN
cana-2597	80	4	snippets	snippet	NOUN
cana-2597	80	5	for	for	ADP
cana-2597	80	6	our	our	PRON
cana-2597	80	7	pipeline	pipeline	NOUN
cana-2597	80	8	are	be	AUX
cana-2597	80	9	as	as	SCONJ
cana-2597	80	10	follows	follow	VERB
cana-2597	80	11	:	:	PUNCT
cana-2597	80	12	communications	communication	NOUN
cana-2597	80	13	on	on	ADP
cana-2597	80	14	applied	apply	VERB
cana-2597	80	15	nonlinear	nonlinear	ADJ
cana-2597	80	16	analysis	analysis	NOUN
cana-2597	80	17	issn	issn	NOUN
cana-2597	80	18	:	:	PUNCT
cana-2597	80	19	1074	1074	NUM
cana-2597	80	20	-	-	PUNCT
cana-2597	80	21	133x	133x	NUM
cana-2597	80	22	vol	vol	NOUN
cana-2597	80	23	32	32	NUM
cana-2597	81	1	no	no	NOUN
cana-2597	81	2	.	.	PUNCT
cana-2597	82	1	3s	3s	NUM
cana-2597	82	2	(	(	PUNCT
cana-2597	82	3	2025	2025	NUM
cana-2597	82	4	)	)	PUNCT
cana-2597	82	5	202	202	NUM
cana-2597	82	6	https://internationalpubls.com	https://internationalpubls.com	PROPN
cana-2597	82	7	producer	producer	NOUN
cana-2597	82	8	code	code	NOUN
cana-2597	82	9	:	:	PUNCT
cana-2597	82	10	consumer	consumer	NOUN
cana-2597	82	11	code	code	NOUN
cana-2597	82	12	:	:	PUNCT
cana-2597	82	13	here	here	ADV
cana-2597	82	14	,	,	PUNCT
cana-2597	82	15	tweets	tweet	NOUN
cana-2597	82	16	are	be	AUX
cana-2597	82	17	scraped	scrape	VERB
cana-2597	82	18	asynchronously	asynchronously	ADV
cana-2597	82	19	and	and	CCONJ
cana-2597	82	20	sent	send	VERB
cana-2597	82	21	to	to	ADP
cana-2597	82	22	the	the	DET
cana-2597	82	23	disaster_tweets	disaster_tweet	NOUN
cana-2597	82	24	topic	topic	NOUN
cana-2597	82	25	in	in	ADP
cana-2597	82	26	kafka	kafka	PROPN
cana-2597	82	27	for	for	ADP
cana-2597	82	28	further	further	ADJ
cana-2597	82	29	processing	processing	NOUN
cana-2597	82	30	.	.	PUNCT
cana-2597	83	1	this	this	DET
cana-2597	83	2	consumer	consumer	NOUN
cana-2597	83	3	continuously	continuously	ADV
cana-2597	83	4	listens	listen	VERB
cana-2597	83	5	for	for	ADP
cana-2597	83	6	new	new	ADJ
cana-2597	83	7	tweets	tweet	NOUN
cana-2597	83	8	,	,	PUNCT
cana-2597	83	9	processing	process	VERB
cana-2597	83	10	them	they	PRON
cana-2597	83	11	in	in	ADP
cana-2597	83	12	real	real	ADJ
cana-2597	83	13	time	time	NOUN
cana-2597	83	14	,	,	PUNCT
cana-2597	83	15	which	which	PRON
cana-2597	83	16	is	be	AUX
cana-2597	83	17	essential	essential	ADJ
cana-2597	83	18	for	for	ADP
cana-2597	83	19	tasks	task	NOUN
cana-2597	83	20	like	like	ADP
cana-2597	83	21	disaster	disaster	NOUN
cana-2597	83	22	response	response	NOUN
cana-2597	83	23	monitoring	monitoring	NOUN
cana-2597	83	24	.	.	PUNCT
cana-2597	84	1	3.4	3.4	NUM
cana-2597	84	2	real	real	ADJ
cana-2597	84	3	-	-	PUNCT
cana-2597	84	4	time	time	NOUN
cana-2597	84	5	processing	processing	NOUN
cana-2597	84	6	with	with	ADP
cana-2597	84	7	spark	spark	NOUN
cana-2597	84	8	streaming	streaming	NOUN
cana-2597	84	9	once	once	SCONJ
cana-2597	84	10	the	the	DET
cana-2597	84	11	tweets	tweet	NOUN
cana-2597	84	12	are	be	AUX
cana-2597	84	13	ingested	ingest	VERB
cana-2597	84	14	into	into	ADP
cana-2597	84	15	kafka	kafka	PROPN
cana-2597	84	16	,	,	PUNCT
cana-2597	84	17	they	they	PRON
cana-2597	84	18	are	be	AUX
cana-2597	84	19	passed	pass	VERB
cana-2597	84	20	to	to	PART
cana-2597	84	21	apache	apache	VERB
cana-2597	84	22	spark	spark	NOUN
cana-2597	84	23	streaming	streaming	NOUN
cana-2597	84	24	,	,	PUNCT
cana-2597	84	25	which	which	PRON
cana-2597	84	26	processes	process	VERB
cana-2597	84	27	the	the	DET
cana-2597	84	28	data	datum	NOUN
cana-2597	84	29	in	in	ADP
cana-2597	84	30	micro	micro	NOUN
cana-2597	84	31	-	-	NOUN
cana-2597	84	32	batches	batch	NOUN
cana-2597	84	33	for	for	ADP
cana-2597	84	34	real	real	ADJ
cana-2597	84	35	-	-	PUNCT
cana-2597	84	36	time	time	NOUN
cana-2597	84	37	analytics	analytic	NOUN
cana-2597	84	38	.	.	PUNCT
cana-2597	85	1	key	key	ADJ
cana-2597	85	2	tasks	task	NOUN
cana-2597	85	3	performed	perform	VERB
cana-2597	85	4	during	during	ADP
cana-2597	85	5	this	this	DET
cana-2597	85	6	stage	stage	NOUN
cana-2597	85	7	include	include	VERB
cana-2597	85	8	sentiment	sentiment	NOUN
cana-2597	85	9	analysis	analysis	NOUN
cana-2597	85	10	,	,	PUNCT
cana-2597	85	11	where	where	SCONJ
cana-2597	85	12	text	text	NOUN
cana-2597	85	13	mining	mining	NOUN
cana-2597	85	14	techniques	technique	NOUN
cana-2597	85	15	are	be	AUX
cana-2597	85	16	applied	apply	VERB
cana-2597	85	17	to	to	PART
cana-2597	85	18	assess	assess	VERB
cana-2597	85	19	the	the	DET
cana-2597	85	20	sentiment	sentiment	NOUN
cana-2597	85	21	of	of	ADP
cana-2597	85	22	tweets	tweet	NOUN
cana-2597	85	23	.	.	PUNCT
cana-2597	86	1	this	this	PRON
cana-2597	86	2	provides	provide	VERB
cana-2597	86	3	immediate	immediate	ADJ
cana-2597	86	4	insight	insight	NOUN
cana-2597	86	5	into	into	ADP
cana-2597	86	6	public	public	ADJ
cana-2597	86	7	reactions	reaction	NOUN
cana-2597	86	8	and	and	CCONJ
cana-2597	86	9	emotions	emotion	NOUN
cana-2597	86	10	concerning	concern	VERB
cana-2597	86	11	specific	specific	ADJ
cana-2597	86	12	disasters	disaster	NOUN
cana-2597	86	13	[	[	X
cana-2597	86	14	12	12	NUM
cana-2597	86	15	]	]	PUNCT
cana-2597	86	16	.	.	PUNCT
cana-2597	87	1	additionally	additionally	ADV
cana-2597	87	2	,	,	PUNCT
cana-2597	87	3	spark	spark	NOUN
cana-2597	87	4	’s	’s	PART
cana-2597	87	5	nlp	nlp	NOUN
cana-2597	87	6	tools	tool	NOUN
cana-2597	87	7	are	be	AUX
cana-2597	87	8	used	use	VERB
cana-2597	87	9	to	to	PART
cana-2597	87	10	extract	extract	VERB
cana-2597	87	11	key	key	ADJ
cana-2597	87	12	disaster	disaster	NOUN
cana-2597	87	13	-	-	PUNCT
cana-2597	87	14	related	relate	VERB
cana-2597	87	15	keywords	keyword	NOUN
cana-2597	87	16	and	and	CCONJ
cana-2597	87	17	location	location	NOUN
cana-2597	87	18	data	datum	NOUN
cana-2597	87	19	from	from	ADP
cana-2597	87	20	the	the	DET
cana-2597	87	21	tweets	tweet	NOUN
cana-2597	87	22	,	,	PUNCT
cana-2597	87	23	helping	help	VERB
cana-2597	87	24	in	in	ADP
cana-2597	87	25	the	the	DET
cana-2597	87	26	identification	identification	NOUN
cana-2597	87	27	of	of	ADP
cana-2597	87	28	affected	affected	ADJ
cana-2597	87	29	regions	region	NOUN
cana-2597	87	30	and	and	CCONJ
cana-2597	87	31	relevant	relevant	ADJ
cana-2597	87	32	discussions	discussion	NOUN
cana-2597	87	33	.	.	PUNCT
cana-2597	88	1	the	the	DET
cana-2597	88	2	use	use	NOUN
cana-2597	88	3	of	of	ADP
cana-2597	88	4	spark	spark	NOUN
cana-2597	88	5	ensures	ensure	VERB
cana-2597	88	6	that	that	SCONJ
cana-2597	88	7	the	the	DET
cana-2597	88	8	system	system	NOUN
cana-2597	88	9	can	can	AUX
cana-2597	88	10	scale	scale	VERB
cana-2597	88	11	to	to	PART
cana-2597	88	12	handle	handle	VERB
cana-2597	88	13	massive	massive	ADJ
cana-2597	88	14	amounts	amount	NOUN
cana-2597	88	15	of	of	ADP
cana-2597	88	16	tweet	tweet	NOUN
cana-2597	88	17	data	datum	NOUN
cana-2597	88	18	efficiently	efficiently	ADV
cana-2597	88	19	,	,	PUNCT
cana-2597	88	20	even	even	ADV
cana-2597	88	21	during	during	ADP
cana-2597	88	22	disaster	disaster	NOUN
cana-2597	88	23	events	event	NOUN
cana-2597	88	24	with	with	ADP
cana-2597	88	25	high	high	ADJ
cana-2597	88	26	tweet	tweet	NOUN
cana-2597	88	27	volumes	volume	NOUN
cana-2597	88	28	.	.	PUNCT
cana-2597	89	1	by	by	ADP
cana-2597	89	2	leveraging	leverage	VERB
cana-2597	89	3	its	its	PRON
cana-2597	89	4	distributed	distribute	VERB
cana-2597	89	5	processing	processing	NOUN
cana-2597	89	6	capabilities	capability	NOUN
cana-2597	89	7	,	,	PUNCT
cana-2597	89	8	the	the	DET
cana-2597	89	9	system	system	NOUN
cana-2597	89	10	guarantees	guarantee	VERB
cana-2597	89	11	that	that	SCONJ
cana-2597	89	12	the	the	DET
cana-2597	89	13	data	data	NOUN
cana-2597	89	14	is	be	AUX
cana-2597	89	15	processed	process	VERB
cana-2597	89	16	quickly	quickly	ADV
cana-2597	89	17	and	and	CCONJ
cana-2597	89	18	effectively	effectively	ADV
cana-2597	89	19	,	,	PUNCT
cana-2597	89	20	making	make	VERB
cana-2597	89	21	the	the	DET
cana-2597	89	22	insights	insight	NOUN
cana-2597	89	23	available	available	ADJ
cana-2597	89	24	in	in	ADP
cana-2597	89	25	real	real	ADJ
cana-2597	89	26	-	-	PUNCT
cana-2597	89	27	time	time	NOUN
cana-2597	89	28	.	.	PUNCT
cana-2597	90	1	this	this	DET
cana-2597	90	2	approach	approach	NOUN
cana-2597	90	3	ensures	ensure	VERB
cana-2597	90	4	fast	fast	ADJ
cana-2597	90	5	,	,	PUNCT
cana-2597	90	6	scalable	scalable	ADJ
cana-2597	90	7	,	,	PUNCT
cana-2597	90	8	and	and	CCONJ
cana-2597	90	9	reliable	reliable	ADJ
cana-2597	90	10	processing	processing	NOUN
cana-2597	90	11	,	,	PUNCT
cana-2597	90	12	which	which	PRON
cana-2597	90	13	is	be	AUX
cana-2597	90	14	essential	essential	ADJ
cana-2597	90	15	during	during	ADP
cana-2597	90	16	critical	critical	ADJ
cana-2597	90	17	disaster	disaster	NOUN
cana-2597	90	18	scenarios	scenario	NOUN
cana-2597	90	19	.	.	PUNCT
cana-2597	91	1	3.5	3.5	NUM
cana-2597	91	2	multimodal	multimodal	NOUN
cana-2597	91	3	model	model	NOUN
cana-2597	91	4	for	for	ADP
cana-2597	91	5	tweet	tweet	NOUN
cana-2597	91	6	categorization	categorization	NOUN
cana-2597	91	7	disaster	disaster	NOUN
cana-2597	91	8	-	-	PUNCT
cana-2597	91	9	related	relate	VERB
cana-2597	91	10	tweets	tweet	NOUN
cana-2597	91	11	often	often	ADV
cana-2597	91	12	contain	contain	VERB
cana-2597	91	13	both	both	DET
cana-2597	91	14	textual	textual	ADJ
cana-2597	91	15	and	and	CCONJ
cana-2597	91	16	visual	visual	ADJ
cana-2597	91	17	information	information	NOUN
cana-2597	91	18	,	,	PUNCT
cana-2597	91	19	making	make	VERB
cana-2597	91	20	a	a	DET
cana-2597	91	21	multimodal	multimodal	ADJ
cana-2597	91	22	approach	approach	NOUN
cana-2597	91	23	necessary	necessary	ADJ
cana-2597	91	24	for	for	ADP
cana-2597	91	25	accurate	accurate	ADJ
cana-2597	91	26	categorization	categorization	NOUN
cana-2597	91	27	.	.	PUNCT
cana-2597	92	1	to	to	PART
cana-2597	92	2	handle	handle	VERB
cana-2597	92	3	this	this	PRON
cana-2597	92	4	,	,	PUNCT
cana-2597	92	5	the	the	DET
cana-2597	92	6	system	system	NOUN
cana-2597	92	7	employs	employ	VERB
cana-2597	92	8	natural	natural	ADJ
cana-2597	92	9	language	language	NOUN
cana-2597	92	10	processing	processing	NOUN
cana-2597	92	11	(	(	PUNCT
cana-2597	92	12	nlp	nlp	NOUN
cana-2597	92	13	)	)	PUNCT
cana-2597	92	14	models	model	NOUN
cana-2597	92	15	for	for	ADP
cana-2597	92	16	text	text	NOUN
cana-2597	92	17	and	and	CCONJ
cana-2597	92	18	convolutional	convolutional	ADJ
cana-2597	92	19	neural	neural	ADJ
cana-2597	92	20	networks	network	NOUN
cana-2597	92	21	(	(	PUNCT
cana-2597	92	22	cnns	cnns	PROPN
cana-2597	92	23	)	)	PUNCT
cana-2597	92	24	for	for	ADP
cana-2597	92	25	images	image	NOUN
cana-2597	92	26	.	.	PUNCT
cana-2597	93	1	the	the	DET
cana-2597	93	2	integration	integration	NOUN
cana-2597	93	3	of	of	ADP
cana-2597	93	4	these	these	DET
cana-2597	93	5	two	two	NUM
cana-2597	93	6	modalities	modality	NOUN
cana-2597	93	7	improves	improve	VERB
cana-2597	93	8	the	the	DET
cana-2597	93	9	classification	classification	NOUN
cana-2597	93	10	of	of	ADP
cana-2597	93	11	disaster	disaster	NOUN
cana-2597	93	12	-	-	PUNCT
cana-2597	93	13	related	relate	VERB
cana-2597	93	14	tweets	tweet	NOUN
cana-2597	93	15	.	.	PUNCT
cana-2597	94	1	communications	communication	NOUN
cana-2597	94	2	on	on	ADP
cana-2597	94	3	applied	apply	VERB
cana-2597	94	4	nonlinear	nonlinear	ADJ
cana-2597	94	5	analysis	analysis	NOUN
cana-2597	94	6	issn	issn	NOUN
cana-2597	94	7	:	:	PUNCT
cana-2597	94	8	1074	1074	NUM
cana-2597	94	9	-	-	PUNCT
cana-2597	94	10	133x	133x	NUM
cana-2597	94	11	vol	vol	NOUN
cana-2597	94	12	32	32	NUM
cana-2597	94	13	no	no	NOUN
cana-2597	94	14	.	.	PUNCT
cana-2597	95	1	3s	3s	NUM
cana-2597	95	2	(	(	PUNCT
cana-2597	95	3	2025	2025	NUM
cana-2597	95	4	)	)	PUNCT
cana-2597	95	5	203	203	NUM
cana-2597	95	6	https://internationalpubls.com	https://internationalpubls.com	SYM
cana-2597	95	7	3.5.1	3.5.1	NUM
cana-2597	95	8	text	text	NOUN
cana-2597	95	9	processing	processing	NOUN
cana-2597	95	10	using	use	VERB
cana-2597	95	11	nlp	nlp	NOUN
cana-2597	95	12	models	model	NOUN
cana-2597	95	13	for	for	ADP
cana-2597	95	14	text	text	NOUN
cana-2597	95	15	classification	classification	NOUN
cana-2597	95	16	,	,	PUNCT
cana-2597	95	17	bert	bert	PROPN
cana-2597	95	18	(	(	PUNCT
cana-2597	95	19	bidirectional	bidirectional	ADJ
cana-2597	95	20	encoder	encoder	NOUN
cana-2597	95	21	representations	representation	VERB
cana-2597	95	22	from	from	ADP
cana-2597	95	23	transformers	transformer	NOUN
cana-2597	95	24	)	)	PUNCT
cana-2597	95	25	and	and	CCONJ
cana-2597	95	26	xlnet	xlnet	PROPN
cana-2597	95	27	models	model	NOUN
cana-2597	95	28	are	be	AUX
cana-2597	95	29	fine	fine	ADV
cana-2597	95	30	-	-	PUNCT
cana-2597	95	31	tuned	tune	VERB
cana-2597	95	32	on	on	ADP
cana-2597	95	33	a	a	DET
cana-2597	95	34	dataset	dataset	NOUN
cana-2597	95	35	of	of	ADP
cana-2597	95	36	disaster	disaster	NOUN
cana-2597	95	37	-	-	PUNCT
cana-2597	95	38	related	relate	VERB
cana-2597	95	39	tweets	tweet	NOUN
cana-2597	95	40	.	.	PUNCT
cana-2597	96	1	these	these	DET
cana-2597	96	2	models	model	NOUN
cana-2597	96	3	excel	excel	VERB
cana-2597	96	4	at	at	ADP
cana-2597	96	5	contextual	contextual	ADJ
cana-2597	96	6	understanding	understanding	NOUN
cana-2597	96	7	,	,	PUNCT
cana-2597	96	8	which	which	PRON
cana-2597	96	9	is	be	AUX
cana-2597	96	10	essential	essential	ADJ
cana-2597	96	11	for	for	ADP
cana-2597	96	12	accurately	accurately	ADV
cana-2597	96	13	identifying	identify	VERB
cana-2597	96	14	disaster	disaster	NOUN
cana-2597	96	15	-	-	PUNCT
cana-2597	96	16	related	relate	VERB
cana-2597	96	17	information	information	NOUN
cana-2597	96	18	within	within	ADP
cana-2597	96	19	short	short	ADJ
cana-2597	96	20	,	,	PUNCT
cana-2597	96	21	often	often	ADV
cana-2597	96	22	informal	informal	ADJ
cana-2597	96	23	tweets	tweet	NOUN
cana-2597	96	24	.	.	PUNCT
cana-2597	97	1	bert	bert	PROPN
cana-2597	97	2	,	,	PUNCT
cana-2597	97	3	in	in	ADP
cana-2597	97	4	particular	particular	ADJ
cana-2597	97	5	,	,	PUNCT
cana-2597	97	6	considers	consider	VERB
cana-2597	97	7	the	the	DET
cana-2597	97	8	context	context	NOUN
cana-2597	97	9	of	of	ADP
cana-2597	97	10	words	word	NOUN
cana-2597	97	11	bidirectionally	bidirectionally	ADV
cana-2597	97	12	,	,	PUNCT
cana-2597	97	13	which	which	PRON
cana-2597	97	14	allows	allow	VERB
cana-2597	97	15	for	for	ADP
cana-2597	97	16	a	a	DET
cana-2597	97	17	more	more	ADV
cana-2597	97	18	nuanced	nuanced	ADJ
cana-2597	97	19	understanding	understanding	NOUN
cana-2597	97	20	of	of	ADP
cana-2597	97	21	the	the	DET
cana-2597	97	22	text	text	NOUN
cana-2597	97	23	.	.	PUNCT
cana-2597	98	1	this	this	DET
cana-2597	98	2	bidirectional	bidirectional	ADJ
cana-2597	98	3	context	context	NOUN
cana-2597	98	4	improves	improve	VERB
cana-2597	98	5	the	the	DET
cana-2597	98	6	model	model	NOUN
cana-2597	98	7	’s	’s	PART
cana-2597	98	8	ability	ability	NOUN
cana-2597	98	9	to	to	PART
cana-2597	98	10	classify	classify	VERB
cana-2597	98	11	tweets	tweet	NOUN
cana-2597	98	12	with	with	ADP
cana-2597	98	13	higher	high	ADJ
cana-2597	98	14	accuracy	accuracy	NOUN
cana-2597	98	15	[	[	X
cana-2597	98	16	13	13	NUM
cana-2597	98	17	]	]	PUNCT
cana-2597	98	18	.	.	PUNCT
cana-2597	99	1	xlnet	xlnet	PROPN
cana-2597	99	2	,	,	PUNCT
cana-2597	99	3	on	on	ADP
cana-2597	99	4	the	the	DET
cana-2597	99	5	other	other	ADJ
cana-2597	99	6	hand	hand	NOUN
cana-2597	99	7	,	,	PUNCT
cana-2597	99	8	is	be	AUX
cana-2597	99	9	trained	train	VERB
cana-2597	99	10	using	use	VERB
cana-2597	99	11	a	a	DET
cana-2597	99	12	permutation	permutation	NOUN
cana-2597	99	13	-	-	PUNCT
cana-2597	99	14	based	base	VERB
cana-2597	99	15	language	language	NOUN
cana-2597	99	16	modeling	modeling	NOUN
cana-2597	99	17	approach	approach	NOUN
cana-2597	99	18	,	,	PUNCT
cana-2597	99	19	which	which	PRON
cana-2597	99	20	allows	allow	VERB
cana-2597	99	21	it	it	PRON
cana-2597	99	22	to	to	PART
cana-2597	99	23	handle	handle	VERB
cana-2597	99	24	longer	long	ADJ
cana-2597	99	25	sequences	sequence	NOUN
cana-2597	99	26	more	more	ADV
cana-2597	99	27	effectively	effectively	ADV
cana-2597	99	28	.	.	PUNCT
cana-2597	100	1	this	this	PRON
cana-2597	100	2	makes	make	VERB
cana-2597	100	3	xlnet	xlnet	PROPN
cana-2597	100	4	particularly	particularly	ADV
cana-2597	100	5	suited	suit	VERB
cana-2597	100	6	for	for	ADP
cana-2597	100	7	longer	long	ADJ
cana-2597	100	8	tweets	tweet	NOUN
cana-2597	100	9	or	or	CCONJ
cana-2597	100	10	more	more	ADJ
cana-2597	100	11	complex	complex	ADJ
cana-2597	100	12	sentence	sentence	NOUN
cana-2597	100	13	structures	structure	NOUN
cana-2597	100	14	,	,	PUNCT
cana-2597	100	15	where	where	SCONJ
cana-2597	100	16	context	context	NOUN
cana-2597	100	17	and	and	CCONJ
cana-2597	100	18	sequence	sequence	NOUN
cana-2597	100	19	are	be	AUX
cana-2597	100	20	important	important	ADJ
cana-2597	100	21	for	for	ADP
cana-2597	100	22	classification	classification	NOUN
cana-2597	100	23	accuracy	accuracy	NOUN
cana-2597	100	24	[	[	X
cana-2597	100	25	14	14	NUM
cana-2597	100	26	]	]	PUNCT
cana-2597	100	27	.	.	PUNCT
cana-2597	101	1	both	both	DET
cana-2597	101	2	models	model	NOUN
cana-2597	101	3	,	,	PUNCT
cana-2597	101	4	when	when	SCONJ
cana-2597	101	5	fine	fine	ADV
cana-2597	101	6	-	-	PUNCT
cana-2597	101	7	tuned	tune	VERB
cana-2597	101	8	,	,	PUNCT
cana-2597	101	9	enhance	enhance	VERB
cana-2597	101	10	the	the	DET
cana-2597	101	11	system	system	NOUN
cana-2597	101	12	's	's	PART
cana-2597	101	13	ability	ability	NOUN
cana-2597	101	14	to	to	PART
cana-2597	101	15	differentiate	differentiate	VERB
cana-2597	101	16	disaster	disaster	NOUN
cana-2597	101	17	-	-	PUNCT
cana-2597	101	18	related	relate	VERB
cana-2597	101	19	tweets	tweet	NOUN
cana-2597	101	20	from	from	ADP
cana-2597	101	21	unrelated	unrelated	ADJ
cana-2597	101	22	content	content	NOUN
cana-2597	101	23	based	base	VERB
cana-2597	101	24	on	on	ADP
cana-2597	101	25	their	their	PRON
cana-2597	101	26	context	context	NOUN
cana-2597	101	27	.	.	PUNCT
cana-2597	102	1	3.5.2	3.5.2	NUM
cana-2597	102	2	image	image	NOUN
cana-2597	102	3	processing	processing	NOUN
cana-2597	102	4	using	use	VERB
cana-2597	102	5	cnns	cnn	NOUN
cana-2597	102	6	in	in	ADP
cana-2597	102	7	addition	addition	NOUN
cana-2597	102	8	to	to	ADP
cana-2597	102	9	text	text	NOUN
cana-2597	102	10	,	,	PUNCT
cana-2597	102	11	many	many	ADJ
cana-2597	102	12	disaster	disaster	NOUN
cana-2597	102	13	-	-	PUNCT
cana-2597	102	14	related	relate	VERB
cana-2597	102	15	tweets	tweet	NOUN
cana-2597	102	16	include	include	VERB
cana-2597	102	17	images	image	NOUN
cana-2597	102	18	that	that	PRON
cana-2597	102	19	provide	provide	VERB
cana-2597	102	20	visual	visual	ADJ
cana-2597	102	21	evidence	evidence	NOUN
cana-2597	102	22	of	of	ADP
cana-2597	102	23	the	the	DET
cana-2597	102	24	event	event	NOUN
cana-2597	102	25	.	.	PUNCT
cana-2597	103	1	for	for	ADP
cana-2597	103	2	this	this	PRON
cana-2597	103	3	,	,	PUNCT
cana-2597	103	4	the	the	DET
cana-2597	103	5	system	system	NOUN
cana-2597	103	6	uses	use	VERB
cana-2597	103	7	efficientnetb3	efficientnetb3	PROPN
cana-2597	103	8	,	,	PUNCT
cana-2597	103	9	densenet201	densenet201	PROPN
cana-2597	103	10	,	,	PUNCT
cana-2597	103	11	and	and	CCONJ
cana-2597	103	12	resnet50	resnet50	NOUN
cana-2597	103	13	—	—	PUNCT
cana-2597	103	14	all	all	DET
cana-2597	103	15	powerful	powerful	ADJ
cana-2597	103	16	cnn	cnn	PROPN
cana-2597	103	17	architectures	architecture	NOUN
cana-2597	103	18	trained	train	VERB
cana-2597	103	19	on	on	ADP
cana-2597	103	20	disaster	disaster	NOUN
cana-2597	103	21	imagery	imagery	NOUN
cana-2597	103	22	datasets	dataset	NOUN
cana-2597	103	23	.	.	PUNCT
cana-2597	104	1	among	among	ADP
cana-2597	104	2	these	these	PRON
cana-2597	104	3	,	,	PUNCT
cana-2597	104	4	efficientnetb3	efficientnetb3	PROPN
cana-2597	104	5	demonstrated	demonstrate	VERB
cana-2597	104	6	the	the	DET
cana-2597	104	7	best	good	ADJ
cana-2597	104	8	performance	performance	NOUN
cana-2597	104	9	,	,	PUNCT
cana-2597	104	10	striking	strike	VERB
cana-2597	104	11	a	a	DET
cana-2597	104	12	balance	balance	NOUN
cana-2597	104	13	between	between	ADP
cana-2597	104	14	accuracy	accuracy	NOUN
cana-2597	104	15	and	and	CCONJ
cana-2597	104	16	computational	computational	ADJ
cana-2597	104	17	cost	cost	NOUN
cana-2597	104	18	.	.	PUNCT
cana-2597	105	1	this	this	PRON
cana-2597	105	2	makes	make	VERB
cana-2597	105	3	it	it	PRON
cana-2597	105	4	particularly	particularly	ADV
cana-2597	105	5	suitable	suitable	ADJ
cana-2597	105	6	for	for	ADP
cana-2597	105	7	real	real	ADJ
cana-2597	105	8	-	-	PUNCT
cana-2597	105	9	time	time	NOUN
cana-2597	105	10	image	image	NOUN
cana-2597	105	11	classification	classification	NOUN
cana-2597	105	12	in	in	ADP
cana-2597	105	13	disaster	disaster	NOUN
cana-2597	105	14	scenarios	scenario	NOUN
cana-2597	105	15	,	,	PUNCT
cana-2597	105	16	where	where	SCONJ
cana-2597	105	17	time	time	NOUN
cana-2597	105	18	is	be	AUX
cana-2597	105	19	of	of	ADP
cana-2597	105	20	the	the	DET
cana-2597	105	21	essence	essence	NOUN
cana-2597	105	22	and	and	CCONJ
cana-2597	105	23	resource	resource	NOUN
cana-2597	105	24	efficiency	efficiency	NOUN
cana-2597	105	25	is	be	AUX
cana-2597	105	26	crucial	crucial	ADJ
cana-2597	105	27	[	[	X
cana-2597	105	28	14	14	NUM
cana-2597	105	29	]	]	PUNCT
cana-2597	105	30	.	.	PUNCT
cana-2597	106	1	densenet201	densenet201	PROPN
cana-2597	106	2	and	and	CCONJ
cana-2597	106	3	resnet50	resnet50	NOUN
cana-2597	106	4	also	also	ADV
cana-2597	106	5	contribute	contribute	VERB
cana-2597	106	6	to	to	ADP
cana-2597	106	7	the	the	DET
cana-2597	106	8	model	model	NOUN
cana-2597	106	9	’s	’s	PART
cana-2597	106	10	ability	ability	NOUN
cana-2597	106	11	to	to	PART
cana-2597	106	12	classify	classify	VERB
cana-2597	106	13	images	image	NOUN
cana-2597	106	14	by	by	ADP
cana-2597	106	15	leveraging	leverage	VERB
cana-2597	106	16	their	their	PRON
cana-2597	106	17	depth	depth	NOUN
cana-2597	106	18	and	and	CCONJ
cana-2597	106	19	feature	feature	NOUN
cana-2597	106	20	extraction	extraction	NOUN
cana-2597	106	21	capabilities	capability	NOUN
cana-2597	106	22	.	.	PUNCT
cana-2597	107	1	by	by	ADP
cana-2597	107	2	using	use	VERB
cana-2597	107	3	a	a	DET
cana-2597	107	4	combination	combination	NOUN
cana-2597	107	5	of	of	ADP
cana-2597	107	6	these	these	DET
cana-2597	107	7	cnns	cnn	NOUN
cana-2597	107	8	,	,	PUNCT
cana-2597	107	9	the	the	DET
cana-2597	107	10	system	system	NOUN
cana-2597	107	11	ensures	ensure	VERB
cana-2597	107	12	that	that	SCONJ
cana-2597	107	13	the	the	DET
cana-2597	107	14	visual	visual	ADJ
cana-2597	107	15	data	datum	NOUN
cana-2597	107	16	from	from	ADP
cana-2597	107	17	tweets	tweet	NOUN
cana-2597	107	18	is	be	AUX
cana-2597	107	19	processed	process	VERB
cana-2597	107	20	effectively	effectively	ADV
cana-2597	107	21	,	,	PUNCT
cana-2597	107	22	providing	provide	VERB
cana-2597	107	23	additional	additional	ADJ
cana-2597	107	24	context	context	NOUN
cana-2597	107	25	to	to	ADP
cana-2597	107	26	the	the	DET
cana-2597	107	27	disaster	disaster	NOUN
cana-2597	107	28	categorization	categorization	NOUN
cana-2597	107	29	.	.	PUNCT
cana-2597	108	1	3.5.3	3.5.3	NUM
cana-2597	108	2	multimodal	multimodal	NOUN
cana-2597	108	3	fusion	fusion	NOUN
cana-2597	108	4	model	model	NOUN
cana-2597	108	5	architecture	architecture	NOUN
cana-2597	108	6	the	the	DET
cana-2597	108	7	integration	integration	NOUN
cana-2597	108	8	of	of	ADP
cana-2597	108	9	text	text	NOUN
cana-2597	108	10	and	and	CCONJ
cana-2597	108	11	image	image	NOUN
cana-2597	108	12	data	datum	NOUN
cana-2597	108	13	occurs	occur	VERB
cana-2597	108	14	in	in	ADP
cana-2597	108	15	the	the	DET
cana-2597	108	16	multimodal	multimodal	NOUN
cana-2597	108	17	fusion	fusion	NOUN
cana-2597	108	18	model	model	NOUN
cana-2597	108	19	,	,	PUNCT
cana-2597	108	20	as	as	SCONJ
cana-2597	108	21	depicted	depict	VERB
cana-2597	108	22	in	in	ADP
cana-2597	108	23	fig	fig	NOUN
cana-2597	108	24	.	.	PUNCT
cana-2597	109	1	2	2	X
cana-2597	109	2	.	.	X
cana-2597	109	3	this	this	DET
cana-2597	109	4	internal	internal	ADJ
cana-2597	109	5	flow	flow	NOUN
cana-2597	109	6	diagram	diagram	NOUN
cana-2597	109	7	shows	show	VERB
cana-2597	109	8	how	how	SCONJ
cana-2597	109	9	features	feature	NOUN
cana-2597	109	10	extracted	extract	VERB
cana-2597	109	11	from	from	ADP
cana-2597	109	12	both	both	DET
cana-2597	109	13	modalities	modality	NOUN
cana-2597	109	14	(	(	PUNCT
cana-2597	109	15	text	text	NOUN
cana-2597	109	16	and	and	CCONJ
cana-2597	109	17	image	image	NOUN
cana-2597	109	18	)	)	PUNCT
cana-2597	109	19	are	be	AUX
cana-2597	109	20	combined	combine	VERB
cana-2597	109	21	before	before	ADP
cana-2597	109	22	the	the	DET
cana-2597	109	23	final	final	ADJ
cana-2597	109	24	classification	classification	NOUN
cana-2597	109	25	.	.	PUNCT
cana-2597	110	1	the	the	DET
cana-2597	110	2	fusion	fusion	NOUN
cana-2597	110	3	of	of	ADP
cana-2597	110	4	both	both	DET
cana-2597	110	5	textual	textual	ADJ
cana-2597	110	6	and	and	CCONJ
cana-2597	110	7	visual	visual	ADJ
cana-2597	110	8	features	feature	NOUN
cana-2597	110	9	enables	enable	VERB
cana-2597	110	10	the	the	DET
cana-2597	110	11	model	model	NOUN
cana-2597	110	12	to	to	PART
cana-2597	110	13	more	more	ADV
cana-2597	110	14	accurately	accurately	ADV
cana-2597	110	15	detect	detect	VERB
cana-2597	110	16	disaster	disaster	NOUN
cana-2597	110	17	-	-	PUNCT
cana-2597	110	18	related	relate	VERB
cana-2597	110	19	content	content	NOUN
cana-2597	110	20	.	.	PUNCT
cana-2597	111	1	by	by	ADP
cana-2597	111	2	leveraging	leverage	VERB
cana-2597	111	3	both	both	DET
cana-2597	111	4	types	type	NOUN
cana-2597	111	5	of	of	ADP
cana-2597	111	6	data	datum	NOUN
cana-2597	111	7	,	,	PUNCT
cana-2597	111	8	the	the	DET
cana-2597	111	9	system	system	NOUN
cana-2597	111	10	improves	improve	VERB
cana-2597	111	11	overall	overall	ADJ
cana-2597	111	12	classification	classification	NOUN
cana-2597	111	13	performance	performance	NOUN
cana-2597	111	14	,	,	PUNCT
cana-2597	111	15	ensuring	ensure	VERB
cana-2597	111	16	that	that	SCONJ
cana-2597	111	17	even	even	ADV
cana-2597	111	18	tweets	tweet	NOUN
cana-2597	111	19	with	with	ADP
cana-2597	111	20	limited	limited	ADJ
cana-2597	111	21	text	text	NOUN
cana-2597	111	22	but	but	CCONJ
cana-2597	111	23	strong	strong	ADJ
cana-2597	111	24	visual	visual	ADJ
cana-2597	111	25	content	content	NOUN
cana-2597	111	26	—	—	PUNCT
cana-2597	111	27	or	or	CCONJ
cana-2597	111	28	vice	vice	NOUN
cana-2597	111	29	versa	versa	ADV
cana-2597	111	30	—	—	PUNCT
cana-2597	111	31	are	be	AUX
cana-2597	111	32	accurately	accurately	ADV
cana-2597	111	33	categorized	categorize	VERB
cana-2597	111	34	.	.	PUNCT
cana-2597	112	1	communications	communication	NOUN
cana-2597	112	2	on	on	ADP
cana-2597	112	3	applied	apply	VERB
cana-2597	112	4	nonlinear	nonlinear	ADJ
cana-2597	112	5	analysis	analysis	NOUN
cana-2597	112	6	issn	issn	NOUN
cana-2597	112	7	:	:	PUNCT
cana-2597	112	8	1074	1074	NUM
cana-2597	112	9	-	-	PUNCT
cana-2597	112	10	133x	133x	NUM
cana-2597	112	11	vol	vol	NOUN
cana-2597	112	12	32	32	NUM
cana-2597	112	13	no	no	NOUN
cana-2597	112	14	.	.	PUNCT
cana-2597	113	1	3s	3s	NUM
cana-2597	113	2	(	(	PUNCT
cana-2597	113	3	2025	2025	NUM
cana-2597	113	4	)	)	PUNCT
cana-2597	113	5	204	204	NUM
cana-2597	113	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2597	113	7	fig	fig	NOUN
cana-2597	113	8	2	2	NUM
cana-2597	113	9	.	.	PUNCT
cana-2597	113	10	multimodal	multimodal	ADJ
cana-2597	113	11	fusion	fusion	NOUN
cana-2597	113	12	model	model	NOUN
cana-2597	113	13	architecture	architecture	NOUN
cana-2597	113	14	this	this	DET
cana-2597	113	15	multimodal	multimodal	NOUN
cana-2597	113	16	approach	approach	NOUN
cana-2597	113	17	provides	provide	VERB
cana-2597	113	18	a	a	DET
cana-2597	113	19	comprehensive	comprehensive	ADJ
cana-2597	113	20	solution	solution	NOUN
cana-2597	113	21	to	to	ADP
cana-2597	113	22	classifying	classify	VERB
cana-2597	113	23	disaster	disaster	NOUN
cana-2597	113	24	-	-	PUNCT
cana-2597	113	25	related	relate	VERB
cana-2597	113	26	tweets	tweet	NOUN
cana-2597	113	27	,	,	PUNCT
cana-2597	113	28	enhancing	enhance	VERB
cana-2597	113	29	the	the	DET
cana-2597	113	30	reliability	reliability	NOUN
cana-2597	113	31	of	of	ADP
cana-2597	113	32	detection	detection	NOUN
cana-2597	113	33	by	by	ADP
cana-2597	113	34	utilizing	utilize	VERB
cana-2597	113	35	both	both	DET
cana-2597	113	36	language	language	NOUN
cana-2597	113	37	and	and	CCONJ
cana-2597	113	38	imagery	imagery	NOUN
cana-2597	113	39	.	.	PUNCT
cana-2597	114	1	3	3	X
cana-2597	114	2	.	.	X
cana-2597	114	3	fusion	fusion	NOUN
cana-2597	114	4	of	of	ADP
cana-2597	114	5	text	text	NOUN
cana-2597	114	6	and	and	CCONJ
cana-2597	114	7	image	image	NOUN
cana-2597	114	8	data	datum	NOUN
cana-2597	114	9	the	the	DET
cana-2597	114	10	system	system	NOUN
cana-2597	114	11	enhances	enhance	VERB
cana-2597	114	12	tweet	tweet	NOUN
cana-2597	114	13	classification	classification	NOUN
cana-2597	114	14	by	by	ADP
cana-2597	114	15	combining	combine	VERB
cana-2597	114	16	text	text	NOUN
cana-2597	114	17	and	and	CCONJ
cana-2597	114	18	image	image	NOUN
cana-2597	114	19	features	feature	VERB
cana-2597	114	20	through	through	ADP
cana-2597	114	21	a	a	DET
cana-2597	114	22	feature	feature	NOUN
cana-2597	114	23	-	-	PUNCT
cana-2597	114	24	level	level	NOUN
cana-2597	114	25	fusion	fusion	NOUN
cana-2597	114	26	approach	approach	NOUN
cana-2597	114	27	.	.	PUNCT
cana-2597	115	1	in	in	ADP
cana-2597	115	2	this	this	DET
cana-2597	115	3	approach	approach	NOUN
cana-2597	115	4	,	,	PUNCT
cana-2597	115	5	the	the	DET
cana-2597	115	6	features	feature	NOUN
cana-2597	115	7	extracted	extract	VERB
cana-2597	115	8	from	from	ADP
cana-2597	115	9	both	both	CCONJ
cana-2597	115	10	the	the	DET
cana-2597	115	11	textual	textual	ADJ
cana-2597	115	12	data	datum	NOUN
cana-2597	115	13	(	(	PUNCT
cana-2597	115	14	using	use	VERB
cana-2597	115	15	nlp	nlp	NOUN
cana-2597	115	16	models	model	NOUN
cana-2597	115	17	)	)	PUNCT
cana-2597	115	18	and	and	CCONJ
cana-2597	115	19	the	the	DET
cana-2597	115	20	visual	visual	ADJ
cana-2597	115	21	data	datum	NOUN
cana-2597	115	22	(	(	PUNCT
cana-2597	115	23	using	use	VERB
cana-2597	115	24	cnns	cnn	NOUN
cana-2597	115	25	)	)	PUNCT
cana-2597	115	26	are	be	AUX
cana-2597	115	27	combined	combine	VERB
cana-2597	115	28	before	before	ADP
cana-2597	115	29	being	be	AUX
cana-2597	115	30	passed	pass	VERB
cana-2597	115	31	to	to	ADP
cana-2597	115	32	the	the	DET
cana-2597	115	33	classifier	classifier	NOUN
cana-2597	115	34	.	.	PUNCT
cana-2597	116	1	this	this	DET
cana-2597	116	2	fusion	fusion	NOUN
cana-2597	116	3	ensures	ensure	VERB
cana-2597	116	4	that	that	SCONJ
cana-2597	116	5	the	the	DET
cana-2597	116	6	model	model	NOUN
cana-2597	116	7	leverages	leverage	VERB
cana-2597	116	8	the	the	DET
cana-2597	116	9	complementary	complementary	ADJ
cana-2597	116	10	strengths	strength	NOUN
cana-2597	116	11	of	of	ADP
cana-2597	116	12	both	both	CCONJ
cana-2597	116	13	textual	textual	ADJ
cana-2597	116	14	and	and	CCONJ
cana-2597	116	15	visual	visual	ADJ
cana-2597	116	16	data	datum	NOUN
cana-2597	116	17	,	,	PUNCT
cana-2597	116	18	leading	lead	VERB
cana-2597	116	19	to	to	ADP
cana-2597	116	20	a	a	DET
cana-2597	116	21	more	more	ADV
cana-2597	116	22	accurate	accurate	ADJ
cana-2597	116	23	classification	classification	NOUN
cana-2597	116	24	of	of	ADP
cana-2597	116	25	disaster	disaster	NOUN
cana-2597	116	26	-	-	PUNCT
cana-2597	116	27	related	relate	VERB
cana-2597	116	28	tweets	tweet	NOUN
cana-2597	116	29	.	.	PUNCT
cana-2597	117	1	studies	study	NOUN
cana-2597	117	2	have	have	AUX
cana-2597	117	3	shown	show	VERB
cana-2597	117	4	that	that	SCONJ
cana-2597	117	5	this	this	DET
cana-2597	117	6	multimodal	multimodal	ADJ
cana-2597	117	7	fusion	fusion	NOUN
cana-2597	117	8	outperforms	outperform	NOUN
cana-2597	117	9	approaches	approach	NOUN
cana-2597	117	10	that	that	PRON
cana-2597	117	11	rely	rely	VERB
cana-2597	117	12	on	on	ADP
cana-2597	117	13	a	a	DET
cana-2597	117	14	single	single	ADJ
cana-2597	117	15	modality	modality	NOUN
cana-2597	117	16	,	,	PUNCT
cana-2597	117	17	such	such	ADJ
cana-2597	117	18	as	as	ADP
cana-2597	117	19	only	only	ADJ
cana-2597	117	20	text	text	NOUN
cana-2597	117	21	or	or	CCONJ
cana-2597	117	22	only	only	ADJ
cana-2597	117	23	images	image	NOUN
cana-2597	117	24	,	,	PUNCT
cana-2597	117	25	particularly	particularly	ADV
cana-2597	117	26	in	in	ADP
cana-2597	117	27	complex	complex	ADJ
cana-2597	117	28	scenarios	scenario	NOUN
cana-2597	117	29	like	like	ADP
cana-2597	117	30	disaster	disaster	NOUN
cana-2597	117	31	detection	detection	NOUN
cana-2597	117	32	[	[	X
cana-2597	117	33	9	9	NUM
cana-2597	117	34	]	]	PUNCT
cana-2597	117	35	.	.	PUNCT
cana-2597	118	1	the	the	DET
cana-2597	118	2	fusion	fusion	NOUN
cana-2597	118	3	model	model	NOUN
cana-2597	118	4	is	be	AUX
cana-2597	118	5	vital	vital	ADJ
cana-2597	118	6	for	for	ADP
cana-2597	118	7	handling	handle	VERB
cana-2597	118	8	tweets	tweet	NOUN
cana-2597	118	9	with	with	ADP
cana-2597	118	10	mixed	mixed	ADJ
cana-2597	118	11	or	or	CCONJ
cana-2597	118	12	incomplete	incomplete	ADJ
cana-2597	118	13	data	datum	NOUN
cana-2597	118	14	,	,	PUNCT
cana-2597	118	15	where	where	SCONJ
cana-2597	118	16	visual	visual	ADJ
cana-2597	118	17	information	information	NOUN
cana-2597	118	18	or	or	CCONJ
cana-2597	118	19	text	text	NOUN
cana-2597	118	20	alone	alone	ADV
cana-2597	118	21	may	may	AUX
cana-2597	118	22	not	not	PART
cana-2597	118	23	be	be	AUX
cana-2597	118	24	sufficient	sufficient	ADJ
cana-2597	118	25	to	to	PART
cana-2597	118	26	classify	classify	VERB
cana-2597	118	27	the	the	DET
cana-2597	118	28	tweet	tweet	NOUN
cana-2597	118	29	accurately	accurately	ADV
cana-2597	118	30	.	.	PUNCT
cana-2597	119	1	3.6	3.6	NUM
cana-2597	119	2	.	.	PUNCT
cana-2597	120	1	filtering	filter	VERB
cana-2597	120	2	irrelevant	irrelevant	ADJ
cana-2597	120	3	tweets	tweet	NOUN
cana-2597	120	4	with	with	ADP
cana-2597	120	5	gemini	gemini	PROPN
cana-2597	120	6	llm	llm	PROPN
cana-2597	120	7	to	to	PART
cana-2597	120	8	handle	handle	VERB
cana-2597	120	9	the	the	DET
cana-2597	120	10	vast	vast	ADJ
cana-2597	120	11	influx	influx	NOUN
cana-2597	120	12	of	of	ADP
cana-2597	120	13	tweets	tweet	NOUN
cana-2597	120	14	,	,	PUNCT
cana-2597	120	15	the	the	DET
cana-2597	120	16	system	system	NOUN
cana-2597	120	17	employs	employ	VERB
cana-2597	120	18	gemini	gemini	PROPN
cana-2597	120	19	llm	llm	PROPN
cana-2597	120	20	,	,	PUNCT
cana-2597	120	21	a	a	DET
cana-2597	120	22	large	large	ADJ
cana-2597	120	23	language	language	NOUN
cana-2597	120	24	model	model	NOUN
cana-2597	120	25	(	(	PUNCT
cana-2597	120	26	llm	llm	NOUN
cana-2597	120	27	)	)	PUNCT
cana-2597	120	28	designed	design	VERB
cana-2597	120	29	specifically	specifically	ADV
cana-2597	120	30	for	for	ADP
cana-2597	120	31	filtering	filter	VERB
cana-2597	120	32	out	out	ADP
cana-2597	120	33	irrelevant	irrelevant	ADJ
cana-2597	120	34	content	content	NOUN
cana-2597	120	35	.	.	PUNCT
cana-2597	121	1	tweets	tweet	NOUN
cana-2597	121	2	related	relate	VERB
cana-2597	121	3	to	to	ADP
cana-2597	121	4	disasters	disaster	NOUN
cana-2597	121	5	can	can	AUX
cana-2597	121	6	often	often	ADV
cana-2597	121	7	be	be	AUX
cana-2597	121	8	mixed	mix	VERB
cana-2597	121	9	with	with	ADP
cana-2597	121	10	spam	spam	NOUN
cana-2597	121	11	,	,	PUNCT
cana-2597	121	12	advertisements	advertisement	NOUN
cana-2597	121	13	,	,	PUNCT
cana-2597	121	14	or	or	CCONJ
cana-2597	121	15	other	other	ADJ
cana-2597	121	16	noise	noise	NOUN
cana-2597	121	17	,	,	PUNCT
cana-2597	121	18	which	which	PRON
cana-2597	121	19	may	may	AUX
cana-2597	121	20	not	not	PART
cana-2597	121	21	contribute	contribute	VERB
cana-2597	121	22	meaningfully	meaningfully	ADV
cana-2597	121	23	to	to	ADP
cana-2597	121	24	disaster	disaster	NOUN
cana-2597	121	25	management	management	NOUN
cana-2597	121	26	efforts	effort	NOUN
cana-2597	121	27	.	.	PUNCT
cana-2597	122	1	gemini	gemini	PROPN
cana-2597	122	2	llm	llm	PROPN
cana-2597	122	3	analyzes	analyze	VERB
cana-2597	122	4	the	the	DET
cana-2597	122	5	context	context	NOUN
cana-2597	122	6	of	of	ADP
cana-2597	122	7	tweets	tweet	NOUN
cana-2597	122	8	and	and	CCONJ
cana-2597	122	9	effectively	effectively	ADV
cana-2597	122	10	differentiates	differentiate	NOUN
cana-2597	122	11	between	between	ADP
cana-2597	122	12	disaster	disaster	NOUN
cana-2597	122	13	-	-	PUNCT
cana-2597	122	14	relevant	relevant	ADJ
cana-2597	122	15	information	information	NOUN
cana-2597	122	16	and	and	CCONJ
cana-2597	122	17	extraneous	extraneous	ADJ
cana-2597	122	18	content	content	NOUN
cana-2597	122	19	.	.	PUNCT
cana-2597	123	1	the	the	DET
cana-2597	123	2	model	model	NOUN
cana-2597	123	3	’s	’s	PART
cana-2597	123	4	ability	ability	NOUN
cana-2597	123	5	to	to	PART
cana-2597	123	6	understand	understand	VERB
cana-2597	123	7	and	and	CCONJ
cana-2597	123	8	communications	communication	NOUN
cana-2597	123	9	on	on	ADP
cana-2597	123	10	applied	apply	VERB
cana-2597	123	11	nonlinear	nonlinear	ADJ
cana-2597	123	12	analysis	analysis	NOUN
cana-2597	123	13	issn	issn	NOUN
cana-2597	123	14	:	:	PUNCT
cana-2597	123	15	1074	1074	NUM
cana-2597	123	16	-	-	PUNCT
cana-2597	123	17	133x	133x	NUM
cana-2597	123	18	vol	vol	NOUN
cana-2597	123	19	32	32	NUM
cana-2597	123	20	no	no	NOUN
cana-2597	123	21	.	.	PUNCT
cana-2597	124	1	3s	3s	NUM
cana-2597	124	2	(	(	PUNCT
cana-2597	124	3	2025	2025	NUM
cana-2597	124	4	)	)	PUNCT
cana-2597	124	5	205	205	NUM
cana-2597	124	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2597	124	7	assess	assess	VERB
cana-2597	124	8	tweet	tweet	NOUN
cana-2597	124	9	context	context	NOUN
cana-2597	124	10	in	in	ADP
cana-2597	124	11	real	real	ADJ
cana-2597	124	12	-	-	PUNCT
cana-2597	124	13	time	time	NOUN
cana-2597	124	14	significantly	significantly	ADV
cana-2597	124	15	reduces	reduce	VERB
cana-2597	124	16	the	the	DET
cana-2597	124	17	clutter	clutter	NOUN
cana-2597	124	18	in	in	ADP
cana-2597	124	19	the	the	DET
cana-2597	124	20	dataset	dataset	NOUN
cana-2597	124	21	,	,	PUNCT
cana-2597	124	22	ensuring	ensure	VERB
cana-2597	124	23	that	that	SCONJ
cana-2597	124	24	only	only	ADJ
cana-2597	124	25	meaningful	meaningful	ADJ
cana-2597	124	26	tweets	tweet	NOUN
cana-2597	124	27	are	be	AUX
cana-2597	124	28	processed	process	VERB
cana-2597	124	29	and	and	CCONJ
cana-2597	124	30	analyzed	analyze	VERB
cana-2597	124	31	[	[	X
cana-2597	124	32	15	15	NUM
cana-2597	124	33	]	]	PUNCT
cana-2597	124	34	.	.	PUNCT
cana-2597	125	1	this	this	DET
cana-2597	125	2	filtering	filter	VERB
cana-2597	125	3	mechanism	mechanism	NOUN
cana-2597	125	4	is	be	AUX
cana-2597	125	5	crucial	crucial	ADJ
cana-2597	125	6	in	in	ADP
cana-2597	125	7	maintaining	maintain	VERB
cana-2597	125	8	the	the	DET
cana-2597	125	9	system	system	NOUN
cana-2597	125	10	’s	’s	PART
cana-2597	125	11	overall	overall	ADJ
cana-2597	125	12	efficiency	efficiency	NOUN
cana-2597	125	13	and	and	CCONJ
cana-2597	125	14	focus	focus	VERB
cana-2597	125	15	during	during	ADP
cana-2597	125	16	disaster	disaster	NOUN
cana-2597	125	17	events	event	NOUN
cana-2597	125	18	when	when	SCONJ
cana-2597	125	19	accurate	accurate	ADJ
cana-2597	125	20	and	and	CCONJ
cana-2597	125	21	relevant	relevant	ADJ
cana-2597	125	22	information	information	NOUN
cana-2597	125	23	is	be	AUX
cana-2597	125	24	essential	essential	ADJ
cana-2597	125	25	.	.	PUNCT
cana-2597	126	1	3.7	3.7	NUM
cana-2597	126	2	.	.	PUNCT
cana-2597	126	3	data	datum	NOUN
cana-2597	126	4	storage	storage	NOUN
cana-2597	126	5	and	and	CCONJ
cana-2597	126	6	search	search	NOUN
cana-2597	126	7	after	after	ADP
cana-2597	126	8	processing	processing	NOUN
cana-2597	126	9	,	,	PUNCT
cana-2597	126	10	the	the	DET
cana-2597	126	11	system	system	NOUN
cana-2597	126	12	stores	store	NOUN
cana-2597	126	13	tweets	tweet	NOUN
cana-2597	126	14	in	in	ADP
cana-2597	126	15	mongodb	mongodb	NOUN
cana-2597	126	16	and	and	CCONJ
cana-2597	126	17	utilizes	utilizes	ADJ
cana-2597	126	18	elasticsearch	elasticsearch	NOUN
cana-2597	126	19	for	for	ADP
cana-2597	126	20	indexing	indexing	NOUN
cana-2597	126	21	and	and	CCONJ
cana-2597	126	22	searching	searching	NOUN
cana-2597	126	23	.	.	PUNCT
cana-2597	127	1	mongodb	mongodb	NOUN
cana-2597	127	2	is	be	AUX
cana-2597	127	3	used	use	VERB
cana-2597	127	4	for	for	ADP
cana-2597	127	5	storing	store	VERB
cana-2597	127	6	unstructured	unstructured	ADJ
cana-2597	127	7	data	datum	NOUN
cana-2597	127	8	,	,	PUNCT
cana-2597	127	9	such	such	ADJ
cana-2597	127	10	as	as	ADP
cana-2597	127	11	tweet	tweet	NOUN
cana-2597	127	12	content	content	NOUN
cana-2597	127	13	,	,	PUNCT
cana-2597	127	14	images	image	NOUN
cana-2597	127	15	,	,	PUNCT
cana-2597	127	16	and	and	CCONJ
cana-2597	127	17	associated	associated	ADJ
cana-2597	127	18	metadata	metadata	NOUN
cana-2597	127	19	,	,	PUNCT
cana-2597	127	20	which	which	PRON
cana-2597	127	21	makes	make	VERB
cana-2597	127	22	it	it	PRON
cana-2597	127	23	well	well	ADV
cana-2597	127	24	-	-	PUNCT
cana-2597	127	25	suited	suit	VERB
cana-2597	127	26	for	for	ADP
cana-2597	127	27	handling	handle	VERB
cana-2597	127	28	diverse	diverse	ADJ
cana-2597	127	29	tweet	tweet	NOUN
cana-2597	127	30	formats	format	NOUN
cana-2597	127	31	.	.	PUNCT
cana-2597	128	1	it	it	PRON
cana-2597	128	2	ensures	ensure	VERB
cana-2597	128	3	the	the	DET
cana-2597	128	4	scalability	scalability	NOUN
cana-2597	128	5	of	of	ADP
cana-2597	128	6	the	the	DET
cana-2597	128	7	storage	storage	NOUN
cana-2597	128	8	,	,	PUNCT
cana-2597	128	9	as	as	SCONJ
cana-2597	128	10	the	the	DET
cana-2597	128	11	volume	volume	NOUN
cana-2597	128	12	of	of	ADP
cana-2597	128	13	tweets	tweet	NOUN
cana-2597	128	14	can	can	AUX
cana-2597	128	15	increase	increase	VERB
cana-2597	128	16	dramatically	dramatically	ADV
cana-2597	128	17	during	during	ADP
cana-2597	128	18	disaster	disaster	NOUN
cana-2597	128	19	situations	situation	NOUN
cana-2597	128	20	.	.	PUNCT
cana-2597	129	1	in	in	ADP
cana-2597	129	2	parallel	parallel	NOUN
cana-2597	129	3	,	,	PUNCT
cana-2597	129	4	elasticsearch	elasticsearch	NOUN
cana-2597	129	5	provides	provide	VERB
cana-2597	129	6	the	the	DET
cana-2597	129	7	capability	capability	NOUN
cana-2597	129	8	for	for	ADP
cana-2597	129	9	fast	fast	ADJ
cana-2597	129	10	,	,	PUNCT
cana-2597	129	11	full	full	ADJ
cana-2597	129	12	-	-	PUNCT
cana-2597	129	13	text	text	NOUN
cana-2597	129	14	search	search	NOUN
cana-2597	129	15	,	,	PUNCT
cana-2597	129	16	enabling	enable	VERB
cana-2597	129	17	real	real	ADJ
cana-2597	129	18	-	-	PUNCT
cana-2597	129	19	time	time	NOUN
cana-2597	129	20	querying	querying	NOUN
cana-2597	129	21	and	and	CCONJ
cana-2597	129	22	retrieval	retrieval	NOUN
cana-2597	129	23	of	of	ADP
cana-2597	129	24	tweets	tweet	NOUN
cana-2597	129	25	based	base	VERB
cana-2597	129	26	on	on	ADP
cana-2597	129	27	disaster	disaster	NOUN
cana-2597	129	28	-	-	PUNCT
cana-2597	129	29	related	relate	VERB
cana-2597	129	30	keywords	keyword	NOUN
cana-2597	129	31	,	,	PUNCT
cana-2597	129	32	geographic	geographic	ADJ
cana-2597	129	33	locations	location	NOUN
cana-2597	129	34	,	,	PUNCT
cana-2597	129	35	or	or	CCONJ
cana-2597	129	36	other	other	ADJ
cana-2597	129	37	metadata	metadata	NOUN
cana-2597	129	38	.	.	PUNCT
cana-2597	130	1	this	this	DET
cana-2597	130	2	combination	combination	NOUN
cana-2597	130	3	allows	allow	VERB
cana-2597	130	4	users	user	NOUN
cana-2597	130	5	to	to	PART
cana-2597	130	6	search	search	VERB
cana-2597	130	7	the	the	DET
cana-2597	130	8	vast	vast	ADJ
cana-2597	130	9	database	database	NOUN
cana-2597	130	10	quickly	quickly	ADV
cana-2597	130	11	and	and	CCONJ
cana-2597	130	12	efficiently	efficiently	ADV
cana-2597	130	13	.	.	PUNCT
cana-2597	131	1	to	to	PART
cana-2597	131	2	visualize	visualize	VERB
cana-2597	131	3	trends	trend	NOUN
cana-2597	131	4	and	and	CCONJ
cana-2597	131	5	provide	provide	VERB
cana-2597	131	6	insights	insight	NOUN
cana-2597	131	7	into	into	ADP
cana-2597	131	8	tweet	tweet	NOUN
cana-2597	131	9	patterns	pattern	NOUN
cana-2597	131	10	,	,	PUNCT
cana-2597	131	11	the	the	DET
cana-2597	131	12	system	system	NOUN
cana-2597	131	13	also	also	ADV
cana-2597	131	14	integrates	integrate	VERB
cana-2597	131	15	kibana	kibana	PROPN
cana-2597	131	16	,	,	PUNCT
cana-2597	131	17	a	a	DET
cana-2597	131	18	data	data	NOUN
cana-2597	131	19	visualization	visualization	NOUN
cana-2597	131	20	tool	tool	NOUN
cana-2597	131	21	that	that	PRON
cana-2597	131	22	works	work	VERB
cana-2597	131	23	seamlessly	seamlessly	ADV
cana-2597	131	24	with	with	ADP
cana-2597	131	25	elasticsearch	elasticsearch	NOUN
cana-2597	131	26	.	.	PUNCT
cana-2597	132	1	kibana	kibana	PROPN
cana-2597	132	2	allows	allow	VERB
cana-2597	132	3	real	real	ADJ
cana-2597	132	4	-	-	PUNCT
cana-2597	132	5	time	time	NOUN
cana-2597	132	6	monitoring	monitoring	NOUN
cana-2597	132	7	of	of	ADP
cana-2597	132	8	tweet	tweet	NOUN
cana-2597	132	9	trends	trend	NOUN
cana-2597	132	10	,	,	PUNCT
cana-2597	132	11	helping	help	VERB
cana-2597	132	12	disaster	disaster	NOUN
cana-2597	132	13	response	response	NOUN
cana-2597	132	14	teams	team	NOUN
cana-2597	132	15	make	make	VERB
cana-2597	132	16	data	datum	NOUN
cana-2597	132	17	-	-	PUNCT
cana-2597	132	18	driven	drive	VERB
cana-2597	132	19	decisions	decision	NOUN
cana-2597	132	20	based	base	VERB
cana-2597	132	21	on	on	ADP
cana-2597	132	22	the	the	DET
cana-2597	132	23	most	most	ADJ
cana-2597	132	24	up	up	ADP
cana-2597	132	25	-	-	PUNCT
cana-2597	132	26	to	to	ADP
cana-2597	132	27	-	-	PUNCT
cana-2597	132	28	date	date	NOUN
cana-2597	132	29	information	information	NOUN
cana-2597	133	1	[	[	X
cana-2597	133	2	16	16	NUM
cana-2597	133	3	]	]	PUNCT
cana-2597	133	4	.	.	PUNCT
cana-2597	134	1	3.8	3.8	NUM
cana-2597	134	2	.	.	PUNCT
cana-2597	134	3	performance	performance	NOUN
cana-2597	134	4	evaluation	evaluation	NOUN
cana-2597	134	5	the	the	DET
cana-2597	134	6	system	system	NOUN
cana-2597	134	7	's	's	PART
cana-2597	134	8	performance	performance	NOUN
cana-2597	134	9	is	be	AUX
cana-2597	134	10	evaluated	evaluate	VERB
cana-2597	134	11	using	use	VERB
cana-2597	134	12	several	several	ADJ
cana-2597	134	13	key	key	ADJ
cana-2597	134	14	metrics	metric	NOUN
cana-2597	134	15	to	to	PART
cana-2597	134	16	ensure	ensure	VERB
cana-2597	134	17	the	the	DET
cana-2597	134	18	effectiveness	effectiveness	NOUN
cana-2597	134	19	of	of	ADP
cana-2597	134	20	tweet	tweet	NOUN
cana-2597	134	21	classification	classification	NOUN
cana-2597	134	22	and	and	CCONJ
cana-2597	134	23	processing	processing	NOUN
cana-2597	134	24	.	.	PUNCT
cana-2597	135	1	accuracy	accuracy	NOUN
cana-2597	135	2	is	be	AUX
cana-2597	135	3	the	the	DET
cana-2597	135	4	primary	primary	ADJ
cana-2597	135	5	metric	metric	NOUN
cana-2597	135	6	used	use	VERB
cana-2597	135	7	to	to	PART
cana-2597	135	8	assess	assess	VERB
cana-2597	135	9	the	the	DET
cana-2597	135	10	overall	overall	ADJ
cana-2597	135	11	proportion	proportion	NOUN
cana-2597	135	12	of	of	ADP
cana-2597	135	13	correctly	correctly	ADV
cana-2597	135	14	classified	classified	ADJ
cana-2597	135	15	tweets	tweet	NOUN
cana-2597	135	16	.	.	PUNCT
cana-2597	136	1	however	however	ADV
cana-2597	136	2	,	,	PUNCT
cana-2597	136	3	given	give	VERB
cana-2597	136	4	the	the	DET
cana-2597	136	5	imbalance	imbalance	NOUN
cana-2597	136	6	in	in	ADP
cana-2597	136	7	datasets	dataset	NOUN
cana-2597	136	8	—	—	PUNCT
cana-2597	136	9	where	where	SCONJ
cana-2597	136	10	some	some	DET
cana-2597	136	11	disaster	disaster	NOUN
cana-2597	136	12	types	type	NOUN
cana-2597	136	13	may	may	AUX
cana-2597	136	14	be	be	AUX
cana-2597	136	15	more	more	ADV
cana-2597	136	16	frequent	frequent	ADJ
cana-2597	136	17	than	than	ADP
cana-2597	136	18	others	other	NOUN
cana-2597	136	19	—	—	PUNCT
cana-2597	136	20	additional	additional	ADJ
cana-2597	136	21	metrics	metric	NOUN
cana-2597	136	22	like	like	ADP
cana-2597	136	23	precision	precision	NOUN
cana-2597	136	24	,	,	PUNCT
cana-2597	136	25	recall	recall	NOUN
cana-2597	136	26	,	,	PUNCT
cana-2597	136	27	and	and	CCONJ
cana-2597	136	28	the	the	DET
cana-2597	136	29	f1	f1	NOUN
cana-2597	136	30	-	-	PUNCT
cana-2597	136	31	score	score	NOUN
cana-2597	136	32	provide	provide	VERB
cana-2597	136	33	a	a	DET
cana-2597	136	34	more	more	ADV
cana-2597	136	35	granular	granular	ADJ
cana-2597	136	36	view	view	NOUN
cana-2597	136	37	of	of	ADP
cana-2597	136	38	performance	performance	NOUN
cana-2597	136	39	.	.	PUNCT
cana-2597	137	1	precision	precision	NOUN
cana-2597	137	2	measures	measure	VERB
cana-2597	137	3	the	the	DET
cana-2597	137	4	proportion	proportion	NOUN
cana-2597	137	5	of	of	ADP
cana-2597	137	6	correctly	correctly	ADV
cana-2597	137	7	identified	identify	VERB
cana-2597	137	8	disaster	disaster	NOUN
cana-2597	137	9	tweets	tweet	NOUN
cana-2597	137	10	out	out	ADP
cana-2597	137	11	of	of	ADP
cana-2597	137	12	all	all	DET
cana-2597	137	13	tweets	tweet	NOUN
cana-2597	137	14	classified	classify	VERB
cana-2597	137	15	as	as	ADP
cana-2597	137	16	disaster	disaster	NOUN
cana-2597	137	17	-	-	PUNCT
cana-2597	137	18	related	relate	VERB
cana-2597	137	19	,	,	PUNCT
cana-2597	137	20	while	while	SCONJ
cana-2597	137	21	recall	recall	NOUN
cana-2597	137	22	evaluates	evaluate	VERB
cana-2597	137	23	the	the	DET
cana-2597	137	24	system	system	NOUN
cana-2597	137	25	's	's	PART
cana-2597	137	26	ability	ability	NOUN
cana-2597	137	27	to	to	PART
cana-2597	137	28	identify	identify	VERB
cana-2597	137	29	all	all	DET
cana-2597	137	30	relevant	relevant	ADJ
cana-2597	137	31	disaster	disaster	NOUN
cana-2597	137	32	tweets	tweet	NOUN
cana-2597	137	33	.	.	PUNCT
cana-2597	138	1	the	the	DET
cana-2597	138	2	f1	f1	NOUN
cana-2597	138	3	-	-	PUNCT
cana-2597	138	4	score	score	NOUN
cana-2597	138	5	,	,	PUNCT
cana-2597	138	6	which	which	PRON
cana-2597	138	7	is	be	AUX
cana-2597	138	8	the	the	DET
cana-2597	138	9	harmonic	harmonic	ADJ
cana-2597	138	10	mean	mean	NOUN
cana-2597	138	11	of	of	ADP
cana-2597	138	12	precision	precision	NOUN
cana-2597	138	13	and	and	CCONJ
cana-2597	138	14	recall	recall	NOUN
cana-2597	138	15	,	,	PUNCT
cana-2597	138	16	provides	provide	VERB
cana-2597	138	17	a	a	DET
cana-2597	138	18	balanced	balanced	ADJ
cana-2597	138	19	evaluation	evaluation	NOUN
cana-2597	138	20	,	,	PUNCT
cana-2597	138	21	particularly	particularly	ADV
cana-2597	138	22	in	in	ADP
cana-2597	138	23	cases	case	NOUN
cana-2597	138	24	of	of	ADP
cana-2597	138	25	imbalanced	imbalanced	ADJ
cana-2597	138	26	datasets	dataset	NOUN
cana-2597	138	27	[	[	X
cana-2597	138	28	17	17	NUM
cana-2597	138	29	]	]	PUNCT
cana-2597	138	30	.	.	PUNCT
cana-2597	139	1	finally	finally	ADV
cana-2597	139	2	,	,	PUNCT
cana-2597	139	3	the	the	DET
cana-2597	139	4	system	system	NOUN
cana-2597	139	5	also	also	ADV
cana-2597	139	6	monitors	monitor	VERB
cana-2597	139	7	latency	latency	NOUN
cana-2597	139	8	,	,	PUNCT
cana-2597	139	9	measuring	measure	VERB
cana-2597	139	10	the	the	DET
cana-2597	139	11	time	time	NOUN
cana-2597	139	12	between	between	ADP
cana-2597	139	13	tweet	tweet	NOUN
cana-2597	139	14	ingestion	ingestion	NOUN
cana-2597	139	15	and	and	CCONJ
cana-2597	139	16	classification	classification	NOUN
cana-2597	139	17	.	.	PUNCT
cana-2597	140	1	this	this	PRON
cana-2597	140	2	ensures	ensure	VERB
cana-2597	140	3	that	that	SCONJ
cana-2597	140	4	the	the	DET
cana-2597	140	5	system	system	NOUN
cana-2597	140	6	can	can	AUX
cana-2597	140	7	handle	handle	VERB
cana-2597	140	8	real	real	ADJ
cana-2597	140	9	-	-	PUNCT
cana-2597	140	10	time	time	NOUN
cana-2597	140	11	tweet	tweet	NOUN
cana-2597	140	12	processing	processing	NOUN
cana-2597	140	13	,	,	PUNCT
cana-2597	140	14	an	an	DET
cana-2597	140	15	essential	essential	ADJ
cana-2597	140	16	feature	feature	NOUN
cana-2597	140	17	during	during	ADP
cana-2597	140	18	fast	fast	ADV
cana-2597	140	19	-	-	PUNCT
cana-2597	140	20	paced	pace	VERB
cana-2597	140	21	disaster	disaster	NOUN
cana-2597	140	22	events	event	NOUN
cana-2597	140	23	.	.	PUNCT
cana-2597	141	1	these	these	DET
cana-2597	141	2	combined	combined	ADJ
cana-2597	141	3	efforts	effort	NOUN
cana-2597	141	4	ensure	ensure	VERB
cana-2597	141	5	that	that	SCONJ
cana-2597	141	6	the	the	DET
cana-2597	141	7	system	system	NOUN
cana-2597	141	8	is	be	AUX
cana-2597	141	9	both	both	CCONJ
cana-2597	141	10	accurate	accurate	ADJ
cana-2597	141	11	and	and	CCONJ
cana-2597	141	12	responsive	responsive	ADJ
cana-2597	141	13	,	,	PUNCT
cana-2597	141	14	capable	capable	ADJ
cana-2597	141	15	of	of	ADP
cana-2597	141	16	processing	process	VERB
cana-2597	141	17	large	large	ADJ
cana-2597	141	18	amounts	amount	NOUN
cana-2597	141	19	of	of	ADP
cana-2597	141	20	tweet	tweet	NOUN
cana-2597	141	21	data	datum	NOUN
cana-2597	141	22	efficiently	efficiently	ADV
cana-2597	141	23	during	during	ADP
cana-2597	141	24	critical	critical	ADJ
cana-2597	141	25	times	time	NOUN
cana-2597	141	26	,	,	PUNCT
cana-2597	141	27	and	and	CCONJ
cana-2597	141	28	providing	provide	VERB
cana-2597	141	29	actionable	actionable	ADJ
cana-2597	141	30	insights	insight	NOUN
cana-2597	141	31	to	to	ADP
cana-2597	141	32	disaster	disaster	NOUN
cana-2597	141	33	response	response	NOUN
cana-2597	141	34	teams	team	NOUN
cana-2597	141	35	.	.	PUNCT
cana-2597	142	1	iv	iv	X
cana-2597	142	2	.	.	PUNCT
cana-2597	142	3	implementation	implementation	NOUN
cana-2597	142	4	the	the	DET
cana-2597	142	5	proposed	propose	VERB
cana-2597	142	6	solution	solution	NOUN
cana-2597	142	7	integrates	integrate	VERB
cana-2597	142	8	various	various	ADJ
cana-2597	142	9	components	component	NOUN
cana-2597	142	10	for	for	ADP
cana-2597	142	11	real	real	ADJ
cana-2597	142	12	-	-	PUNCT
cana-2597	142	13	time	time	NOUN
cana-2597	142	14	disaster	disaster	NOUN
cana-2597	142	15	tweet	tweet	NOUN
cana-2597	142	16	collection	collection	NOUN
cana-2597	142	17	,	,	PUNCT
cana-2597	142	18	processing	processing	NOUN
cana-2597	142	19	,	,	PUNCT
cana-2597	142	20	and	and	CCONJ
cana-2597	142	21	classification	classification	NOUN
cana-2597	142	22	.	.	PUNCT
cana-2597	143	1	the	the	DET
cana-2597	143	2	system	system	NOUN
cana-2597	143	3	is	be	AUX
cana-2597	143	4	built	build	VERB
cana-2597	143	5	using	use	VERB
cana-2597	143	6	web	web	NOUN
cana-2597	143	7	scraping	scraping	NOUN
cana-2597	143	8	tools	tool	NOUN
cana-2597	143	9	,	,	PUNCT
cana-2597	143	10	apache	apache	PROPN
cana-2597	143	11	kafka	kafka	PROPN
cana-2597	143	12	,	,	PUNCT
cana-2597	143	13	apache	apache	NOUN
cana-2597	143	14	spark	spark	NOUN
cana-2597	143	15	,	,	PUNCT
cana-2597	143	16	mongodb	mongodb	NOUN
cana-2597	143	17	,	,	PUNCT
cana-2597	143	18	elasticsearch	elasticsearch	NOUN
cana-2597	143	19	,	,	PUNCT
cana-2597	143	20	and	and	CCONJ
cana-2597	143	21	a	a	DET
cana-2597	143	22	multimodal	multimodal	ADJ
cana-2597	143	23	classification	classification	NOUN
cana-2597	143	24	model	model	NOUN
cana-2597	143	25	that	that	PRON
cana-2597	143	26	leverages	leverage	VERB
cana-2597	143	27	bert	bert	NOUN
cana-2597	143	28	,	,	PUNCT
cana-2597	143	29	efficientnetb3	efficientnetb3	PROPN
cana-2597	143	30	,	,	PUNCT
cana-2597	143	31	and	and	CCONJ
cana-2597	143	32	other	other	ADJ
cana-2597	143	33	state	state	NOUN
cana-2597	143	34	-	-	PUNCT
cana-2597	143	35	of	of	ADP
cana-2597	143	36	-	-	PUNCT
cana-2597	143	37	the	the	DET
cana-2597	143	38	-	-	PUNCT
cana-2597	143	39	art	art	NOUN
cana-2597	143	40	machine	machine	NOUN
cana-2597	143	41	learning	learning	NOUN
cana-2597	143	42	tools	tool	NOUN
cana-2597	143	43	.	.	PUNCT
cana-2597	144	1	4.1	4.1	NUM
cana-2597	144	2	dataset	dataset	VERB
cana-2597	144	3	the	the	DET
cana-2597	144	4	dataset	dataset	NOUN
cana-2597	144	5	used	use	VERB
cana-2597	144	6	in	in	ADP
cana-2597	144	7	this	this	DET
cana-2597	144	8	research	research	NOUN
cana-2597	144	9	consists	consist	VERB
cana-2597	144	10	of	of	ADP
cana-2597	144	11	social	social	ADJ
cana-2597	144	12	media	medium	NOUN
cana-2597	144	13	posts	post	NOUN
cana-2597	144	14	with	with	ADP
cana-2597	144	15	associated	associated	ADJ
cana-2597	144	16	text	text	NOUN
cana-2597	144	17	and	and	CCONJ
cana-2597	144	18	image	image	NOUN
cana-2597	144	19	files	file	NOUN
cana-2597	144	20	.	.	PUNCT
cana-2597	145	1	each	each	DET
cana-2597	145	2	post	post	NOUN
cana-2597	145	3	is	be	AUX
cana-2597	145	4	classified	classify	VERB
cana-2597	145	5	into	into	ADP
cana-2597	145	6	one	one	NUM
cana-2597	145	7	of	of	ADP
cana-2597	145	8	several	several	ADJ
cana-2597	145	9	categories	category	NOUN
cana-2597	145	10	,	,	PUNCT
cana-2597	145	11	including	include	VERB
cana-2597	145	12	damage	damage	NOUN
cana-2597	145	13	-	-	PUNCT
cana-2597	145	14	related	relate	VERB
cana-2597	145	15	events	event	NOUN
cana-2597	145	16	like	like	ADP
cana-2597	145	17	floods	flood	NOUN
cana-2597	145	18	or	or	CCONJ
cana-2597	145	19	fires	fire	NOUN
cana-2597	145	20	and	and	CCONJ
cana-2597	145	21	non	non	ADJ
cana-2597	145	22	-	-	ADJ
cana-2597	145	23	damage	damage	ADJ
cana-2597	145	24	events	event	NOUN
cana-2597	145	25	.	.	PUNCT
cana-2597	146	1	the	the	DET
cana-2597	146	2	dataset	dataset	NOUN
cana-2597	146	3	structure	structure	NOUN
cana-2597	146	4	looks	look	VERB
cana-2597	146	5	like	like	ADP
cana-2597	146	6	this	this	PRON
cana-2597	146	7	:	:	PUNCT
cana-2597	146	8	communications	communication	NOUN
cana-2597	146	9	on	on	ADP
cana-2597	146	10	applied	apply	VERB
cana-2597	146	11	nonlinear	nonlinear	ADJ
cana-2597	146	12	analysis	analysis	NOUN
cana-2597	146	13	issn	issn	NOUN
cana-2597	146	14	:	:	PUNCT
cana-2597	146	15	1074	1074	NUM
cana-2597	146	16	-	-	PUNCT
cana-2597	146	17	133x	133x	NUM
cana-2597	146	18	vol	vol	NOUN
cana-2597	146	19	32	32	NUM
cana-2597	146	20	no	no	NOUN
cana-2597	146	21	.	.	PUNCT
cana-2597	147	1	3s	3s	NUM
cana-2597	147	2	(	(	PUNCT
cana-2597	147	3	2025	2025	NUM
cana-2597	147	4	)	)	PUNCT
cana-2597	147	5	206	206	NUM
cana-2597	147	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2597	147	7	table	table	NOUN
cana-2597	147	8	1	1	NUM
cana-2597	147	9	.	.	PUNCT
cana-2597	147	10	dataset	dataset	NOUN
cana-2597	147	11	overview	overview	NOUN
cana-2597	147	12	the	the	DET
cana-2597	147	13	dataset	dataset	NOUN
cana-2597	147	14	consists	consist	NOUN
cana-2597	147	15	of	of	ADP
cana-2597	147	16	columns	column	NOUN
cana-2597	147	17	for	for	ADP
cana-2597	147	18	the	the	DET
cana-2597	147	19	filename	filename	NOUN
cana-2597	147	20	,	,	PUNCT
cana-2597	147	21	tweet	tweet	NOUN
cana-2597	147	22	(	(	PUNCT
cana-2597	147	23	text	text	NOUN
cana-2597	147	24	data	datum	NOUN
cana-2597	147	25	)	)	PUNCT
cana-2597	147	26	,	,	PUNCT
cana-2597	147	27	label	label	NOUN
cana-2597	147	28	(	(	PUNCT
cana-2597	147	29	classification	classification	NOUN
cana-2597	147	30	of	of	ADP
cana-2597	147	31	the	the	DET
cana-2597	147	32	event	event	NOUN
cana-2597	147	33	)	)	PUNCT
cana-2597	147	34	,	,	PUNCT
cana-2597	147	35	and	and	CCONJ
cana-2597	147	36	image	image	NOUN
cana-2597	147	37	(	(	PUNCT
cana-2597	147	38	associated	associate	VERB
cana-2597	147	39	visual	visual	ADJ
cana-2597	147	40	data	datum	NOUN
cana-2597	147	41	)	)	PUNCT
cana-2597	147	42	.	.	PUNCT
cana-2597	148	1	the	the	DET
cana-2597	148	2	label	label	NOUN
cana-2597	148	3	column	column	NOUN
cana-2597	148	4	indicates	indicate	VERB
cana-2597	148	5	whether	whether	SCONJ
cana-2597	148	6	the	the	DET
cana-2597	148	7	post	post	NOUN
cana-2597	148	8	refers	refer	VERB
cana-2597	148	9	to	to	ADP
cana-2597	148	10	flood	flood	NOUN
cana-2597	148	11	,	,	PUNCT
cana-2597	148	12	fire	fire	NOUN
cana-2597	148	13	,	,	PUNCT
cana-2597	148	14	or	or	CCONJ
cana-2597	148	15	non	non	ADJ
cana-2597	148	16	-	-	ADJ
cana-2597	148	17	damage	damage	ADJ
cana-2597	148	18	events	event	NOUN
cana-2597	148	19	.	.	PUNCT
cana-2597	149	1	this	this	DET
cana-2597	149	2	multimodal	multimodal	ADJ
cana-2597	149	3	dataset	dataset	NOUN
cana-2597	149	4	is	be	AUX
cana-2597	149	5	essential	essential	ADJ
cana-2597	149	6	for	for	ADP
cana-2597	149	7	training	training	NOUN
cana-2597	149	8	models	model	NOUN
cana-2597	149	9	that	that	PRON
cana-2597	149	10	can	can	AUX
cana-2597	149	11	handle	handle	VERB
cana-2597	149	12	both	both	CCONJ
cana-2597	149	13	textual	textual	ADJ
cana-2597	149	14	and	and	CCONJ
cana-2597	149	15	visual	visual	ADJ
cana-2597	149	16	data	datum	NOUN
cana-2597	149	17	to	to	PART
cana-2597	149	18	predict	predict	VERB
cana-2597	149	19	disaster	disaster	NOUN
cana-2597	149	20	-	-	PUNCT
cana-2597	149	21	related	relate	VERB
cana-2597	149	22	events	event	NOUN
cana-2597	149	23	.	.	PUNCT
cana-2597	150	1	4.2	4.2	NUM
cana-2597	150	2	.	.	PUNCT
cana-2597	151	1	data	datum	NOUN
cana-2597	151	2	collection	collection	NOUN
cana-2597	151	3	using	use	VERB
cana-2597	151	4	web	web	NOUN
cana-2597	151	5	scraping	scraping	NOUN
cana-2597	151	6	given	give	VERB
cana-2597	151	7	the	the	DET
cana-2597	151	8	limitations	limitation	NOUN
cana-2597	151	9	of	of	ADP
cana-2597	151	10	the	the	DET
cana-2597	151	11	twitter	twitter	NOUN
cana-2597	151	12	api	api	NOUN
cana-2597	151	13	,	,	PUNCT
cana-2597	151	14	selenium	selenium	NOUN
cana-2597	151	15	and	and	CCONJ
cana-2597	151	16	beautifulsoup	beautifulsoup	NOUN
cana-2597	151	17	were	be	AUX
cana-2597	151	18	employed	employ	VERB
cana-2597	151	19	to	to	PART
cana-2597	151	20	scrape	scrape	VERB
cana-2597	151	21	tweets	tweet	NOUN
cana-2597	151	22	directly	directly	ADV
cana-2597	151	23	from	from	ADP
cana-2597	151	24	twitter	twitter	NOUN
cana-2597	151	25	’s	’s	PART
cana-2597	151	26	web	web	NOUN
cana-2597	151	27	interface	interface	NOUN
cana-2597	151	28	.	.	PUNCT
cana-2597	152	1	this	this	DET
cana-2597	152	2	method	method	NOUN
cana-2597	152	3	allows	allow	VERB
cana-2597	152	4	us	we	PRON
cana-2597	152	5	to	to	PART
cana-2597	152	6	collect	collect	VERB
cana-2597	152	7	the	the	DET
cana-2597	152	8	necessary	necessary	ADJ
cana-2597	152	9	tweets	tweet	NOUN
cana-2597	152	10	,	,	PUNCT
cana-2597	152	11	user	user	NOUN
cana-2597	152	12	metadata	metadata	NOUN
cana-2597	152	13	,	,	PUNCT
cana-2597	152	14	and	and	CCONJ
cana-2597	152	15	other	other	ADJ
cana-2597	152	16	related	related	ADJ
cana-2597	152	17	information	information	NOUN
cana-2597	152	18	for	for	ADP
cana-2597	152	19	further	further	ADJ
cana-2597	152	20	processing	processing	NOUN
cana-2597	152	21	.	.	PUNCT
cana-2597	153	1	steps	step	NOUN
cana-2597	153	2	involved	involve	VERB
cana-2597	153	3	:	:	PUNCT
cana-2597	153	4	1	1	X
cana-2597	153	5	.	.	X
cana-2597	153	6	login	login	NOUN
cana-2597	153	7	management	management	NOUN
cana-2597	153	8	:	:	PUNCT
cana-2597	153	9	selenium	selenium	NOUN
cana-2597	153	10	automates	automate	VERB
cana-2597	153	11	the	the	DET
cana-2597	153	12	login	login	NOUN
cana-2597	153	13	process	process	NOUN
cana-2597	153	14	and	and	CCONJ
cana-2597	153	15	manages	manage	VERB
cana-2597	153	16	cookies	cookie	NOUN
cana-2597	153	17	to	to	PART
cana-2597	153	18	maintain	maintain	VERB
cana-2597	153	19	a	a	DET
cana-2597	153	20	valid	valid	ADJ
cana-2597	153	21	session	session	NOUN
cana-2597	153	22	.	.	PUNCT
cana-2597	154	1	2	2	X
cana-2597	154	2	.	.	X
cana-2597	154	3	automated	automate	VERB
cana-2597	154	4	scrolling	scrolling	NOUN
cana-2597	154	5	:	:	PUNCT
cana-2597	154	6	selenium	selenium	NOUN
cana-2597	154	7	simulates	simulate	NOUN
cana-2597	154	8	user	user	NOUN
cana-2597	154	9	3	3	X
cana-2597	154	10	.	.	PUNCT
cana-2597	155	1	scrolling	scroll	VERB
cana-2597	155	2	to	to	PART
cana-2597	155	3	load	load	VERB
cana-2597	155	4	older	old	ADJ
cana-2597	155	5	tweets	tweet	NOUN
cana-2597	155	6	dynamically	dynamically	ADV
cana-2597	155	7	.	.	PUNCT
cana-2597	156	1	4	4	X
cana-2597	156	2	.	.	X
cana-2597	156	3	content	content	NOUN
cana-2597	156	4	extraction	extraction	PROPN
cana-2597	156	5	:	:	PUNCT
cana-2597	156	6	beautifulsoup	beautifulsoup	NOUN
cana-2597	156	7	is	be	AUX
cana-2597	156	8	used	use	VERB
cana-2597	156	9	to	to	PART
cana-2597	156	10	extract	extract	VERB
cana-2597	156	11	tweet	tweet	NOUN
cana-2597	156	12	content	content	NOUN
cana-2597	156	13	,	,	PUNCT
cana-2597	156	14	metadata	metadata	NOUN
cana-2597	156	15	(	(	PUNCT
cana-2597	156	16	such	such	ADJ
cana-2597	156	17	as	as	ADP
cana-2597	156	18	timestamp	timestamp	NOUN
cana-2597	156	19	,	,	PUNCT
cana-2597	156	20	user	user	NOUN
cana-2597	156	21	location	location	NOUN
cana-2597	156	22	)	)	PUNCT
cana-2597	156	23	,	,	PUNCT
cana-2597	156	24	and	and	CCONJ
cana-2597	156	25	associated	associated	ADJ
cana-2597	156	26	images	image	NOUN
cana-2597	156	27	.	.	PUNCT
cana-2597	157	1	5	5	X
cana-2597	157	2	.	.	X
cana-2597	157	3	proxy	proxy	NOUN
cana-2597	157	4	and	and	CCONJ
cana-2597	157	5	anti	anti	ADJ
cana-2597	157	6	-	-	ADJ
cana-2597	157	7	bot	bot	ADJ
cana-2597	157	8	mechanisms	mechanism	NOUN
cana-2597	157	9	:	:	PUNCT
cana-2597	157	10	techniques	technique	NOUN
cana-2597	157	11	like	like	ADP
cana-2597	157	12	proxy	proxy	NOUN
cana-2597	157	13	rotation	rotation	NOUN
cana-2597	157	14	and	and	CCONJ
cana-2597	157	15	user	user	NOUN
cana-2597	157	16	-	-	PUNCT
cana-2597	157	17	agent	agent	NOUN
cana-2597	157	18	spoofing	spoofing	NOUN
cana-2597	157	19	were	be	AUX
cana-2597	157	20	implemented	implement	VERB
cana-2597	157	21	to	to	PART
cana-2597	157	22	mimic	mimic	VERB
cana-2597	157	23	human	human	ADJ
cana-2597	157	24	browsing	browse	VERB
cana-2597	157	25	behavior	behavior	NOUN
cana-2597	157	26	and	and	CCONJ
cana-2597	157	27	bypass	bypass	NOUN
cana-2597	157	28	twitter	twitter	NOUN
cana-2597	157	29	's	's	PART
cana-2597	157	30	anti	anti	ADJ
cana-2597	157	31	-	-	ADJ
cana-2597	157	32	bot	bot	ADJ
cana-2597	157	33	detection	detection	NOUN
cana-2597	157	34	.	.	PUNCT
cana-2597	158	1	once	once	ADV
cana-2597	158	2	scraped	scrape	VERB
cana-2597	158	3	,	,	PUNCT
cana-2597	158	4	the	the	DET
cana-2597	158	5	data	datum	NOUN
cana-2597	158	6	(	(	PUNCT
cana-2597	158	7	tweets	tweet	NOUN
cana-2597	158	8	,	,	PUNCT
cana-2597	158	9	metadata	metadata	NOUN
cana-2597	158	10	,	,	PUNCT
cana-2597	158	11	and	and	CCONJ
cana-2597	158	12	images	image	NOUN
cana-2597	158	13	)	)	PUNCT
cana-2597	158	14	is	be	AUX
cana-2597	158	15	streamed	stream	VERB
cana-2597	158	16	into	into	ADP
cana-2597	158	17	apache	apache	PROPN
cana-2597	158	18	kafka	kafka	PROPN
cana-2597	158	19	for	for	ADP
cana-2597	158	20	further	further	ADJ
cana-2597	158	21	processing	processing	NOUN
cana-2597	158	22	.	.	PUNCT
cana-2597	159	1	code	code	NOUN
cana-2597	159	2	for	for	ADP
cana-2597	159	3	webscraping	webscrape	VERB
cana-2597	159	4	communications	communication	NOUN
cana-2597	159	5	on	on	ADP
cana-2597	159	6	applied	apply	VERB
cana-2597	159	7	nonlinear	nonlinear	ADJ
cana-2597	159	8	analysis	analysis	NOUN
cana-2597	159	9	issn	issn	NOUN
cana-2597	159	10	:	:	PUNCT
cana-2597	159	11	1074	1074	NUM
cana-2597	159	12	-	-	PUNCT
cana-2597	159	13	133x	133x	NUM
cana-2597	159	14	vol	vol	NOUN
cana-2597	159	15	32	32	NUM
cana-2597	159	16	no	no	NOUN
cana-2597	159	17	.	.	PUNCT
cana-2597	160	1	3s	3s	NUM
cana-2597	160	2	(	(	PUNCT
cana-2597	160	3	2025	2025	NUM
cana-2597	160	4	)	)	PUNCT
cana-2597	160	5	207	207	NUM
cana-2597	160	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2597	160	7	4.3	4.3	NUM
cana-2597	160	8	.	.	PUNCT
cana-2597	160	9	real	real	ADJ
cana-2597	160	10	-	-	PUNCT
cana-2597	160	11	time	time	NOUN
cana-2597	160	12	data	datum	NOUN
cana-2597	160	13	ingestion	ingestion	NOUN
cana-2597	160	14	with	with	ADP
cana-2597	160	15	apache	apache	PROPN
cana-2597	160	16	kafka	kafka	PROPN
cana-2597	160	17	the	the	DET
cana-2597	160	18	scraped	scrape	VERB
cana-2597	160	19	data	datum	NOUN
cana-2597	160	20	is	be	AUX
cana-2597	160	21	immediately	immediately	ADV
cana-2597	160	22	streamed	stream	VERB
cana-2597	160	23	into	into	ADP
cana-2597	160	24	apache	apache	PROPN
cana-2597	160	25	kafka	kafka	PROPN
cana-2597	160	26	for	for	ADP
cana-2597	160	27	real	real	ADJ
cana-2597	160	28	-	-	PUNCT
cana-2597	160	29	time	time	NOUN
cana-2597	160	30	ingestion	ingestion	NOUN
cana-2597	160	31	.	.	PUNCT
cana-2597	161	1	kafka	kafka	PROPN
cana-2597	161	2	acts	act	VERB
cana-2597	161	3	as	as	ADP
cana-2597	161	4	a	a	DET
cana-2597	161	5	message	message	NOUN
cana-2597	161	6	broker	broker	NOUN
cana-2597	161	7	that	that	PRON
cana-2597	161	8	handles	handle	VERB
cana-2597	161	9	high	high	ADJ
cana-2597	161	10	-	-	PUNCT
cana-2597	161	11	throughput	throughput	NOUN
cana-2597	161	12	streams	stream	NOUN
cana-2597	161	13	of	of	ADP
cana-2597	161	14	tweets	tweet	NOUN
cana-2597	161	15	,	,	PUNCT
cana-2597	161	16	partitioning	partition	VERB
cana-2597	161	17	the	the	DET
cana-2597	161	18	data	datum	NOUN
cana-2597	161	19	into	into	ADP
cana-2597	161	20	different	different	ADJ
cana-2597	161	21	topics	topic	NOUN
cana-2597	161	22	based	base	VERB
cana-2597	161	23	on	on	ADP
cana-2597	161	24	keywords	keyword	NOUN
cana-2597	161	25	or	or	CCONJ
cana-2597	161	26	geographic	geographic	ADJ
cana-2597	161	27	locations	location	NOUN
cana-2597	161	28	.	.	PUNCT
cana-2597	162	1	1	1	X
cana-2597	162	2	.	.	X
cana-2597	162	3	kafka	kafka	PROPN
cana-2597	162	4	producer	producer	PROPN
cana-2597	162	5	api	api	NOUN
cana-2597	162	6	streams	stream	VERB
cana-2597	162	7	the	the	DET
cana-2597	162	8	tweets	tweet	NOUN
cana-2597	162	9	to	to	ADP
cana-2597	162	10	respective	respective	ADJ
cana-2597	162	11	topics	topic	NOUN
cana-2597	162	12	(	(	PUNCT
cana-2597	162	13	e.g.	e.g.	ADV
cana-2597	162	14	,	,	PUNCT
cana-2597	162	15	“	"	PUNCT
cana-2597	162	16	#	#	SYM
cana-2597	162	17	flood	flood	NOUN
cana-2597	162	18	”	"	PUNCT
cana-2597	162	19	,	,	PUNCT
cana-2597	162	20	“	"	PUNCT
cana-2597	162	21	#	#	SYM
cana-2597	162	22	fire	fire	NOUN
cana-2597	162	23	”	"	PUNCT
cana-2597	162	24	)	)	PUNCT
cana-2597	162	25	.	.	PUNCT
cana-2597	163	1	2	2	X
cana-2597	163	2	.	.	X
cana-2597	163	3	partitioning	partition	VERB
cana-2597	163	4	:	:	PUNCT
cana-2597	163	5	tweets	tweet	NOUN
cana-2597	163	6	are	be	AUX
cana-2597	163	7	partitioned	partition	VERB
cana-2597	163	8	by	by	ADP
cana-2597	163	9	disaster	disaster	NOUN
cana-2597	163	10	types	type	NOUN
cana-2597	163	11	(	(	PUNCT
cana-2597	163	12	e.g.	e.g.	ADV
cana-2597	163	13	,	,	PUNCT
cana-2597	163	14	“	"	PUNCT
cana-2597	163	15	earthquake	earthquake	NOUN
cana-2597	163	16	”	"	PUNCT
cana-2597	163	17	,	,	PUNCT
cana-2597	163	18	“	"	PUNCT
cana-2597	163	19	flood	flood	NOUN
cana-2597	163	20	”	"	PUNCT
cana-2597	163	21	)	)	PUNCT
cana-2597	163	22	to	to	PART
cana-2597	163	23	allow	allow	VERB
cana-2597	163	24	for	for	ADP
cana-2597	163	25	parallel	parallel	ADJ
cana-2597	163	26	processing	processing	NOUN
cana-2597	163	27	.	.	PUNCT
cana-2597	164	1	this	this	DET
cana-2597	164	2	partitioning	partitioning	NOUN
cana-2597	164	3	ensures	ensure	VERB
cana-2597	164	4	scalable	scalable	ADJ
cana-2597	164	5	,	,	PUNCT
cana-2597	164	6	real	real	ADJ
cana-2597	164	7	-	-	PUNCT
cana-2597	164	8	time	time	NOUN
cana-2597	164	9	disaster	disaster	NOUN
cana-2597	164	10	monitoring	monitoring	NOUN
cana-2597	164	11	.	.	PUNCT
cana-2597	165	1	4.4	4.4	NUM
cana-2597	165	2	.	.	PUNCT
cana-2597	166	1	real	real	ADJ
cana-2597	166	2	-	-	PUNCT
cana-2597	166	3	time	time	NOUN
cana-2597	166	4	data	datum	NOUN
cana-2597	166	5	processing	processing	NOUN
cana-2597	166	6	with	with	ADP
cana-2597	166	7	apache	apache	NOUN
cana-2597	166	8	spark	spark	NOUN
cana-2597	166	9	once	once	SCONJ
cana-2597	166	10	the	the	DET
cana-2597	166	11	tweets	tweet	NOUN
cana-2597	166	12	are	be	AUX
cana-2597	166	13	ingested	ingest	VERB
cana-2597	166	14	,	,	PUNCT
cana-2597	166	15	apache	apache	NOUN
cana-2597	166	16	spark	spark	NOUN
cana-2597	166	17	streaming	streaming	NOUN
cana-2597	166	18	processes	process	VERB
cana-2597	166	19	the	the	DET
cana-2597	166	20	data	datum	NOUN
cana-2597	166	21	in	in	ADP
cana-2597	166	22	micro	micro	NOUN
cana-2597	166	23	-	-	NOUN
cana-2597	166	24	batches	batch	NOUN
cana-2597	166	25	.	.	PUNCT
cana-2597	167	1	this	this	DET
cana-2597	167	2	realtime	realtime	ADJ
cana-2597	167	3	processing	processing	NOUN
cana-2597	167	4	pipeline	pipeline	NOUN
cana-2597	167	5	performs	perform	VERB
cana-2597	167	6	the	the	DET
cana-2597	167	7	following	following	ADJ
cana-2597	167	8	tasks	task	NOUN
cana-2597	167	9	:	:	PUNCT
cana-2597	167	10	code	code	NOUN
cana-2597	167	11	for	for	ADP
cana-2597	167	12	sparksession	sparksession	NOUN
cana-2597	167	13	1	1	NUM
cana-2597	167	14	.	.	PUNCT
cana-2597	167	15	sentiment	sentiment	NOUN
cana-2597	167	16	analysis	analysis	NOUN
cana-2597	167	17	:	:	PUNCT
cana-2597	167	18	nlp	nlp	ADJ
cana-2597	167	19	techniques	technique	NOUN
cana-2597	167	20	are	be	AUX
cana-2597	167	21	used	use	VERB
cana-2597	167	22	to	to	PART
cana-2597	167	23	assess	assess	VERB
cana-2597	167	24	the	the	DET
cana-2597	167	25	sentiment	sentiment	NOUN
cana-2597	167	26	of	of	ADP
cana-2597	167	27	tweets	tweet	NOUN
cana-2597	167	28	related	relate	VERB
cana-2597	167	29	to	to	ADP
cana-2597	167	30	disaster	disaster	NOUN
cana-2597	167	31	events	event	NOUN
cana-2597	167	32	.	.	PUNCT
cana-2597	168	1	2	2	X
cana-2597	168	2	.	.	X
cana-2597	168	3	location	location	NOUN
cana-2597	168	4	extraction	extraction	NOUN
cana-2597	168	5	:	:	PUNCT
cana-2597	168	6	extracts	extract	VERB
cana-2597	168	7	disaster	disaster	NOUN
cana-2597	168	8	locations	location	NOUN
cana-2597	168	9	and	and	CCONJ
cana-2597	168	10	severity	severity	NOUN
cana-2597	168	11	from	from	ADP
cana-2597	168	12	the	the	DET
cana-2597	168	13	tweet	tweet	NOUN
cana-2597	168	14	content	content	NOUN
cana-2597	168	15	using	use	VERB
cana-2597	168	16	geospatial	geospatial	ADJ
cana-2597	168	17	analysis	analysis	NOUN
cana-2597	168	18	and	and	CCONJ
cana-2597	168	19	named	name	VERB
cana-2597	168	20	entity	entity	NOUN
cana-2597	168	21	recognition	recognition	NOUN
cana-2597	168	22	(	(	PUNCT
cana-2597	168	23	ner	ner	NOUN
cana-2597	168	24	)	)	PUNCT
cana-2597	168	25	.	.	PUNCT
cana-2597	169	1	by	by	ADP
cana-2597	169	2	leveraging	leverage	VERB
cana-2597	169	3	spark	spark	NOUN
cana-2597	169	4	’s	’s	PART
cana-2597	169	5	distributed	distribute	VERB
cana-2597	169	6	processing	processing	NOUN
cana-2597	169	7	capabilities	capability	NOUN
cana-2597	169	8	,	,	PUNCT
cana-2597	169	9	the	the	DET
cana-2597	169	10	system	system	NOUN
cana-2597	169	11	can	can	AUX
cana-2597	169	12	handle	handle	VERB
cana-2597	169	13	a	a	DET
cana-2597	169	14	massive	massive	ADJ
cana-2597	169	15	influx	influx	NOUN
cana-2597	169	16	of	of	ADP
cana-2597	169	17	tweets	tweet	NOUN
cana-2597	169	18	,	,	PUNCT
cana-2597	169	19	even	even	ADV
cana-2597	169	20	during	during	ADP
cana-2597	169	21	high	high	ADJ
cana-2597	169	22	-	-	PUNCT
cana-2597	169	23	traffic	traffic	NOUN
cana-2597	169	24	periods	period	NOUN
cana-2597	169	25	,	,	PUNCT
cana-2597	169	26	ensuring	ensure	VERB
cana-2597	169	27	insights	insight	NOUN
cana-2597	169	28	are	be	AUX
cana-2597	169	29	delivered	deliver	VERB
cana-2597	169	30	in	in	ADP
cana-2597	169	31	real	real	ADJ
cana-2597	169	32	-	-	PUNCT
cana-2597	169	33	time	time	NOUN
cana-2597	169	34	.	.	PUNCT
cana-2597	170	1	4.5	4.5	NUM
cana-2597	170	2	.	.	PUNCT
cana-2597	171	1	multimodal	multimodal	NOUN
cana-2597	171	2	classification	classification	NOUN
cana-2597	171	3	model	model	NOUN
cana-2597	171	4	the	the	DET
cana-2597	171	5	core	core	NOUN
cana-2597	171	6	of	of	ADP
cana-2597	171	7	the	the	DET
cana-2597	171	8	system	system	NOUN
cana-2597	171	9	is	be	AUX
cana-2597	171	10	a	a	DET
cana-2597	171	11	multimodal	multimodal	ADJ
cana-2597	171	12	classification	classification	NOUN
cana-2597	171	13	model	model	NOUN
cana-2597	171	14	designed	design	VERB
cana-2597	171	15	to	to	PART
cana-2597	171	16	process	process	VERB
cana-2597	171	17	both	both	PRON
cana-2597	171	18	text	text	NOUN
cana-2597	171	19	and	and	CCONJ
cana-2597	171	20	image	image	NOUN
cana-2597	171	21	data	datum	NOUN
cana-2597	171	22	.	.	PUNCT
cana-2597	172	1	for	for	ADP
cana-2597	172	2	text	text	NOUN
cana-2597	172	3	classification	classification	NOUN
cana-2597	172	4	,	,	PUNCT
cana-2597	172	5	the	the	DET
cana-2597	172	6	system	system	NOUN
cana-2597	172	7	leverages	leverage	VERB
cana-2597	172	8	bert	bert	PROPN
cana-2597	172	9	and	and	CCONJ
cana-2597	172	10	xlnet	xlnet	PROPN
cana-2597	172	11	,	,	PUNCT
cana-2597	172	12	two	two	NUM
cana-2597	172	13	powerful	powerful	ADJ
cana-2597	172	14	natural	natural	ADJ
cana-2597	172	15	language	language	NOUN
cana-2597	172	16	processing	processing	NOUN
cana-2597	172	17	models	model	NOUN
cana-2597	172	18	that	that	PRON
cana-2597	172	19	have	have	AUX
cana-2597	172	20	been	be	AUX
cana-2597	172	21	fine	fine	ADV
cana-2597	172	22	-	-	PUNCT
cana-2597	172	23	tuned	tune	VERB
cana-2597	172	24	on	on	ADP
cana-2597	172	25	disaster	disaster	NOUN
cana-2597	172	26	-	-	PUNCT
cana-2597	172	27	specific	specific	ADJ
cana-2597	172	28	datasets	dataset	NOUN
cana-2597	172	29	.	.	PUNCT
cana-2597	173	1	these	these	DET
cana-2597	173	2	models	model	NOUN
cana-2597	173	3	are	be	AUX
cana-2597	173	4	able	able	ADJ
cana-2597	173	5	to	to	PART
cana-2597	173	6	capture	capture	VERB
cana-2597	173	7	the	the	DET
cana-2597	173	8	context	context	NOUN
cana-2597	173	9	and	and	CCONJ
cana-2597	173	10	semantic	semantic	ADJ
cana-2597	173	11	meaning	meaning	NOUN
cana-2597	173	12	from	from	ADP
cana-2597	173	13	the	the	DET
cana-2597	173	14	tweets	tweet	NOUN
cana-2597	173	15	,	,	PUNCT
cana-2597	173	16	making	make	VERB
cana-2597	173	17	them	they	PRON
cana-2597	173	18	particularly	particularly	ADV
cana-2597	173	19	effective	effective	ADJ
cana-2597	173	20	for	for	ADP
cana-2597	173	21	identifying	identify	VERB
cana-2597	173	22	and	and	CCONJ
cana-2597	173	23	classifying	classify	VERB
cana-2597	173	24	disaster	disaster	NOUN
cana-2597	173	25	-	-	PUNCT
cana-2597	173	26	related	relate	VERB
cana-2597	173	27	text	text	NOUN
cana-2597	173	28	content	content	NOUN
cana-2597	173	29	.	.	PUNCT
cana-2597	174	1	in	in	ADP
cana-2597	174	2	parallel	parallel	NOUN
cana-2597	174	3	,	,	PUNCT
cana-2597	174	4	efficientnet	efficientnet	NOUN
cana-2597	174	5	-	-	PUNCT
cana-2597	174	6	b3	b3	NOUN
cana-2597	174	7	is	be	AUX
cana-2597	174	8	used	use	VERB
cana-2597	174	9	for	for	ADP
cana-2597	174	10	image	image	NOUN
cana-2597	174	11	classification	classification	NOUN
cana-2597	174	12	.	.	PUNCT
cana-2597	175	1	this	this	DET
cana-2597	175	2	model	model	NOUN
cana-2597	175	3	has	have	AUX
cana-2597	175	4	been	be	AUX
cana-2597	175	5	trained	train	VERB
cana-2597	175	6	on	on	ADP
cana-2597	175	7	a	a	DET
cana-2597	175	8	curated	curate	VERB
cana-2597	175	9	dataset	dataset	NOUN
cana-2597	175	10	of	of	ADP
cana-2597	175	11	disaster	disaster	NOUN
cana-2597	175	12	images	image	NOUN
cana-2597	175	13	,	,	PUNCT
cana-2597	175	14	enabling	enable	VERB
cana-2597	175	15	it	it	PRON
cana-2597	175	16	to	to	PART
cana-2597	175	17	accurately	accurately	ADV
cana-2597	175	18	categorize	categorize	VERB
cana-2597	175	19	visual	visual	ADJ
cana-2597	175	20	data	datum	NOUN
cana-2597	175	21	into	into	ADP
cana-2597	175	22	classes	class	NOUN
cana-2597	175	23	such	such	ADJ
cana-2597	175	24	as	as	ADP
cana-2597	175	25	flood	flood	NOUN
cana-2597	175	26	,	,	PUNCT
cana-2597	175	27	fire	fire	NOUN
cana-2597	175	28	,	,	PUNCT
cana-2597	175	29	and	and	CCONJ
cana-2597	175	30	non	non	ADJ
cana-2597	175	31	-	-	NOUN
cana-2597	175	32	damage	damage	NOUN
cana-2597	175	33	.	.	PUNCT
cana-2597	176	1	to	to	PART
cana-2597	176	2	combine	combine	VERB
cana-2597	176	3	the	the	DET
cana-2597	176	4	strengths	strength	NOUN
cana-2597	176	5	of	of	ADP
cana-2597	176	6	both	both	DET
cana-2597	176	7	text	text	NOUN
cana-2597	176	8	and	and	CCONJ
cana-2597	176	9	image	image	NOUN
cana-2597	176	10	data	datum	NOUN
cana-2597	176	11	,	,	PUNCT
cana-2597	176	12	the	the	DET
cana-2597	176	13	system	system	NOUN
cana-2597	176	14	utilizes	utilize	VERB
cana-2597	176	15	a	a	DET
cana-2597	176	16	model	model	NOUN
cana-2597	176	17	fusion	fusion	NOUN
cana-2597	176	18	approach	approach	NOUN
cana-2597	176	19	.	.	PUNCT
cana-2597	177	1	the	the	DET
cana-2597	177	2	embeddings	embedding	NOUN
cana-2597	177	3	produced	produce	VERB
cana-2597	177	4	by	by	ADP
cana-2597	177	5	the	the	DET
cana-2597	177	6	bert	bert	PROPN
cana-2597	177	7	/	/	SYM
cana-2597	177	8	xlnet	xlnet	PROPN
cana-2597	177	9	models	model	NOUN
cana-2597	177	10	for	for	ADP
cana-2597	177	11	text	text	NOUN
cana-2597	177	12	and	and	CCONJ
cana-2597	177	13	the	the	DET
cana-2597	177	14	efficientnet	efficientnet	NOUN
cana-2597	177	15	-	-	PUNCT
cana-2597	177	16	b3	b3	NOUN
cana-2597	177	17	model	model	NOUN
cana-2597	177	18	for	for	ADP
cana-2597	177	19	images	image	NOUN
cana-2597	177	20	are	be	AUX
cana-2597	177	21	concatenated	concatenate	VERB
cana-2597	177	22	in	in	ADP
cana-2597	177	23	a	a	DET
cana-2597	177	24	fusion	fusion	NOUN
cana-2597	177	25	layer	layer	NOUN
cana-2597	177	26	.	.	PUNCT
cana-2597	178	1	this	this	DET
cana-2597	178	2	unified	unify	VERB
cana-2597	178	3	representation	representation	NOUN
cana-2597	178	4	of	of	ADP
cana-2597	178	5	the	the	DET
cana-2597	178	6	tweet	tweet	NOUN
cana-2597	178	7	,	,	PUNCT
cana-2597	178	8	which	which	PRON
cana-2597	178	9	now	now	ADV
cana-2597	178	10	contains	contain	VERB
cana-2597	178	11	both	both	DET
cana-2597	178	12	textual	textual	ADJ
cana-2597	178	13	and	and	CCONJ
cana-2597	178	14	visual	visual	ADJ
cana-2597	178	15	information	information	NOUN
cana-2597	178	16	,	,	PUNCT
cana-2597	178	17	is	be	AUX
cana-2597	178	18	then	then	ADV
cana-2597	178	19	passed	pass	VERB
cana-2597	178	20	through	through	ADP
cana-2597	178	21	fully	fully	ADV
cana-2597	178	22	connected	connect	VERB
cana-2597	178	23	layers	layer	NOUN
cana-2597	178	24	.	.	PUNCT
cana-2597	179	1	the	the	DET
cana-2597	179	2	final	final	ADJ
cana-2597	179	3	output	output	NOUN
cana-2597	179	4	is	be	AUX
cana-2597	179	5	a	a	DET
cana-2597	179	6	classification	classification	NOUN
cana-2597	179	7	of	of	ADP
cana-2597	179	8	the	the	DET
cana-2597	179	9	tweet	tweet	NOUN
cana-2597	179	10	,	,	PUNCT
cana-2597	179	11	indicating	indicate	VERB
cana-2597	179	12	whether	whether	SCONJ
cana-2597	179	13	it	it	PRON
cana-2597	179	14	pertains	pertain	VERB
cana-2597	179	15	to	to	ADP
cana-2597	179	16	a	a	DET
cana-2597	179	17	flood	flood	NOUN
cana-2597	179	18	,	,	PUNCT
cana-2597	179	19	fire	fire	NOUN
cana-2597	179	20	,	,	PUNCT
cana-2597	179	21	or	or	CCONJ
cana-2597	179	22	non	non	ADJ
cana-2597	179	23	-	-	ADJ
cana-2597	179	24	disaster	disaster	ADJ
cana-2597	179	25	event	event	NOUN
cana-2597	179	26	.	.	PUNCT
cana-2597	180	1	by	by	ADP
cana-2597	180	2	communications	communication	NOUN
cana-2597	180	3	on	on	ADP
cana-2597	180	4	applied	apply	VERB
cana-2597	180	5	nonlinear	nonlinear	ADJ
cana-2597	180	6	analysis	analysis	NOUN
cana-2597	180	7	issn	issn	NOUN
cana-2597	180	8	:	:	PUNCT
cana-2597	180	9	1074	1074	NUM
cana-2597	180	10	-	-	PUNCT
cana-2597	180	11	133x	133x	NUM
cana-2597	180	12	vol	vol	NOUN
cana-2597	180	13	32	32	NUM
cana-2597	180	14	no	no	NOUN
cana-2597	180	15	.	.	PUNCT
cana-2597	181	1	3s	3s	NUM
cana-2597	181	2	(	(	PUNCT
cana-2597	181	3	2025	2025	NUM
cana-2597	181	4	)	)	PUNCT
cana-2597	181	5	208	208	NUM
cana-2597	181	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2597	181	7	fusing	fuse	VERB
cana-2597	181	8	these	these	DET
cana-2597	181	9	two	two	NUM
cana-2597	181	10	modalities	modality	NOUN
cana-2597	181	11	,	,	PUNCT
cana-2597	181	12	the	the	DET
cana-2597	181	13	system	system	NOUN
cana-2597	181	14	is	be	AUX
cana-2597	181	15	able	able	ADJ
cana-2597	181	16	to	to	PART
cana-2597	181	17	improve	improve	VERB
cana-2597	181	18	accuracy	accuracy	NOUN
cana-2597	181	19	and	and	CCONJ
cana-2597	181	20	leverage	leverage	VERB
cana-2597	181	21	the	the	DET
cana-2597	181	22	complementary	complementary	ADJ
cana-2597	181	23	strengths	strength	NOUN
cana-2597	181	24	of	of	ADP
cana-2597	181	25	text	text	NOUN
cana-2597	181	26	and	and	CCONJ
cana-2597	181	27	image	image	NOUN
cana-2597	181	28	data	datum	NOUN
cana-2597	181	29	.	.	PUNCT
cana-2597	182	1	4.6	4.6	NUM
cana-2597	182	2	.	.	PUNCT
cana-2597	182	3	training	training	NOUN
cana-2597	182	4	procedure	procedure	NOUN
cana-2597	182	5	in	in	ADP
cana-2597	182	6	the	the	DET
cana-2597	182	7	training	training	NOUN
cana-2597	182	8	procedure	procedure	NOUN
cana-2597	182	9	,	,	PUNCT
cana-2597	182	10	we	we	PRON
cana-2597	182	11	utilize	utilize	VERB
cana-2597	182	12	crossentropyloss	crossentropyloss	NOUN
cana-2597	182	13	as	as	ADP
cana-2597	182	14	the	the	DET
cana-2597	182	15	loss	loss	NOUN
cana-2597	182	16	function	function	NOUN
cana-2597	182	17	,	,	PUNCT
cana-2597	182	18	which	which	PRON
cana-2597	182	19	is	be	AUX
cana-2597	182	20	well	well	ADV
cana-2597	182	21	-	-	PUNCT
cana-2597	182	22	suited	suit	VERB
cana-2597	182	23	for	for	ADP
cana-2597	182	24	multi	multi	ADJ
cana-2597	182	25	-	-	ADJ
cana-2597	182	26	class	class	ADJ
cana-2597	182	27	classification	classification	NOUN
cana-2597	182	28	tasks	task	NOUN
cana-2597	182	29	.	.	PUNCT
cana-2597	183	1	this	this	PRON
cana-2597	183	2	ensures	ensure	VERB
cana-2597	183	3	that	that	SCONJ
cana-2597	183	4	the	the	DET
cana-2597	183	5	model	model	NOUN
cana-2597	183	6	is	be	AUX
cana-2597	183	7	optimized	optimize	VERB
cana-2597	183	8	to	to	PART
cana-2597	183	9	correctly	correctly	ADV
cana-2597	183	10	classify	classify	VERB
cana-2597	183	11	tweets	tweet	NOUN
cana-2597	183	12	into	into	ADP
cana-2597	183	13	disaster	disaster	NOUN
cana-2597	183	14	categories	category	NOUN
cana-2597	183	15	such	such	ADJ
cana-2597	183	16	as	as	ADP
cana-2597	183	17	flood	flood	NOUN
cana-2597	183	18	,	,	PUNCT
cana-2597	183	19	fire	fire	NOUN
cana-2597	183	20	,	,	PUNCT
cana-2597	183	21	or	or	CCONJ
cana-2597	183	22	non	non	ADJ
cana-2597	183	23	-	-	NOUN
cana-2597	183	24	damage	damage	NOUN
cana-2597	183	25	.	.	PUNCT
cana-2597	184	1	for	for	ADP
cana-2597	184	2	optimization	optimization	NOUN
cana-2597	184	3	,	,	PUNCT
cana-2597	184	4	we	we	PRON
cana-2597	184	5	use	use	VERB
cana-2597	184	6	the	the	DET
cana-2597	184	7	adamw	adamw	NOUN
cana-2597	184	8	optimizer	optimizer	NOUN
cana-2597	184	9	with	with	ADP
cana-2597	184	10	a	a	DET
cana-2597	184	11	learning	learn	VERB
cana-2597	184	12	rate	rate	NOUN
cana-2597	184	13	of	of	ADP
cana-2597	184	14	1e-4	1e-4	PROPN
cana-2597	184	15	.	.	PUNCT
cana-2597	184	16	adamw	adamw	PROPN
cana-2597	184	17	is	be	AUX
cana-2597	184	18	known	know	VERB
cana-2597	184	19	for	for	ADP
cana-2597	184	20	its	its	PRON
cana-2597	184	21	efficiency	efficiency	NOUN
cana-2597	184	22	in	in	ADP
cana-2597	184	23	handling	handle	VERB
cana-2597	184	24	large	large	ADJ
cana-2597	184	25	-	-	PUNCT
cana-2597	184	26	scale	scale	NOUN
cana-2597	184	27	datasets	dataset	NOUN
cana-2597	184	28	and	and	CCONJ
cana-2597	184	29	includes	include	VERB
cana-2597	184	30	built	build	VERB
cana-2597	184	31	-	-	PUNCT
cana-2597	184	32	in	in	ADP
cana-2597	184	33	weight	weight	NOUN
cana-2597	184	34	decay	decay	NOUN
cana-2597	184	35	regularization	regularization	NOUN
cana-2597	184	36	,	,	PUNCT
cana-2597	184	37	which	which	PRON
cana-2597	184	38	helps	help	VERB
cana-2597	184	39	prevent	prevent	VERB
cana-2597	184	40	overfitting	overfitting	NOUN
cana-2597	184	41	during	during	ADP
cana-2597	184	42	training	training	NOUN
cana-2597	184	43	.	.	PUNCT
cana-2597	185	1	the	the	DET
cana-2597	185	2	model	model	NOUN
cana-2597	185	3	is	be	AUX
cana-2597	185	4	trained	train	VERB
cana-2597	185	5	over	over	ADP
cana-2597	185	6	20	20	NUM
cana-2597	185	7	epochs	epoch	NOUN
cana-2597	185	8	,	,	PUNCT
cana-2597	185	9	with	with	ADP
cana-2597	185	10	early	early	ADJ
cana-2597	185	11	stopping	stop	VERB
cana-2597	185	12	in	in	ADP
cana-2597	185	13	place	place	NOUN
cana-2597	185	14	based	base	VERB
cana-2597	185	15	on	on	ADP
cana-2597	185	16	the	the	DET
cana-2597	185	17	validation	validation	NOUN
cana-2597	185	18	accuracy	accuracy	NOUN
cana-2597	185	19	.	.	PUNCT
cana-2597	186	1	this	this	PRON
cana-2597	186	2	ensures	ensure	VERB
cana-2597	186	3	that	that	SCONJ
cana-2597	186	4	training	training	NOUN
cana-2597	186	5	is	be	AUX
cana-2597	186	6	halted	halt	VERB
cana-2597	186	7	when	when	SCONJ
cana-2597	186	8	no	no	DET
cana-2597	186	9	significant	significant	ADJ
cana-2597	186	10	improvement	improvement	NOUN
cana-2597	186	11	is	be	AUX
cana-2597	186	12	observed	observe	VERB
cana-2597	186	13	in	in	ADP
cana-2597	186	14	validation	validation	NOUN
cana-2597	186	15	performance	performance	NOUN
cana-2597	186	16	,	,	PUNCT
cana-2597	186	17	reducing	reduce	VERB
cana-2597	186	18	overfitting	overfitte	VERB
cana-2597	186	19	and	and	CCONJ
cana-2597	186	20	unnecessary	unnecessary	ADJ
cana-2597	186	21	training	training	NOUN
cana-2597	186	22	time	time	NOUN
cana-2597	186	23	.	.	PUNCT
cana-2597	187	1	during	during	ADP
cana-2597	187	2	each	each	DET
cana-2597	187	3	epoch	epoch	NOUN
cana-2597	187	4	,	,	PUNCT
cana-2597	187	5	the	the	DET
cana-2597	187	6	model	model	NOUN
cana-2597	187	7	is	be	AUX
cana-2597	187	8	switched	switch	VERB
cana-2597	187	9	to	to	ADP
cana-2597	187	10	training	training	NOUN
cana-2597	187	11	mode	mode	NOUN
cana-2597	187	12	,	,	PUNCT
cana-2597	187	13	and	and	CCONJ
cana-2597	187	14	the	the	DET
cana-2597	187	15	loss	loss	NOUN
cana-2597	187	16	is	be	AUX
cana-2597	187	17	computed	compute	VERB
cana-2597	187	18	using	use	VERB
cana-2597	187	19	both	both	PRON
cana-2597	187	20	text	text	NOUN
cana-2597	187	21	and	and	CCONJ
cana-2597	187	22	image	image	NOUN
cana-2597	187	23	inputs	input	NOUN
cana-2597	187	24	.	.	PUNCT
cana-2597	188	1	the	the	DET
cana-2597	188	2	optimizer	optimizer	NOUN
cana-2597	188	3	updates	update	VERB
cana-2597	188	4	the	the	DET
cana-2597	188	5	model	model	NOUN
cana-2597	188	6	weights	weight	NOUN
cana-2597	188	7	based	base	VERB
cana-2597	188	8	on	on	ADP
cana-2597	188	9	the	the	DET
cana-2597	188	10	calculated	calculate	VERB
cana-2597	188	11	gradients	gradient	NOUN
cana-2597	188	12	.	.	PUNCT
cana-2597	189	1	for	for	ADP
cana-2597	189	2	each	each	DET
cana-2597	189	3	batch	batch	NOUN
cana-2597	189	4	,	,	PUNCT
cana-2597	189	5	the	the	DET
cana-2597	189	6	loss	loss	NOUN
cana-2597	189	7	is	be	AUX
cana-2597	189	8	computed	compute	VERB
cana-2597	189	9	,	,	PUNCT
cana-2597	189	10	backpropagated	backpropagate	VERB
cana-2597	189	11	,	,	PUNCT
cana-2597	189	12	and	and	CCONJ
cana-2597	189	13	accumulated	accumulate	VERB
cana-2597	189	14	,	,	PUNCT
cana-2597	189	15	allowing	allow	VERB
cana-2597	189	16	us	we	PRON
cana-2597	189	17	to	to	PART
cana-2597	189	18	track	track	VERB
cana-2597	189	19	the	the	DET
cana-2597	189	20	running	run	VERB
cana-2597	189	21	loss	loss	NOUN
cana-2597	189	22	throughout	throughout	ADP
cana-2597	189	23	the	the	DET
cana-2597	189	24	training	training	NOUN
cana-2597	189	25	process	process	NOUN
cana-2597	189	26	.	.	PUNCT
cana-2597	190	1	the	the	DET
cana-2597	190	2	loss	loss	NOUN
cana-2597	190	3	for	for	ADP
cana-2597	190	4	each	each	DET
cana-2597	190	5	epoch	epoch	NOUN
cana-2597	190	6	is	be	AUX
cana-2597	190	7	printed	print	VERB
cana-2597	190	8	at	at	ADP
cana-2597	190	9	the	the	DET
cana-2597	190	10	end	end	NOUN
cana-2597	190	11	to	to	PART
cana-2597	190	12	monitor	monitor	VERB
cana-2597	190	13	the	the	DET
cana-2597	190	14	model	model	NOUN
cana-2597	190	15	's	's	PART
cana-2597	190	16	learning	learn	VERB
cana-2597	190	17	progress	progress	NOUN
cana-2597	190	18	fig	fig	NOUN
cana-2597	190	19	3	3	NUM
cana-2597	190	20	.	.	PUNCT
cana-2597	190	21	train	train	NOUN
cana-2597	190	22	&	&	CCONJ
cana-2597	190	23	validation	validation	PROPN
cana-2597	190	24	accuracy	accuracy	NOUN
cana-2597	190	25	over	over	ADP
cana-2597	190	26	epochs	epoch	NOUN
cana-2597	190	27	fig	fig	NOUN
cana-2597	190	28	4	4	NUM
cana-2597	190	29	.	.	PUNCT
cana-2597	190	30	train	train	NOUN
cana-2597	190	31	&	&	CCONJ
cana-2597	190	32	validation	validation	PROPN
cana-2597	190	33	accuracy	accuracy	NOUN
cana-2597	190	34	over	over	ADP
cana-2597	190	35	epochs	epoch	NOUN
cana-2597	190	36	4.7	4.7	NUM
cana-2597	190	37	.	.	PUNCT
cana-2597	191	1	evaluation	evaluation	NOUN
cana-2597	191	2	metrics	metric	NOUN
cana-2597	191	3	the	the	DET
cana-2597	191	4	evaluation	evaluation	NOUN
cana-2597	191	5	of	of	ADP
cana-2597	191	6	the	the	DET
cana-2597	191	7	model	model	NOUN
cana-2597	191	8	's	's	PART
cana-2597	191	9	performance	performance	NOUN
cana-2597	191	10	is	be	AUX
cana-2597	191	11	carried	carry	VERB
cana-2597	191	12	out	out	ADP
cana-2597	191	13	using	use	VERB
cana-2597	191	14	several	several	ADJ
cana-2597	191	15	key	key	ADJ
cana-2597	191	16	metrics	metric	NOUN
cana-2597	191	17	.	.	PUNCT
cana-2597	192	1	accuracy	accuracy	NOUN
cana-2597	192	2	provides	provide	VERB
cana-2597	192	3	a	a	DET
cana-2597	192	4	basic	basic	ADJ
cana-2597	192	5	measure	measure	NOUN
cana-2597	192	6	of	of	ADP
cana-2597	192	7	overall	overall	ADJ
cana-2597	192	8	correctness	correctness	NOUN
cana-2597	192	9	by	by	ADP
cana-2597	192	10	determining	determine	VERB
cana-2597	192	11	the	the	DET
cana-2597	192	12	proportion	proportion	NOUN
cana-2597	192	13	of	of	ADP
cana-2597	192	14	correct	correct	ADJ
cana-2597	192	15	predictions	prediction	NOUN
cana-2597	192	16	.	.	PUNCT
cana-2597	193	1	to	to	PART
cana-2597	193	2	further	far	ADV
cana-2597	193	3	assess	assess	VERB
cana-2597	193	4	the	the	DET
cana-2597	193	5	model	model	NOUN
cana-2597	193	6	's	's	PART
cana-2597	193	7	effectiveness	effectiveness	NOUN
cana-2597	193	8	in	in	ADP
cana-2597	193	9	distinguishing	distinguish	VERB
cana-2597	193	10	between	between	ADP
cana-2597	193	11	different	different	ADJ
cana-2597	193	12	disaster	disaster	NOUN
cana-2597	193	13	categories	category	NOUN
cana-2597	193	14	,	,	PUNCT
cana-2597	193	15	precision	precision	NOUN
cana-2597	193	16	,	,	PUNCT
cana-2597	193	17	recall	recall	NOUN
cana-2597	193	18	,	,	PUNCT
cana-2597	193	19	and	and	CCONJ
cana-2597	193	20	the	the	DET
cana-2597	193	21	f1	f1	NOUN
cana-2597	193	22	-	-	PUNCT
cana-2597	193	23	score	score	NOUN
cana-2597	193	24	are	be	AUX
cana-2597	193	25	used	use	VERB
cana-2597	193	26	.	.	PUNCT
cana-2597	194	1	these	these	DET
cana-2597	194	2	metrics	metric	NOUN
cana-2597	194	3	provide	provide	VERB
cana-2597	194	4	deeper	deep	ADJ
cana-2597	194	5	insights	insight	NOUN
cana-2597	194	6	,	,	PUNCT
cana-2597	194	7	especially	especially	ADV
cana-2597	194	8	in	in	ADP
cana-2597	194	9	handling	handle	VERB
cana-2597	194	10	imbalanced	imbalanced	ADJ
cana-2597	194	11	data	datum	NOUN
cana-2597	194	12	by	by	ADP
cana-2597	194	13	focusing	focus	VERB
cana-2597	194	14	on	on	ADP
cana-2597	194	15	the	the	DET
cana-2597	194	16	balance	balance	NOUN
cana-2597	194	17	between	between	ADP
cana-2597	194	18	correctly	correctly	ADV
cana-2597	194	19	identified	identify	VERB
cana-2597	194	20	disasters	disaster	NOUN
cana-2597	194	21	and	and	CCONJ
cana-2597	194	22	the	the	DET
cana-2597	194	23	false	false	ADJ
cana-2597	194	24	positives	positive	NOUN
cana-2597	194	25	or	or	CCONJ
cana-2597	194	26	negatives	negative	NOUN
cana-2597	194	27	.	.	PUNCT
cana-2597	195	1	additionally	additionally	ADV
cana-2597	195	2	,	,	PUNCT
cana-2597	195	3	the	the	DET
cana-2597	195	4	confusion	confusion	NOUN
cana-2597	195	5	matrix	matrix	NOUN
cana-2597	195	6	is	be	AUX
cana-2597	195	7	used	use	VERB
cana-2597	195	8	to	to	PART
cana-2597	195	9	offer	offer	VERB
cana-2597	195	10	a	a	DET
cana-2597	195	11	more	more	ADV
cana-2597	195	12	detailed	detailed	ADJ
cana-2597	195	13	view	view	NOUN
cana-2597	195	14	of	of	ADP
cana-2597	195	15	the	the	DET
cana-2597	195	16	model	model	NOUN
cana-2597	195	17	's	's	PART
cana-2597	195	18	classification	classification	NOUN
cana-2597	195	19	performance	performance	NOUN
cana-2597	195	20	.	.	PUNCT
cana-2597	196	1	it	it	PRON
cana-2597	196	2	visualizes	visualize	VERB
cana-2597	196	3	how	how	SCONJ
cana-2597	196	4	well	well	ADV
cana-2597	196	5	the	the	DET
cana-2597	196	6	model	model	NOUN
cana-2597	196	7	is	be	AUX
cana-2597	196	8	distinguishing	distinguish	VERB
cana-2597	196	9	between	between	ADP
cana-2597	196	10	the	the	DET
cana-2597	196	11	three	three	NUM
cana-2597	196	12	classes	class	NOUN
cana-2597	196	13	—	—	PUNCT
cana-2597	196	14	flood	flood	NOUN
cana-2597	196	15	,	,	PUNCT
cana-2597	196	16	fire	fire	NOUN
cana-2597	196	17	,	,	PUNCT
cana-2597	196	18	and	and	CCONJ
cana-2597	196	19	non	non	ADJ
cana-2597	196	20	-	-	NOUN
cana-2597	196	21	damage	damage	NOUN
cana-2597	196	22	—	—	PUNCT
cana-2597	196	23	by	by	ADP
cana-2597	196	24	showing	show	VERB
cana-2597	196	25	the	the	DET
cana-2597	196	26	counts	count	NOUN
cana-2597	196	27	of	of	ADP
cana-2597	196	28	true	true	ADJ
cana-2597	196	29	positives	positive	NOUN
cana-2597	196	30	,	,	PUNCT
cana-2597	196	31	false	false	ADJ
cana-2597	196	32	positives	positive	NOUN
cana-2597	196	33	,	,	PUNCT
cana-2597	196	34	true	true	ADJ
cana-2597	196	35	negatives	negative	NOUN
cana-2597	196	36	,	,	PUNCT
cana-2597	196	37	and	and	CCONJ
cana-2597	196	38	false	false	ADJ
cana-2597	196	39	negatives	negative	NOUN
cana-2597	196	40	for	for	ADP
cana-2597	196	41	each	each	DET
cana-2597	196	42	class	class	NOUN
cana-2597	196	43	.	.	PUNCT
cana-2597	197	1	communications	communication	NOUN
cana-2597	197	2	on	on	ADP
cana-2597	197	3	applied	apply	VERB
cana-2597	197	4	nonlinear	nonlinear	ADJ
cana-2597	197	5	analysis	analysis	NOUN
cana-2597	197	6	issn	issn	NOUN
cana-2597	197	7	:	:	PUNCT
cana-2597	197	8	1074	1074	NUM
cana-2597	197	9	-	-	PUNCT
cana-2597	197	10	133x	133x	NUM
cana-2597	197	11	vol	vol	NOUN
cana-2597	197	12	32	32	NUM
cana-2597	197	13	no	no	NOUN
cana-2597	197	14	.	.	PUNCT
cana-2597	198	1	3s	3s	NUM
cana-2597	198	2	(	(	PUNCT
cana-2597	198	3	2025	2025	NUM
cana-2597	198	4	)	)	PUNCT
cana-2597	198	5	209	209	NUM
cana-2597	198	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2597	198	7	moreover	moreover	ADV
cana-2597	198	8	,	,	PUNCT
cana-2597	198	9	throughout	throughout	ADP
cana-2597	198	10	the	the	DET
cana-2597	198	11	training	training	NOUN
cana-2597	198	12	process	process	NOUN
cana-2597	198	13	,	,	PUNCT
cana-2597	198	14	we	we	PRON
cana-2597	198	15	track	track	VERB
cana-2597	198	16	training	training	NOUN
cana-2597	198	17	loss	loss	NOUN
cana-2597	198	18	,	,	PUNCT
cana-2597	198	19	validation	validation	NOUN
cana-2597	198	20	loss	loss	NOUN
cana-2597	198	21	,	,	PUNCT
cana-2597	198	22	training	training	NOUN
cana-2597	198	23	accuracy	accuracy	NOUN
cana-2597	198	24	,	,	PUNCT
cana-2597	198	25	and	and	CCONJ
cana-2597	198	26	validation	validation	NOUN
cana-2597	198	27	accuracy	accuracy	NOUN
cana-2597	198	28	.	.	PUNCT
cana-2597	199	1	these	these	DET
cana-2597	199	2	metrics	metric	NOUN
cana-2597	199	3	,	,	PUNCT
cana-2597	199	4	plotted	plot	VERB
cana-2597	199	5	over	over	ADP
cana-2597	199	6	the	the	DET
cana-2597	199	7	epochs	epoch	NOUN
cana-2597	199	8	,	,	PUNCT
cana-2597	199	9	help	help	VERB
cana-2597	199	10	in	in	ADP
cana-2597	199	11	analyzing	analyze	VERB
cana-2597	199	12	the	the	DET
cana-2597	199	13	model	model	NOUN
cana-2597	199	14	’s	’s	PART
cana-2597	199	15	convergence	convergence	NOUN
cana-2597	199	16	behavior	behavior	NOUN
cana-2597	199	17	,	,	PUNCT
cana-2597	199	18	indicating	indicate	VERB
cana-2597	199	19	whether	whether	SCONJ
cana-2597	199	20	the	the	DET
cana-2597	199	21	model	model	NOUN
cana-2597	199	22	is	be	AUX
cana-2597	199	23	improving	improve	VERB
cana-2597	199	24	,	,	PUNCT
cana-2597	199	25	overfitting	overfitting	NOUN
cana-2597	199	26	,	,	PUNCT
cana-2597	199	27	or	or	CCONJ
cana-2597	199	28	underfitting	underfitting	NOUN
cana-2597	199	29	.	.	PUNCT
cana-2597	200	1	by	by	ADP
cana-2597	200	2	monitoring	monitor	VERB
cana-2597	200	3	these	these	DET
cana-2597	200	4	metrics	metric	NOUN
cana-2597	200	5	,	,	PUNCT
cana-2597	200	6	we	we	PRON
cana-2597	200	7	can	can	AUX
cana-2597	200	8	adjust	adjust	VERB
cana-2597	200	9	the	the	DET
cana-2597	200	10	model	model	NOUN
cana-2597	200	11	training	training	NOUN
cana-2597	200	12	to	to	PART
cana-2597	200	13	optimize	optimize	VERB
cana-2597	200	14	its	its	PRON
cana-2597	200	15	performance	performance	NOUN
cana-2597	200	16	for	for	ADP
cana-2597	200	17	disaster	disaster	NOUN
cana-2597	200	18	tweet	tweet	NOUN
cana-2597	200	19	classification	classification	NOUN
cana-2597	200	20	.	.	PUNCT
cana-2597	201	1	fig	fig	NOUN
cana-2597	201	2	5	5	NUM
cana-2597	201	3	.	.	PUNCT
cana-2597	201	4	confusion	confusion	NOUN
cana-2597	201	5	matrix	matrix	NOUN
cana-2597	201	6	this	this	DET
cana-2597	201	7	implementation	implementation	NOUN
cana-2597	201	8	outlines	outline	VERB
cana-2597	201	9	the	the	DET
cana-2597	201	10	full	full	ADJ
cana-2597	201	11	pipeline	pipeline	NOUN
cana-2597	201	12	of	of	ADP
cana-2597	201	13	our	our	PRON
cana-2597	201	14	disaster	disaster	NOUN
cana-2597	201	15	tweet	tweet	NOUN
cana-2597	201	16	processing	processing	NOUN
cana-2597	201	17	system	system	NOUN
cana-2597	201	18	,	,	PUNCT
cana-2597	201	19	from	from	ADP
cana-2597	201	20	web	web	NOUN
cana-2597	201	21	scraping	scraping	NOUN
cana-2597	201	22	and	and	CCONJ
cana-2597	201	23	real	real	ADJ
cana-2597	201	24	-	-	PUNCT
cana-2597	201	25	time	time	NOUN
cana-2597	201	26	streaming	streaming	NOUN
cana-2597	201	27	with	with	ADP
cana-2597	201	28	kafka	kafka	PROPN
cana-2597	201	29	,	,	PUNCT
cana-2597	201	30	to	to	AUX
cana-2597	201	31	multimodal	multimodal	VERB
cana-2597	201	32	classification	classification	NOUN
cana-2597	201	33	using	use	VERB
cana-2597	201	34	bert	bert	PROPN
cana-2597	201	35	and	and	CCONJ
cana-2597	201	36	efficientnet	efficientnet	NOUN
cana-2597	201	37	,	,	PUNCT
cana-2597	201	38	tweet	tweet	NOUN
cana-2597	201	39	filtering	filtering	NOUN
cana-2597	201	40	with	with	ADP
cana-2597	201	41	gemini	gemini	PROPN
cana-2597	201	42	llm	llm	PROPN
cana-2597	201	43	,	,	PUNCT
cana-2597	201	44	and	and	CCONJ
cana-2597	201	45	data	datum	NOUN
cana-2597	201	46	storage	storage	NOUN
cana-2597	201	47	with	with	ADP
cana-2597	201	48	mongodb	mongodb	NOUN
cana-2597	201	49	and	and	CCONJ
cana-2597	201	50	elasticsearch	elasticsearch	NOUN
cana-2597	201	51	.	.	PUNCT
cana-2597	202	1	the	the	DET
cana-2597	202	2	system	system	NOUN
cana-2597	202	3	delivers	deliver	VERB
cana-2597	202	4	real	real	ADJ
cana-2597	202	5	-	-	PUNCT
cana-2597	202	6	time	time	NOUN
cana-2597	202	7	disaster	disaster	NOUN
cana-2597	202	8	insights	insight	NOUN
cana-2597	202	9	through	through	ADP
cana-2597	202	10	a	a	DET
cana-2597	202	11	combination	combination	NOUN
cana-2597	202	12	of	of	ADP
cana-2597	202	13	advanced	advanced	ADJ
cana-2597	202	14	data	datum	NOUN
cana-2597	202	15	processing	processing	NOUN
cana-2597	202	16	techniques	technique	NOUN
cana-2597	202	17	,	,	PUNCT
cana-2597	202	18	machine	machine	NOUN
cana-2597	202	19	learning	learning	NOUN
cana-2597	202	20	models	model	NOUN
cana-2597	202	21	,	,	PUNCT
cana-2597	202	22	and	and	CCONJ
cana-2597	202	23	visualization	visualization	NOUN
cana-2597	202	24	tools	tool	NOUN
cana-2597	202	25	.	.	PUNCT
cana-2597	203	1	v.	v.	ADP
cana-2597	203	2	results	result	NOUN
cana-2597	203	3	and	and	CCONJ
cana-2597	203	4	discussion	discussion	VERB
cana-2597	203	5	the	the	DET
cana-2597	203	6	results	result	NOUN
cana-2597	203	7	of	of	ADP
cana-2597	203	8	the	the	DET
cana-2597	203	9	comparative	comparative	ADJ
cana-2597	203	10	analysis	analysis	NOUN
cana-2597	203	11	between	between	ADP
cana-2597	203	12	the	the	DET
cana-2597	203	13	efficientnet	efficientnet	NOUN
cana-2597	203	14	-	-	PUNCT
cana-2597	203	15	bert	bert	NOUN
cana-2597	203	16	and	and	CCONJ
cana-2597	203	17	resnet	resnet	ADJ
cana-2597	203	18	-	-	PUNCT
cana-2597	203	19	bert	bert	NOUN
cana-2597	203	20	models	model	NOUN
cana-2597	203	21	demonstrate	demonstrate	VERB
cana-2597	203	22	distinct	distinct	ADJ
cana-2597	203	23	performance	performance	NOUN
cana-2597	203	24	characteristics	characteristic	NOUN
cana-2597	203	25	in	in	ADP
cana-2597	203	26	classifying	classify	VERB
cana-2597	203	27	disaster	disaster	NOUN
cana-2597	203	28	-	-	PUNCT
cana-2597	203	29	related	relate	VERB
cana-2597	203	30	tweets	tweet	NOUN
cana-2597	203	31	.	.	PUNCT
cana-2597	204	1	both	both	DET
cana-2597	204	2	models	model	NOUN
cana-2597	204	3	achieved	achieve	VERB
cana-2597	204	4	high	high	ADJ
cana-2597	204	5	accuracy	accuracy	NOUN
cana-2597	204	6	rates	rate	NOUN
cana-2597	204	7	,	,	PUNCT
cana-2597	204	8	but	but	CCONJ
cana-2597	204	9	efficientnet	efficientnet	ADJ
cana-2597	204	10	-	-	PUNCT
cana-2597	204	11	bert	bert	NOUN
cana-2597	204	12	outperformed	outperform	VERB
cana-2597	204	13	resnet	resnet	NOUN
cana-2597	204	14	-	-	PUNCT
cana-2597	204	15	bert	bert	NOUN
cana-2597	204	16	across	across	ADP
cana-2597	204	17	all	all	DET
cana-2597	204	18	key	key	ADJ
cana-2597	204	19	metrics	metric	NOUN
cana-2597	204	20	,	,	PUNCT
cana-2597	204	21	indicating	indicate	VERB
cana-2597	204	22	its	its	PRON
cana-2597	204	23	superior	superior	ADJ
cana-2597	204	24	ability	ability	NOUN
cana-2597	204	25	to	to	PART
cana-2597	204	26	differentiate	differentiate	VERB
cana-2597	204	27	between	between	ADP
cana-2597	204	28	relevant	relevant	ADJ
cana-2597	204	29	and	and	CCONJ
cana-2597	204	30	irrelevant	irrelevant	ADJ
cana-2597	204	31	tweets	tweet	NOUN
cana-2597	204	32	in	in	ADP
cana-2597	204	33	the	the	DET
cana-2597	204	34	context	context	NOUN
cana-2597	204	35	of	of	ADP
cana-2597	204	36	disaster	disaster	NOUN
cana-2597	204	37	management	management	NOUN
cana-2597	204	38	.	.	PUNCT
cana-2597	205	1	metrics	metric	NOUN
cana-2597	205	2	efficientnet	efficientnet	ADJ
cana-2597	205	3	-	-	PUNCT
cana-2597	205	4	bert	bert	NOUN
cana-2597	205	5	resnet	resnet	NOUN
cana-2597	205	6	-	-	PUNCT
cana-2597	205	7	bert	bert	NOUN
cana-2597	205	8	accuracy	accuracy	NOUN
cana-2597	205	9	0.9366	0.9366	NUM
cana-2597	205	10	0.9250	0.9250	NUM
cana-2597	205	11	precision	precision	NOUN
cana-2597	205	12	0.9385	0.9385	NUM
cana-2597	205	13	0.9200	0.9200	NUM
cana-2597	205	14	recall	recall	VERB
cana-2597	205	15	0.9366	0.9366	NUM
cana-2597	205	16	0.9180	0.9180	NUM
cana-2597	205	17	f1	f1	NOUN
cana-2597	205	18	score	score	NOUN
cana-2597	205	19	0.9368	0.9368	NUM
cana-2597	205	20	0.9190	0.9190	NUM
cana-2597	205	21	table	table	NOUN
cana-2597	205	22	2	2	NUM
cana-2597	205	23	.	.	PUNCT
cana-2597	205	24	summary	summary	NOUN
cana-2597	205	25	of	of	ADP
cana-2597	205	26	performance	performance	NOUN
cana-2597	205	27	metrics	metric	NOUN
cana-2597	205	28	5.1	5.1	NUM
cana-2597	205	29	.	.	PUNCT
cana-2597	206	1	model	model	NOUN
cana-2597	206	2	performance	performance	NOUN
cana-2597	206	3	the	the	DET
cana-2597	206	4	efficientnet	efficientnet	ADJ
cana-2597	206	5	-	-	PUNCT
cana-2597	206	6	bert	bert	NOUN
cana-2597	206	7	model	model	NOUN
cana-2597	206	8	achieved	achieve	VERB
cana-2597	206	9	an	an	DET
cana-2597	206	10	impressive	impressive	ADJ
cana-2597	206	11	accuracy	accuracy	NOUN
cana-2597	206	12	of	of	ADP
cana-2597	206	13	93.66	93.66	NUM
cana-2597	206	14	%	%	NOUN
cana-2597	206	15	,	,	PUNCT
cana-2597	206	16	indicating	indicate	VERB
cana-2597	206	17	that	that	SCONJ
cana-2597	206	18	it	it	PRON
cana-2597	206	19	correctly	correctly	ADV
cana-2597	206	20	classified	classify	VERB
cana-2597	206	21	a	a	DET
cana-2597	206	22	high	high	ADJ
cana-2597	206	23	percentage	percentage	NOUN
cana-2597	206	24	of	of	ADP
cana-2597	206	25	tweets	tweet	NOUN
cana-2597	206	26	across	across	ADP
cana-2597	206	27	the	the	DET
cana-2597	206	28	flood	flood	NOUN
cana-2597	206	29	,	,	PUNCT
cana-2597	206	30	fire	fire	NOUN
cana-2597	206	31	,	,	PUNCT
cana-2597	206	32	and	and	CCONJ
cana-2597	206	33	non	non	ADJ
cana-2597	206	34	-	-	ADJ
cana-2597	206	35	damage	damage	ADJ
cana-2597	206	36	categories	category	NOUN
cana-2597	206	37	.	.	PUNCT
cana-2597	207	1	this	this	PRON
cana-2597	207	2	suggests	suggest	VERB
cana-2597	207	3	that	that	SCONJ
cana-2597	207	4	the	the	DET
cana-2597	207	5	model	model	NOUN
cana-2597	207	6	has	have	AUX
cana-2597	207	7	learned	learn	VERB
cana-2597	207	8	to	to	PART
cana-2597	207	9	differentiate	differentiate	VERB
cana-2597	207	10	effectively	effectively	ADV
cana-2597	207	11	between	between	ADP
cana-2597	207	12	relevant	relevant	ADJ
cana-2597	207	13	and	and	CCONJ
cana-2597	207	14	irrelevant	irrelevant	ADJ
cana-2597	207	15	tweets	tweet	NOUN
cana-2597	207	16	in	in	ADP
cana-2597	207	17	the	the	DET
cana-2597	207	18	context	context	NOUN
cana-2597	207	19	of	of	ADP
cana-2597	207	20	disaster	disaster	NOUN
cana-2597	207	21	management	management	NOUN
cana-2597	207	22	.	.	PUNCT
cana-2597	208	1	in	in	ADP
cana-2597	208	2	contrast	contrast	NOUN
cana-2597	208	3	,	,	PUNCT
cana-2597	208	4	the	the	DET
cana-2597	208	5	resnet	resnet	NOUN
cana-2597	208	6	-	-	PUNCT
cana-2597	208	7	bert	bert	NOUN
cana-2597	208	8	model	model	NOUN
cana-2597	208	9	performed	perform	VERB
cana-2597	208	10	well	well	ADV
cana-2597	208	11	with	with	ADP
cana-2597	208	12	an	an	DET
cana-2597	208	13	accuracy	accuracy	NOUN
cana-2597	208	14	of	of	ADP
cana-2597	208	15	92.50	92.50	NUM
cana-2597	208	16	%	%	NOUN
cana-2597	208	17	,	,	PUNCT
cana-2597	208	18	but	but	CCONJ
cana-2597	208	19	demonstrated	demonstrate	VERB
cana-2597	208	20	slightly	slightly	ADV
cana-2597	208	21	lower	low	ADJ
cana-2597	208	22	scores	score	NOUN
cana-2597	208	23	across	across	ADP
cana-2597	208	24	all	all	DET
cana-2597	208	25	metrics	metric	NOUN
cana-2597	208	26	compared	compare	VERB
cana-2597	208	27	to	to	ADP
cana-2597	208	28	communications	communication	NOUN
cana-2597	208	29	on	on	ADP
cana-2597	208	30	applied	apply	VERB
cana-2597	208	31	nonlinear	nonlinear	ADJ
cana-2597	208	32	analysis	analysis	NOUN
cana-2597	208	33	issn	issn	NOUN
cana-2597	208	34	:	:	PUNCT
cana-2597	208	35	1074	1074	NUM
cana-2597	208	36	-	-	PUNCT
cana-2597	208	37	133x	133x	NUM
cana-2597	208	38	vol	vol	NOUN
cana-2597	208	39	32	32	NUM
cana-2597	208	40	no	no	NOUN
cana-2597	208	41	.	.	PUNCT
cana-2597	209	1	3s	3s	NUM
cana-2597	209	2	(	(	PUNCT
cana-2597	209	3	2025	2025	NUM
cana-2597	209	4	)	)	PUNCT
cana-2597	209	5	210	210	NUM
cana-2597	209	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2597	209	7	efficientnet	efficientnet	NOUN
cana-2597	209	8	-	-	PUNCT
cana-2597	209	9	bert	bert	NOUN
cana-2597	209	10	.	.	PUNCT
cana-2597	210	1	this	this	PRON
cana-2597	210	2	indicates	indicate	VERB
cana-2597	210	3	that	that	SCONJ
cana-2597	210	4	,	,	PUNCT
cana-2597	210	5	while	while	SCONJ
cana-2597	210	6	resnet	resnet	NOUN
cana-2597	210	7	-	-	PUNCT
cana-2597	210	8	bert	bert	NOUN
cana-2597	210	9	is	be	AUX
cana-2597	210	10	a	a	DET
cana-2597	210	11	strong	strong	ADJ
cana-2597	210	12	model	model	NOUN
cana-2597	210	13	,	,	PUNCT
cana-2597	210	14	it	it	PRON
cana-2597	210	15	may	may	AUX
cana-2597	210	16	not	not	PART
cana-2597	210	17	be	be	AUX
cana-2597	210	18	as	as	ADV
cana-2597	210	19	finely	finely	ADV
cana-2597	210	20	tuned	tune	VERB
cana-2597	210	21	for	for	ADP
cana-2597	210	22	this	this	DET
cana-2597	210	23	specific	specific	ADJ
cana-2597	210	24	classification	classification	NOUN
cana-2597	210	25	task	task	NOUN
cana-2597	210	26	.	.	PUNCT
cana-2597	211	1	5.2	5.2	NUM
cana-2597	211	2	.	.	PUNCT
cana-2597	211	3	precision	precision	NOUN
cana-2597	211	4	and	and	CCONJ
cana-2597	211	5	recall	recall	VERB
cana-2597	211	6	efficientnet	efficientnet	ADJ
cana-2597	211	7	-	-	PUNCT
cana-2597	211	8	bert	bert	NOUN
cana-2597	211	9	's	's	PART
cana-2597	211	10	precision	precision	NOUN
cana-2597	211	11	of	of	ADP
cana-2597	211	12	0.9385	0.9385	NUM
cana-2597	211	13	signifies	signify	VERB
cana-2597	211	14	a	a	DET
cana-2597	211	15	robust	robust	ADJ
cana-2597	211	16	ability	ability	NOUN
cana-2597	211	17	to	to	PART
cana-2597	211	18	minimize	minimize	VERB
cana-2597	211	19	false	false	ADJ
cana-2597	211	20	positives	positive	NOUN
cana-2597	211	21	,	,	PUNCT
cana-2597	211	22	meaning	mean	VERB
cana-2597	211	23	that	that	SCONJ
cana-2597	211	24	when	when	SCONJ
cana-2597	211	25	it	it	PRON
cana-2597	211	26	predicts	predict	VERB
cana-2597	211	27	a	a	DET
cana-2597	211	28	tweet	tweet	NOUN
cana-2597	211	29	as	as	ADP
cana-2597	211	30	relevant	relevant	ADJ
cana-2597	211	31	to	to	ADP
cana-2597	211	32	disasters	disaster	NOUN
cana-2597	211	33	,	,	PUNCT
cana-2597	211	34	it	it	PRON
cana-2597	211	35	is	be	AUX
cana-2597	211	36	correct	correct	ADJ
cana-2597	211	37	93.85	93.85	NUM
cana-2597	211	38	%	%	NOUN
cana-2597	211	39	of	of	ADP
cana-2597	211	40	the	the	DET
cana-2597	211	41	time	time	NOUN
cana-2597	211	42	.	.	PUNCT
cana-2597	212	1	this	this	PRON
cana-2597	212	2	is	be	AUX
cana-2597	212	3	crucial	crucial	ADJ
cana-2597	212	4	in	in	ADP
cana-2597	212	5	disaster	disaster	NOUN
cana-2597	212	6	response	response	NOUN
cana-2597	212	7	contexts	contexts	NOUN
cana-2597	212	8	,	,	PUNCT
cana-2597	212	9	where	where	SCONJ
cana-2597	212	10	false	false	ADJ
cana-2597	212	11	alarms	alarm	NOUN
cana-2597	212	12	can	can	AUX
cana-2597	212	13	lead	lead	VERB
cana-2597	212	14	to	to	ADP
cana-2597	212	15	unnecessary	unnecessary	ADJ
cana-2597	212	16	resource	resource	NOUN
cana-2597	212	17	allocation	allocation	NOUN
cana-2597	212	18	.	.	PUNCT
cana-2597	213	1	the	the	DET
cana-2597	213	2	recall	recall	NOUN
cana-2597	213	3	score	score	NOUN
cana-2597	213	4	of	of	ADP
cana-2597	213	5	0.9366	0.9366	NUM
cana-2597	213	6	highlights	highlight	NOUN
cana-2597	213	7	that	that	PRON
cana-2597	213	8	the	the	DET
cana-2597	213	9	model	model	NOUN
cana-2597	213	10	effectively	effectively	ADV
cana-2597	213	11	captures	capture	VERB
cana-2597	213	12	the	the	DET
cana-2597	213	13	majority	majority	NOUN
cana-2597	213	14	of	of	ADP
cana-2597	213	15	actual	actual	ADJ
cana-2597	213	16	disaster	disaster	NOUN
cana-2597	213	17	-	-	PUNCT
cana-2597	213	18	related	relate	VERB
cana-2597	213	19	tweets	tweet	NOUN
cana-2597	213	20	.	.	PUNCT
cana-2597	214	1	this	this	DET
cana-2597	214	2	balance	balance	NOUN
cana-2597	214	3	between	between	ADP
cana-2597	214	4	precision	precision	NOUN
cana-2597	214	5	and	and	CCONJ
cana-2597	214	6	recall	recall	NOUN
cana-2597	214	7	is	be	AUX
cana-2597	214	8	vital	vital	ADJ
cana-2597	214	9	for	for	ADP
cana-2597	214	10	ensuring	ensure	VERB
cana-2597	214	11	that	that	SCONJ
cana-2597	214	12	significant	significant	ADJ
cana-2597	214	13	tweets	tweet	NOUN
cana-2597	214	14	are	be	AUX
cana-2597	214	15	not	not	PART
cana-2597	214	16	missed	miss	VERB
cana-2597	214	17	,	,	PUNCT
cana-2597	214	18	which	which	PRON
cana-2597	214	19	can	can	AUX
cana-2597	214	20	be	be	AUX
cana-2597	214	21	critical	critical	ADJ
cana-2597	214	22	during	during	ADP
cana-2597	214	23	emergencies	emergency	NOUN
cana-2597	214	24	.	.	PUNCT
cana-2597	215	1	5.3	5.3	NUM
cana-2597	215	2	.	.	PUNCT
cana-2597	215	3	f1	f1	PROPN
cana-2597	215	4	score	score	NOUN
cana-2597	215	5	the	the	DET
cana-2597	215	6	f1	f1	ADJ
cana-2597	215	7	score	score	NOUN
cana-2597	215	8	of	of	ADP
cana-2597	215	9	0.9368	0.9368	NUM
cana-2597	215	10	for	for	ADP
cana-2597	215	11	efficientnet	efficientnet	NOUN
cana-2597	215	12	-	-	PUNCT
cana-2597	215	13	bert	bert	NOUN
cana-2597	215	14	reflects	reflect	VERB
cana-2597	215	15	a	a	DET
cana-2597	215	16	harmonious	harmonious	ADJ
cana-2597	215	17	trade	trade	NOUN
cana-2597	215	18	-	-	PUNCT
cana-2597	215	19	off	off	NOUN
cana-2597	215	20	between	between	ADP
cana-2597	215	21	precision	precision	NOUN
cana-2597	215	22	and	and	CCONJ
cana-2597	215	23	recall	recall	NOUN
cana-2597	215	24	.	.	PUNCT
cana-2597	216	1	it	it	PRON
cana-2597	216	2	indicates	indicate	VERB
cana-2597	216	3	that	that	SCONJ
cana-2597	216	4	the	the	DET
cana-2597	216	5	model	model	NOUN
cana-2597	216	6	is	be	AUX
cana-2597	216	7	robust	robust	ADJ
cana-2597	216	8	in	in	ADP
cana-2597	216	9	classifying	classify	VERB
cana-2597	216	10	tweets	tweet	NOUN
cana-2597	216	11	without	without	ADP
cana-2597	216	12	being	be	AUX
cana-2597	216	13	overly	overly	ADV
cana-2597	216	14	sensitive	sensitive	ADJ
cana-2597	216	15	to	to	ADP
cana-2597	216	16	either	either	DET
cana-2597	216	17	false	false	ADJ
cana-2597	216	18	positives	positive	NOUN
cana-2597	216	19	or	or	CCONJ
cana-2597	216	20	false	false	ADJ
cana-2597	216	21	negatives	negative	NOUN
cana-2597	216	22	.	.	PUNCT
cana-2597	217	1	on	on	ADP
cana-2597	217	2	the	the	DET
cana-2597	217	3	other	other	ADJ
cana-2597	217	4	hand	hand	NOUN
cana-2597	217	5	,	,	PUNCT
cana-2597	217	6	the	the	DET
cana-2597	217	7	f1	f1	ADJ
cana-2597	217	8	score	score	NOUN
cana-2597	217	9	of	of	ADP
cana-2597	217	10	0.9190	0.9190	NUM
cana-2597	217	11	for	for	ADP
cana-2597	217	12	resnet	resnet	NOUN
cana-2597	217	13	-	-	PUNCT
cana-2597	217	14	bert	bert	NOUN
cana-2597	217	15	,	,	PUNCT
cana-2597	217	16	while	while	SCONJ
cana-2597	217	17	commendable	commendable	ADJ
cana-2597	217	18	,	,	PUNCT
cana-2597	217	19	suggests	suggest	VERB
cana-2597	217	20	that	that	SCONJ
cana-2597	217	21	there	there	PRON
cana-2597	217	22	may	may	AUX
cana-2597	217	23	be	be	AUX
cana-2597	217	24	a	a	DET
cana-2597	217	25	slight	slight	ADJ
cana-2597	217	26	trade	trade	NOUN
cana-2597	217	27	-	-	PUNCT
cana-2597	217	28	off	off	NOUN
cana-2597	217	29	in	in	ADP
cana-2597	217	30	performance	performance	NOUN
cana-2597	217	31	.	.	PUNCT
cana-2597	218	1	this	this	PRON
cana-2597	218	2	potentially	potentially	ADV
cana-2597	218	3	makes	make	VERB
cana-2597	218	4	efficientnet	efficientnet	NOUN
cana-2597	218	5	-	-	PUNCT
cana-2597	218	6	bert	bert	NOUN
cana-2597	218	7	the	the	DET
cana-2597	218	8	more	more	ADV
cana-2597	218	9	reliable	reliable	ADJ
cana-2597	218	10	option	option	NOUN
cana-2597	218	11	in	in	ADP
cana-2597	218	12	scenarios	scenario	NOUN
cana-2597	218	13	where	where	SCONJ
cana-2597	218	14	both	both	DET
cana-2597	218	15	precision	precision	NOUN
cana-2597	218	16	and	and	CCONJ
cana-2597	218	17	recall	recall	NOUN
cana-2597	218	18	are	be	AUX
cana-2597	218	19	equally	equally	ADV
cana-2597	218	20	important	important	ADJ
cana-2597	218	21	.	.	PUNCT
cana-2597	219	1	vi	vi	X
cana-2597	219	2	.	.	NOUN
cana-2597	219	3	conclusion	conclusion	NOUN
cana-2597	219	4	and	and	CCONJ
cana-2597	219	5	future	future	ADJ
cana-2597	219	6	work	work	NOUN
cana-2597	219	7	this	this	DET
cana-2597	219	8	research	research	NOUN
cana-2597	219	9	paper	paper	NOUN
cana-2597	219	10	presents	present	VERB
cana-2597	219	11	the	the	DET
cana-2597	219	12	development	development	NOUN
cana-2597	219	13	of	of	ADP
cana-2597	219	14	an	an	DET
cana-2597	219	15	ai	ai	NOUN
cana-2597	219	16	-	-	PUNCT
cana-2597	219	17	driven	drive	VERB
cana-2597	219	18	disaster	disaster	NOUN
cana-2597	219	19	management	management	NOUN
cana-2597	219	20	system	system	NOUN
cana-2597	219	21	that	that	PRON
cana-2597	219	22	harnesses	harness	VERB
cana-2597	219	23	real	real	ADJ
cana-2597	219	24	-	-	PUNCT
cana-2597	219	25	time	time	NOUN
cana-2597	219	26	social	social	ADJ
cana-2597	219	27	media	medium	NOUN
cana-2597	219	28	data	datum	NOUN
cana-2597	219	29	to	to	PART
cana-2597	219	30	enhance	enhance	VERB
cana-2597	219	31	disaster	disaster	NOUN
cana-2597	219	32	response	response	NOUN
cana-2597	219	33	.	.	PUNCT
cana-2597	220	1	by	by	ADP
cana-2597	220	2	integrating	integrate	VERB
cana-2597	220	3	data	datum	NOUN
cana-2597	220	4	from	from	ADP
cana-2597	220	5	platforms	platform	NOUN
cana-2597	220	6	like	like	ADP
cana-2597	220	7	twitter	twitter	NOUN
cana-2597	220	8	,	,	PUNCT
cana-2597	220	9	instagram	instagram	PROPN
cana-2597	220	10	,	,	PUNCT
cana-2597	220	11	and	and	CCONJ
cana-2597	220	12	facebook	facebook	NOUN
cana-2597	220	13	,	,	PUNCT
cana-2597	220	14	the	the	DET
cana-2597	220	15	system	system	NOUN
cana-2597	220	16	collects	collect	VERB
cana-2597	220	17	both	both	DET
cana-2597	220	18	textual	textual	ADJ
cana-2597	220	19	and	and	CCONJ
cana-2597	220	20	image	image	NOUN
cana-2597	220	21	-	-	PUNCT
cana-2597	220	22	based	base	VERB
cana-2597	220	23	information	information	NOUN
cana-2597	220	24	critical	critical	ADJ
cana-2597	220	25	for	for	ADP
cana-2597	220	26	situational	situational	ADJ
cana-2597	220	27	awareness	awareness	NOUN
cana-2597	220	28	during	during	ADP
cana-2597	220	29	disasters	disaster	NOUN
cana-2597	220	30	.	.	PUNCT
cana-2597	221	1	the	the	DET
cana-2597	221	2	system	system	NOUN
cana-2597	221	3	employs	employ	VERB
cana-2597	221	4	advanced	advanced	ADJ
cana-2597	221	5	machine	machine	NOUN
cana-2597	221	6	learning	learning	NOUN
cana-2597	221	7	models	model	NOUN
cana-2597	221	8	,	,	PUNCT
cana-2597	221	9	such	such	ADJ
cana-2597	221	10	as	as	ADP
cana-2597	221	11	bert	bert	PROPN
cana-2597	221	12	and	and	CCONJ
cana-2597	221	13	xlnet	xlnet	PROPN
cana-2597	221	14	for	for	ADP
cana-2597	221	15	text	text	NOUN
cana-2597	221	16	analysis	analysis	NOUN
cana-2597	221	17	,	,	PUNCT
cana-2597	221	18	and	and	CCONJ
cana-2597	221	19	efficientnet	efficientnet	NOUN
cana-2597	221	20	and	and	CCONJ
cana-2597	221	21	resnet50	resnet50	NOUN
cana-2597	221	22	for	for	ADP
cana-2597	221	23	image	image	NOUN
cana-2597	221	24	classification	classification	NOUN
cana-2597	221	25	,	,	PUNCT
cana-2597	221	26	to	to	PART
cana-2597	221	27	accurately	accurately	ADV
cana-2597	221	28	categorize	categorize	VERB
cana-2597	221	29	and	and	CCONJ
cana-2597	221	30	assess	assess	VERB
cana-2597	221	31	the	the	DET
cana-2597	221	32	severity	severity	NOUN
cana-2597	221	33	of	of	ADP
cana-2597	221	34	disaster	disaster	NOUN
cana-2597	221	35	events	event	NOUN
cana-2597	221	36	.	.	PUNCT
cana-2597	222	1	this	this	DET
cana-2597	222	2	multimodal	multimodal	ADJ
cana-2597	222	3	approach	approach	NOUN
cana-2597	222	4	significantly	significantly	ADV
cana-2597	222	5	improves	improve	VERB
cana-2597	222	6	the	the	DET
cana-2597	222	7	precision	precision	NOUN
cana-2597	222	8	of	of	ADP
cana-2597	222	9	disaster	disaster	NOUN
cana-2597	222	10	detection	detection	NOUN
cana-2597	222	11	,	,	PUNCT
cana-2597	222	12	providing	provide	VERB
cana-2597	222	13	actionable	actionable	ADJ
cana-2597	222	14	insights	insight	NOUN
cana-2597	222	15	that	that	PRON
cana-2597	222	16	help	help	VERB
cana-2597	222	17	responders	responder	NOUN
cana-2597	222	18	prioritize	prioritize	VERB
cana-2597	222	19	and	and	CCONJ
cana-2597	222	20	allocate	allocate	VERB
cana-2597	222	21	resources	resource	NOUN
cana-2597	222	22	effectively	effectively	ADV
cana-2597	222	23	.	.	PUNCT
cana-2597	223	1	the	the	DET
cana-2597	223	2	system	system	NOUN
cana-2597	223	3	’s	’s	PART
cana-2597	223	4	real	real	ADJ
cana-2597	223	5	-	-	PUNCT
cana-2597	223	6	time	time	NOUN
cana-2597	223	7	data	datum	NOUN
cana-2597	223	8	ingestion	ingestion	NOUN
cana-2597	223	9	and	and	CCONJ
cana-2597	223	10	processing	processing	NOUN
cana-2597	223	11	pipeline	pipeline	NOUN
cana-2597	223	12	,	,	PUNCT
cana-2597	223	13	built	build	VERB
cana-2597	223	14	using	use	VERB
cana-2597	223	15	apache	apache	PROPN
cana-2597	223	16	kafka	kafka	PROPN
cana-2597	223	17	and	and	CCONJ
cana-2597	223	18	apache	apache	PROPN
cana-2597	223	19	spark	spark	NOUN
cana-2597	223	20	,	,	PUNCT
cana-2597	223	21	ensures	ensure	VERB
cana-2597	223	22	scalability	scalability	NOUN
cana-2597	223	23	and	and	CCONJ
cana-2597	223	24	efficiency	efficiency	NOUN
cana-2597	223	25	even	even	ADV
cana-2597	223	26	during	during	ADP
cana-2597	223	27	high	high	ADJ
cana-2597	223	28	-	-	PUNCT
cana-2597	223	29	traffic	traffic	NOUN
cana-2597	223	30	disaster	disaster	NOUN
cana-2597	223	31	periods	period	NOUN
cana-2597	223	32	.	.	PUNCT
cana-2597	224	1	by	by	ADP
cana-2597	224	2	fusing	fuse	VERB
cana-2597	224	3	textual	textual	ADJ
cana-2597	224	4	and	and	CCONJ
cana-2597	224	5	visual	visual	ADJ
cana-2597	224	6	data	datum	NOUN
cana-2597	224	7	,	,	PUNCT
cana-2597	224	8	the	the	DET
cana-2597	224	9	system	system	NOUN
cana-2597	224	10	achieves	achieve	VERB
cana-2597	224	11	high	high	ADJ
cana-2597	224	12	classification	classification	NOUN
cana-2597	224	13	accuracy	accuracy	NOUN
cana-2597	224	14	,	,	PUNCT
cana-2597	224	15	ensuring	ensure	VERB
cana-2597	224	16	that	that	SCONJ
cana-2597	224	17	relevant	relevant	ADJ
cana-2597	224	18	disaster	disaster	NOUN
cana-2597	224	19	information	information	NOUN
cana-2597	224	20	is	be	AUX
cana-2597	224	21	quickly	quickly	ADV
cana-2597	224	22	identified	identify	VERB
cana-2597	224	23	.	.	PUNCT
cana-2597	225	1	the	the	DET
cana-2597	225	2	architecture	architecture	NOUN
cana-2597	225	3	’s	’s	PART
cana-2597	225	4	scalability	scalability	NOUN
cana-2597	225	5	allows	allow	VERB
cana-2597	225	6	for	for	ADP
cana-2597	225	7	processing	process	VERB
cana-2597	225	8	large	large	ADJ
cana-2597	225	9	volumes	volume	NOUN
cana-2597	225	10	of	of	ADP
cana-2597	225	11	social	social	ADJ
cana-2597	225	12	media	medium	NOUN
cana-2597	225	13	data	datum	NOUN
cana-2597	225	14	without	without	ADP
cana-2597	225	15	bottlenecks	bottleneck	NOUN
cana-2597	225	16	,	,	PUNCT
cana-2597	225	17	making	make	VERB
cana-2597	225	18	it	it	PRON
cana-2597	225	19	a	a	DET
cana-2597	225	20	robust	robust	ADJ
cana-2597	225	21	solution	solution	NOUN
cana-2597	225	22	for	for	ADP
cana-2597	225	23	managing	manage	VERB
cana-2597	225	24	disaster	disaster	NOUN
cana-2597	225	25	-	-	PUNCT
cana-2597	225	26	related	relate	VERB
cana-2597	225	27	data	datum	NOUN
cana-2597	225	28	in	in	ADP
cana-2597	225	29	real	real	ADJ
cana-2597	225	30	time	time	NOUN
cana-2597	225	31	.	.	PUNCT
cana-2597	226	1	ultimately	ultimately	ADV
cana-2597	226	2	,	,	PUNCT
cana-2597	226	3	this	this	DET
cana-2597	226	4	system	system	NOUN
cana-2597	226	5	demonstrates	demonstrate	VERB
cana-2597	226	6	the	the	DET
cana-2597	226	7	potential	potential	NOUN
cana-2597	226	8	of	of	ADP
cana-2597	226	9	social	social	ADJ
cana-2597	226	10	media	medium	NOUN
cana-2597	226	11	analytics	analytic	NOUN
cana-2597	226	12	in	in	ADP
cana-2597	226	13	improving	improve	VERB
cana-2597	226	14	disaster	disaster	NOUN
cana-2597	226	15	response	response	NOUN
cana-2597	226	16	,	,	PUNCT
cana-2597	226	17	helping	help	VERB
cana-2597	226	18	save	save	VERB
cana-2597	226	19	lives	life	NOUN
cana-2597	226	20	by	by	ADP
cana-2597	226	21	providing	provide	VERB
cana-2597	226	22	timely	timely	ADJ
cana-2597	226	23	and	and	CCONJ
cana-2597	226	24	actionable	actionable	ADJ
cana-2597	226	25	insights	insight	NOUN
cana-2597	226	26	.	.	PUNCT
cana-2597	227	1	looking	look	VERB
cana-2597	227	2	forward	forward	ADV
cana-2597	227	3	,	,	PUNCT
cana-2597	227	4	the	the	DET
cana-2597	227	5	system	system	NOUN
cana-2597	227	6	will	will	AUX
cana-2597	227	7	be	be	AUX
cana-2597	227	8	expanded	expand	VERB
cana-2597	227	9	into	into	ADP
cana-2597	227	10	a	a	DET
cana-2597	227	11	web	web	NOUN
cana-2597	227	12	portal	portal	NOUN
cana-2597	227	13	aimed	aim	VERB
cana-2597	227	14	at	at	ADP
cana-2597	227	15	providing	provide	VERB
cana-2597	227	16	disaster	disaster	NOUN
cana-2597	227	17	response	response	NOUN
cana-2597	227	18	agencies	agency	NOUN
cana-2597	227	19	in	in	ADP
cana-2597	227	20	india	india	PROPN
cana-2597	227	21	with	with	ADP
cana-2597	227	22	a	a	DET
cana-2597	227	23	powerful	powerful	ADJ
cana-2597	227	24	tool	tool	NOUN
cana-2597	227	25	for	for	ADP
cana-2597	227	26	analysing	analyse	VERB
cana-2597	227	27	and	and	CCONJ
cana-2597	227	28	categorizing	categorize	VERB
cana-2597	227	29	disaster	disaster	NOUN
cana-2597	227	30	-	-	PUNCT
cana-2597	227	31	related	relate	VERB
cana-2597	227	32	data	datum	NOUN
cana-2597	227	33	from	from	ADP
cana-2597	227	34	significant	significant	ADJ
cana-2597	227	35	repositories	repository	NOUN
cana-2597	227	36	,	,	PUNCT
cana-2597	227	37	particularly	particularly	ADV
cana-2597	227	38	social	social	ADJ
cana-2597	227	39	media	medium	NOUN
cana-2597	227	40	.	.	PUNCT
cana-2597	228	1	this	this	DET
cana-2597	228	2	portal	portal	NOUN
cana-2597	228	3	will	will	AUX
cana-2597	228	4	automate	automate	VERB
cana-2597	228	5	the	the	DET
cana-2597	228	6	collection	collection	NOUN
cana-2597	228	7	,	,	PUNCT
cana-2597	228	8	processing	processing	NOUN
cana-2597	228	9	,	,	PUNCT
cana-2597	228	10	and	and	CCONJ
cana-2597	228	11	analysis	analysis	NOUN
cana-2597	228	12	of	of	ADP
cana-2597	228	13	data	datum	NOUN
cana-2597	228	14	,	,	PUNCT
cana-2597	228	15	delivering	deliver	VERB
cana-2597	228	16	real	real	ADJ
cana-2597	228	17	-	-	PUNCT
cana-2597	228	18	time	time	NOUN
cana-2597	228	19	actionable	actionable	ADJ
cana-2597	228	20	insights	insight	NOUN
cana-2597	228	21	that	that	PRON
cana-2597	228	22	allow	allow	VERB
cana-2597	228	23	agencies	agency	NOUN
cana-2597	228	24	to	to	PART
cana-2597	228	25	respond	respond	VERB
cana-2597	228	26	more	more	ADV
cana-2597	228	27	quickly	quickly	ADV
cana-2597	228	28	and	and	CCONJ
cana-2597	228	29	efficiently	efficiently	ADV
cana-2597	228	30	.	.	PUNCT
cana-2597	229	1	by	by	ADP
cana-2597	229	2	reducing	reduce	VERB
cana-2597	229	3	response	response	NOUN
cana-2597	229	4	times	time	NOUN
cana-2597	229	5	,	,	PUNCT
cana-2597	229	6	the	the	DET
cana-2597	229	7	portal	portal	NOUN
cana-2597	229	8	will	will	AUX
cana-2597	229	9	help	help	VERB
cana-2597	229	10	agencies	agency	NOUN
cana-2597	229	11	save	save	VERB
cana-2597	229	12	valuable	valuable	ADJ
cana-2597	229	13	time	time	NOUN
cana-2597	229	14	,	,	PUNCT
cana-2597	229	15	money	money	NOUN
cana-2597	229	16	,	,	PUNCT
cana-2597	229	17	and	and	CCONJ
cana-2597	229	18	resources	resource	NOUN
cana-2597	229	19	,	,	PUNCT
cana-2597	229	20	while	while	SCONJ
cana-2597	229	21	also	also	ADV
cana-2597	229	22	enhancing	enhance	VERB
cana-2597	229	23	the	the	DET
cana-2597	229	24	accuracy	accuracy	NOUN
cana-2597	229	25	and	and	CCONJ
cana-2597	229	26	relevance	relevance	NOUN
cana-2597	229	27	of	of	ADP
cana-2597	229	28	the	the	DET
cana-2597	229	29	data	datum	NOUN
cana-2597	229	30	used	use	VERB
cana-2597	229	31	for	for	ADP
cana-2597	229	32	decisionmaking	decisionmake	VERB
cana-2597	229	33	.	.	PUNCT
cana-2597	230	1	this	this	DET
cana-2597	230	2	future	future	ADJ
cana-2597	230	3	development	development	NOUN
cana-2597	230	4	will	will	AUX
cana-2597	230	5	streamline	streamline	VERB
cana-2597	230	6	the	the	DET
cana-2597	230	7	process	process	NOUN
cana-2597	230	8	of	of	ADP
cana-2597	230	9	disaster	disaster	NOUN
cana-2597	230	10	monitoring	monitoring	NOUN
cana-2597	230	11	and	and	CCONJ
cana-2597	230	12	management	management	NOUN
cana-2597	230	13	,	,	PUNCT
cana-2597	230	14	communications	communication	NOUN
cana-2597	230	15	on	on	ADP
cana-2597	230	16	applied	apply	VERB
cana-2597	230	17	nonlinear	nonlinear	ADJ
cana-2597	230	18	analysis	analysis	NOUN
cana-2597	230	19	issn	issn	NOUN
cana-2597	230	20	:	:	PUNCT
cana-2597	230	21	1074	1074	NUM
cana-2597	230	22	-	-	PUNCT
cana-2597	230	23	133x	133x	NUM
cana-2597	230	24	vol	vol	NOUN
cana-2597	230	25	32	32	NUM
cana-2597	230	26	no	no	NOUN
cana-2597	230	27	.	.	PUNCT
cana-2597	231	1	3s	3s	NUM
cana-2597	231	2	(	(	PUNCT
cana-2597	231	3	2025	2025	NUM
cana-2597	231	4	)	)	PUNCT
cana-2597	231	5	211	211	NUM
cana-2597	231	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-2597	231	7	offering	offer	VERB
cana-2597	231	8	a	a	DET
cana-2597	231	9	centralized	centralized	ADJ
cana-2597	231	10	platform	platform	NOUN
cana-2597	231	11	where	where	SCONJ
cana-2597	231	12	agencies	agency	NOUN
cana-2597	231	13	can	can	AUX
cana-2597	231	14	access	access	VERB
cana-2597	231	15	critical	critical	ADJ
cana-2597	231	16	information	information	NOUN
cana-2597	231	17	,	,	PUNCT
cana-2597	231	18	ultimately	ultimately	ADV
cana-2597	231	19	improving	improve	VERB
cana-2597	231	20	disaster	disaster	NOUN
cana-2597	231	21	preparedness	preparedness	NOUN
cana-2597	231	22	and	and	CCONJ
cana-2597	231	23	response	response	NOUN
cana-2597	231	24	.	.	PUNCT
cana-2597	232	1	references	reference	NOUN
cana-2597	232	2	[	[	X
cana-2597	232	3	1	1	NUM
cana-2597	232	4	]	]	SYM
cana-2597	232	5	wiegmann	wiegmann	NOUN
cana-2597	232	6	,	,	PUNCT
cana-2597	232	7	m.	m.	NOUN
cana-2597	232	8	,	,	PUNCT
cana-2597	232	9	kersten	kersten	PROPN
cana-2597	232	10	,	,	PUNCT
cana-2597	232	11	j.	j.	PROPN
cana-2597	232	12	,	,	PUNCT
cana-2597	232	13	senaratne	senaratne	PROPN
cana-2597	232	14	,	,	PUNCT
cana-2597	232	15	h.	h.	NOUN
cana-2597	232	16	,	,	PUNCT
cana-2597	232	17	potthast	potthast	NOUN
cana-2597	232	18	,	,	PUNCT
cana-2597	232	19	m.	m.	NOUN
cana-2597	232	20	,	,	PUNCT
cana-2597	232	21	klan	klan	PROPN
cana-2597	232	22	,	,	PUNCT
cana-2597	232	23	f.	f.	PROPN
cana-2597	232	24	,	,	PUNCT
cana-2597	232	25	and	and	CCONJ
cana-2597	232	26	stein	stein	PROPN
cana-2597	232	27	,	,	PUNCT
cana-2597	232	28	b.	b.	PROPN
cana-2597	232	29	:	:	PUNCT
cana-2597	232	30	“	"	PUNCT
cana-2597	232	31	opportunities	opportunity	NOUN
cana-2597	232	32	and	and	CCONJ
cana-2597	232	33	risks	risk	NOUN
cana-2597	232	34	of	of	ADP
cana-2597	232	35	disaster	disaster	NOUN
cana-2597	232	36	data	datum	NOUN
cana-2597	232	37	from	from	ADP
cana-2597	232	38	social	social	ADJ
cana-2597	232	39	media	medium	NOUN
cana-2597	232	40	:	:	PUNCT
cana-2597	232	41	a	a	DET
cana-2597	232	42	systematic	systematic	ADJ
cana-2597	232	43	review	review	NOUN
cana-2597	232	44	of	of	ADP
cana-2597	232	45	incident	incident	NOUN
cana-2597	232	46	information	information	NOUN
cana-2597	232	47	,	,	PUNCT
cana-2597	232	48	”	"	PUNCT
cana-2597	232	49	nat	nat	PROPN
cana-2597	232	50	.	.	PUNCT
cana-2597	233	1	hazards	hazard	VERB
cana-2597	233	2	earth	earth	PROPN
cana-2597	233	3	syst	syst	PROPN
cana-2597	233	4	.	.	PUNCT
cana-2597	234	1	sci	sci	PROPN
cana-2597	234	2	.	.	PROPN
cana-2597	234	3	,	,	PUNCT
cana-2597	234	4	21	21	NUM
cana-2597	234	5	,	,	PUNCT
cana-2597	234	6	1431–1444	1431–1444	NUM
cana-2597	234	7	,	,	PUNCT
cana-2597	234	8	https://doi.org/10.5194/nhess-21-1431-2021	https://doi.org/10.5194/nhess-21-1431-2021	NUM
cana-2597	234	9	,	,	PUNCT
cana-2597	234	10	2021	2021	NUM
cana-2597	234	11	.	.	PUNCT
cana-2597	235	1	[	[	X
cana-2597	235	2	2	2	NUM
cana-2597	235	3	]	]	PUNCT
cana-2597	235	4	a.	a.	NOUN
cana-2597	235	5	pal	pal	NOUN
cana-2597	235	6	,	,	PUNCT
cana-2597	235	7	j.	j.	PROPN
cana-2597	235	8	wang	wang	PROPN
cana-2597	235	9	,	,	PUNCT
cana-2597	235	10	y.	y.	PROPN
cana-2597	235	11	wu	wu	PROPN
cana-2597	235	12	,	,	PUNCT
cana-2597	235	13	k.	k.	PROPN
cana-2597	235	14	kant	kant	PROPN
cana-2597	235	15	,	,	PUNCT
cana-2597	235	16	z.	z.	PROPN
cana-2597	235	17	liu	liu	PROPN
cana-2597	235	18	,	,	PUNCT
cana-2597	235	19	and	and	CCONJ
cana-2597	235	20	k.	k.	PROPN
cana-2597	235	21	sato	sato	PROPN
cana-2597	235	22	,	,	PUNCT
cana-2597	235	23	"	"	PUNCT
cana-2597	235	24	social	social	ADJ
cana-2597	235	25	media	medium	NOUN
cana-2597	235	26	driven	drive	VERB
cana-2597	235	27	big	big	ADJ
cana-2597	235	28	data	datum	NOUN
cana-2597	235	29	analysis	analysis	NOUN
cana-2597	235	30	for	for	ADP
cana-2597	235	31	disaster	disaster	NOUN
cana-2597	235	32	situation	situation	NOUN
cana-2597	235	33	awareness	awareness	NOUN
cana-2597	235	34	:	:	PUNCT
cana-2597	235	35	a	a	DET
cana-2597	235	36	tutorial	tutorial	NOUN
cana-2597	235	37	,	,	PUNCT
cana-2597	235	38	"	"	PUNCT
cana-2597	235	39	ieee	ieee	NOUN
cana-2597	235	40	transactions	transaction	NOUN
cana-2597	235	41	on	on	ADP
cana-2597	235	42	big	big	ADJ
cana-2597	235	43	data	datum	NOUN
cana-2597	235	44	,	,	PUNCT
cana-2597	235	45	vol	vol	NOUN
cana-2597	235	46	.	.	PROPN
cana-2597	235	47	9	9	NUM
cana-2597	235	48	,	,	PUNCT
cana-2597	235	49	no	no	INTJ
cana-2597	235	50	.	.	NOUN
cana-2597	235	51	1	1	NUM
cana-2597	235	52	,	,	PUNCT
cana-2597	235	53	pp	pp	ADJ
cana-2597	235	54	.	.	PUNCT
cana-2597	236	1	1	1	NUM
cana-2597	236	2	-	-	SYM
cana-2597	236	3	21	21	NUM
cana-2597	236	4	,	,	PUNCT
cana-2597	236	5	feb	feb	PROPN
cana-2597	236	6	.	.	PROPN
cana-2597	236	7	2023	2023	NUM
cana-2597	236	8	,	,	PUNCT
cana-2597	236	9	doi	doi	NOUN
cana-2597	236	10	:	:	PUNCT
cana-2597	236	11	10.1109	10.1109	NUM
cana-2597	236	12	/	/	SYM
cana-2597	236	13	tbdata.2022.3158431	tbdata.2022.3158431	NOUN
cana-2597	236	14	.	.	PUNCT
cana-2597	237	1	[	[	X
cana-2597	237	2	3	3	X
cana-2597	237	3	]	]	X
cana-2597	237	4	h.	h.	PROPN
cana-2597	237	5	hao	hao	PROPN
cana-2597	237	6	and	and	CCONJ
cana-2597	237	7	y.	y.	PROPN
cana-2597	237	8	wang	wang	PROPN
cana-2597	237	9	,	,	PUNCT
cana-2597	237	10	"	"	PUNCT
cana-2597	237	11	leveraging	leverage	VERB
cana-2597	237	12	multimodal	multimodal	ADJ
cana-2597	237	13	social	social	ADJ
cana-2597	237	14	media	medium	NOUN
cana-2597	237	15	data	datum	NOUN
cana-2597	237	16	for	for	ADP
cana-2597	237	17	rapid	rapid	ADJ
cana-2597	237	18	disaster	disaster	NOUN
cana-2597	237	19	damage	damage	NOUN
cana-2597	237	20	assessment	assessment	NOUN
cana-2597	237	21	,	,	PUNCT
cana-2597	237	22	"	"	PUNCT
cana-2597	237	23	international	international	ADJ
cana-2597	237	24	journal	journal	NOUN
cana-2597	237	25	of	of	ADP
cana-2597	237	26	disaster	disaster	NOUN
cana-2597	237	27	risk	risk	NOUN
cana-2597	237	28	reduction	reduction	NOUN
cana-2597	237	29	,	,	PUNCT
cana-2597	237	30	vol	vol	NOUN
cana-2597	237	31	.	.	PROPN
cana-2597	237	32	51	51	NUM
cana-2597	237	33	,	,	PUNCT
cana-2597	237	34	2020	2020	NUM
cana-2597	237	35	,	,	PUNCT
cana-2597	237	36	art	art	NOUN
cana-2597	237	37	.	.	PUNCT
cana-2597	238	1	no	no	INTJ
cana-2597	238	2	.	.	NOUN
cana-2597	239	1	101760	101760	NUM
cana-2597	239	2	.	.	PUNCT
cana-2597	240	1	https://doi.org/10.1016/j.ijdrr.2020.101760	https://doi.org/10.1016/j.ijdrr.2020.101760	NOUN
cana-2597	241	1	[	[	X
cana-2597	241	2	4	4	X
cana-2597	241	3	]	]	X
cana-2597	241	4	jaebeom	jaebeom	VERB
cana-2597	241	5	you	you	PRON
cana-2597	241	6	,	,	PUNCT
cana-2597	241	7	kisung	kisung	PROPN
cana-2597	241	8	lee	lee	PROPN
cana-2597	241	9	,	,	PUNCT
cana-2597	241	10	hyuk	hyuk	PROPN
cana-2597	241	11	-	-	PUNCT
cana-2597	241	12	yoon	yoon	PROPN
cana-2597	241	13	kwon	kwon	PROPN
cana-2597	241	14	,	,	PUNCT
cana-2597	241	15	“	"	PUNCT
cana-2597	241	16	deepscraper	deepscraper	NOUN
cana-2597	241	17	:	:	PUNCT
cana-2597	241	18	a	a	DET
cana-2597	241	19	complete	complete	ADJ
cana-2597	241	20	and	and	CCONJ
cana-2597	241	21	efficient	efficient	ADJ
cana-2597	241	22	tweet	tweet	NOUN
cana-2597	241	23	scraping	scrape	VERB
cana-2597	241	24	method	method	NOUN
cana-2597	241	25	using	use	VERB
cana-2597	241	26	authenticated	authenticate	VERB
cana-2597	241	27	multiprocessing	multiprocessing	NOUN
cana-2597	241	28	,	,	PUNCT
cana-2597	241	29	”	"	PUNCT
cana-2597	241	30	data	data	PROPN
cana-2597	241	31	&	&	CCONJ
cana-2597	241	32	knowledge	knowledge	PROPN
cana-2597	241	33	engineering	engineering	PROPN
cana-2597	241	34	,	,	PUNCT
cana-2597	241	35	volume	volume	NOUN
cana-2597	241	36	149	149	NUM
cana-2597	241	37	,	,	PUNCT
cana-2597	241	38	2024	2024	NUM
cana-2597	241	39	,	,	PUNCT
cana-2597	241	40	102260	102260	NUM
cana-2597	241	41	,	,	PUNCT
cana-2597	241	42	issn	issn	PROPN
cana-2597	241	43	0169	0169	PROPN
cana-2597	241	44	-	-	PUNCT
cana-2597	241	45	023x	023x	PROPN
cana-2597	241	46	,	,	PUNCT
cana-2597	241	47	https://doi.org/10.1016/j.datak.2023.102260	https://doi.org/10.1016/j.datak.2023.102260	NOUN
cana-2597	241	48	.	.	PUNCT
cana-2597	242	1	[	[	X
cana-2597	242	2	5	5	NUM
cana-2597	242	3	]	]	X
cana-2597	242	4	domala	domala	NOUN
cana-2597	242	5	,	,	PUNCT
cana-2597	242	6	jayashree	jayashree	PROPN
cana-2597	242	7	&	&	CCONJ
cana-2597	242	8	dogra	dogra	PROPN
cana-2597	242	9	,	,	PUNCT
cana-2597	242	10	manmohan	manmohan	PROPN
cana-2597	242	11	&	&	CCONJ
cana-2597	242	12	masrani	masrani	PROPN
cana-2597	242	13	,	,	PUNCT
cana-2597	242	14	vinit	vinit	PROPN
cana-2597	242	15	&	&	CCONJ
cana-2597	242	16	fernandes	fernandes	PROPN
cana-2597	242	17	,	,	PUNCT
cana-2597	242	18	dwayne	dwayne	PROPN
cana-2597	242	19	&	&	CCONJ
cana-2597	242	20	d'souza	d'souza	PROPN
cana-2597	242	21	,	,	PUNCT
cana-2597	242	22	kevin	kevin	PROPN
cana-2597	242	23	&	&	CCONJ
cana-2597	242	24	fernandes	fernandes	PROPN
cana-2597	242	25	,	,	PUNCT
cana-2597	242	26	delicia	delicia	PROPN
cana-2597	242	27	&	&	CCONJ
cana-2597	242	28	carvalho	carvalho	PROPN
cana-2597	242	29	,	,	PUNCT
cana-2597	242	30	tejal	tejal	NOUN
cana-2597	242	31	.	.	PUNCT
cana-2597	243	1	(	(	PUNCT
cana-2597	243	2	2020	2020	NUM
cana-2597	243	3	)	)	PUNCT
cana-2597	243	4	.	.	PUNCT
cana-2597	244	1	“	"	PUNCT
cana-2597	244	2	automated	automate	VERB
cana-2597	244	3	identification	identification	NOUN
cana-2597	244	4	of	of	ADP
cana-2597	244	5	disaster	disaster	NOUN
cana-2597	244	6	news	news	NOUN
cana-2597	244	7	for	for	ADP
cana-2597	244	8	crisis	crisis	NOUN
cana-2597	244	9	management	management	NOUN
cana-2597	244	10	using	use	VERB
cana-2597	244	11	machine	machine	NOUN
cana-2597	244	12	learning	learning	NOUN
cana-2597	244	13	and	and	CCONJ
cana-2597	244	14	natural	natural	ADJ
cana-2597	244	15	language	language	NOUN
cana-2597	244	16	processing	processing	NOUN
cana-2597	244	17	.	.	PUNCT
cana-2597	244	18	”	"	PUNCT
cana-2597	245	1	503	503	NUM
cana-2597	245	2	-	-	NUM
cana-2597	245	3	508	508	NUM
cana-2597	245	4	.	.	PUNCT
cana-2597	246	1	10.1109	10.1109	NUM
cana-2597	246	2	/	/	SYM
cana-2597	246	3	icesc48915.2020.9156031	icesc48915.2020.9156031	NOUN
cana-2597	246	4	.	.	PUNCT
cana-2597	247	1	[	[	X
cana-2597	247	2	6	6	NUM
cana-2597	247	3	]	]	X
cana-2597	247	4	mishra	mishra	PROPN
cana-2597	247	5	,	,	PUNCT
cana-2597	247	6	sameer	sameer	PROPN
cana-2597	247	7	&	&	CCONJ
cana-2597	247	8	bisen	bisen	PROPN
cana-2597	247	9	,	,	PUNCT
cana-2597	247	10	atharva	atharva	PROPN
cana-2597	247	11	&	&	CCONJ
cana-2597	247	12	mundhada	mundhada	PROPN
cana-2597	247	13	,	,	PUNCT
cana-2597	247	14	soham	soham	PROPN
cana-2597	247	15	&	&	CCONJ
cana-2597	247	16	singh	singh	PROPN
cana-2597	247	17	,	,	PUNCT
cana-2597	247	18	utkarsh	utkarsh	PROPN
cana-2597	247	19	&	&	CCONJ
cana-2597	247	20	bongirwar	bongirwar	PROPN
cana-2597	247	21	,	,	PUNCT
cana-2597	247	22	vrushali	vrushali	PROPN
cana-2597	247	23	.	.	PUNCT
cana-2597	248	1	(	(	PUNCT
cana-2597	248	2	2022	2022	NUM
cana-2597	248	3	)	)	PUNCT
cana-2597	248	4	.	.	PUNCT
cana-2597	249	1	“	"	PUNCT
cana-2597	249	2	disaster	disaster	NOUN
cana-2597	249	3	information	information	NOUN
cana-2597	249	4	verification	verification	NOUN
cana-2597	249	5	and	and	CCONJ
cana-2597	249	6	validation	validation	NOUN
cana-2597	249	7	application	application	NOUN
cana-2597	249	8	using	use	VERB
cana-2597	249	9	machine	machine	NOUN
cana-2597	249	10	learning	learning	NOUN
cana-2597	249	11	.	.	PUNCT
cana-2597	250	1	international	international	ADJ
cana-2597	250	2	journal	journal	PROPN
cana-2597	250	3	of	of	ADP
cana-2597	250	4	nextgeneration	nextgeneration	PROPN
cana-2597	250	5	computing	computing	NOUN
cana-2597	250	6	”	"	PUNCT
cana-2597	250	7	.	.	PUNCT
cana-2597	251	1	10.47164	10.47164	NUM
cana-2597	251	2	/	/	SYM
cana-2597	251	3	ijngc.v13i5.907	ijngc.v13i5.907	PROPN
cana-2597	251	4	.	.	PUNCT
cana-2597	252	1	[	[	X
cana-2597	252	2	7	7	NUM
cana-2597	252	3	]	]	PUNCT
cana-2597	252	4	ningsih	ningsih	NOUN
cana-2597	252	5	,	,	PUNCT
cana-2597	252	6	a	a	PRON
cana-2597	252	7	&	&	CCONJ
cana-2597	252	8	i	i	PROPN
cana-2597	252	9	d	d	PROPN
cana-2597	252	10	hadiana	hadiana	PROPN
cana-2597	252	11	,	,	PUNCT
cana-2597	252	12	asep	asep	PROPN
cana-2597	252	13	.	.	PUNCT
cana-2597	253	1	(	(	PUNCT
cana-2597	253	2	2021	2021	NUM
cana-2597	253	3	)	)	PUNCT
cana-2597	253	4	.	.	PUNCT
cana-2597	254	1	disaster	disaster	NOUN
cana-2597	254	2	tweets	tweet	VERB
cana-2597	254	3	classification	classification	NOUN
cana-2597	254	4	in	in	ADP
cana-2597	254	5	disaster	disaster	NOUN
cana-2597	254	6	response	response	NOUN
cana-2597	254	7	using	use	VERB
cana-2597	254	8	bidirectional	bidirectional	ADJ
cana-2597	254	9	encoder	encoder	NOUN
cana-2597	254	10	representations	representation	VERB
cana-2597	254	11	from	from	ADP
cana-2597	254	12	transformer	transformer	NOUN
cana-2597	254	13	(	(	PUNCT
cana-2597	254	14	bert	bert	PROPN
cana-2597	254	15	)	)	PUNCT
cana-2597	254	16	.	.	PUNCT
cana-2597	255	1	iop	iop	PROPN
cana-2597	255	2	conference	conference	PROPN
cana-2597	255	3	series	series	PROPN
cana-2597	255	4	:	:	PUNCT
cana-2597	255	5	materials	material	NOUN
cana-2597	255	6	science	science	NOUN
cana-2597	255	7	and	and	CCONJ
cana-2597	255	8	engineering	engineering	NOUN
cana-2597	255	9	.	.	PUNCT
cana-2597	256	1	1115	1115	NUM
cana-2597	256	2	.	.	PUNCT
cana-2597	257	1	012032	012032	NUM
cana-2597	257	2	.	.	PUNCT
cana-2597	258	1	10.1088/1757	10.1088/1757	X
cana-2597	258	2	-	-	SYM
cana-2597	258	3	899x/1115/1/012032	899x/1115/1/012032	NUM
cana-2597	258	4	.	.	PUNCT
cana-2597	259	1	[	[	X
cana-2597	259	2	8	8	NUM
cana-2597	259	3	]	]	X
cana-2597	259	4	kusumasari	kusumasari	PROPN
cana-2597	259	5	,	,	PUNCT
cana-2597	259	6	b.	b.	PROPN
cana-2597	259	7	,	,	PUNCT
cana-2597	259	8	prabowo	prabowo	NOUN
cana-2597	259	9	,	,	PUNCT
cana-2597	259	10	n.p.a	n.p.a	NOUN
cana-2597	259	11	.	.	PUNCT
cana-2597	260	1	“	"	PUNCT
cana-2597	260	2	scraping	scrape	VERB
cana-2597	260	3	social	social	ADJ
cana-2597	260	4	media	medium	NOUN
cana-2597	260	5	data	datum	NOUN
cana-2597	260	6	for	for	ADP
cana-2597	260	7	disaster	disaster	NOUN
cana-2597	260	8	communication	communication	NOUN
cana-2597	260	9	:	:	PUNCT
cana-2597	260	10	how	how	SCONJ
cana-2597	260	11	the	the	DET
cana-2597	260	12	pattern	pattern	NOUN
cana-2597	260	13	of	of	ADP
cana-2597	260	14	twitter	twitter	NOUN
cana-2597	260	15	users	user	NOUN
cana-2597	260	16	affects	affect	VERB
cana-2597	260	17	disasters	disaster	NOUN
cana-2597	260	18	in	in	ADP
cana-2597	260	19	asia	asia	PROPN
cana-2597	260	20	and	and	CCONJ
cana-2597	260	21	the	the	DET
cana-2597	260	22	pacific	pacific	PROPN
cana-2597	260	23	.	.	PUNCT
cana-2597	261	1	”nat	”nat	PRON
cana-2597	261	2	hazards	hazard	NOUN
cana-2597	261	3	103	103	NUM
cana-2597	261	4	,	,	PUNCT
cana-2597	261	5	(	(	PUNCT
cana-2597	261	6	2020	2020	NUM
cana-2597	261	7	)	)	PUNCT
cana-2597	261	8	.	.	PUNCT
cana-2597	262	1	https://doi.org/10.1007/s11069-02004136-z	https://doi.org/10.1007/s11069-02004136-z	PROPN
cana-2597	263	1	[	[	X
cana-2597	263	2	9	9	NUM
cana-2597	263	3	]	]	PUNCT
cana-2597	263	4	m.	m.	NOUN
cana-2597	263	5	laya	laya	PROPN
cana-2597	263	6	,	,	PUNCT
cana-2597	263	7	m.	m.	PROPN
cana-2597	263	8	k.	k.	PROPN
cana-2597	263	9	delimayanti	delimayanti	PROPN
cana-2597	263	10	,	,	PUNCT
cana-2597	263	11	a.	a.	PROPN
cana-2597	263	12	mardiyono	mardiyono	PROPN
cana-2597	263	13	,	,	PUNCT
cana-2597	263	14	f.	f.	PROPN
cana-2597	263	15	setianingrum	setianingrum	PROPN
cana-2597	263	16	,	,	PUNCT
cana-2597	263	17	a.	a.	NOUN
cana-2597	263	18	mahmudah	mahmudah	PROPN
cana-2597	263	19	and	and	CCONJ
cana-2597	263	20	d.	d.	PROPN
cana-2597	263	21	anggraini	anggraini	PROPN
cana-2597	263	22	,	,	PUNCT
cana-2597	263	23	"	"	PUNCT
cana-2597	263	24	classification	classification	NOUN
cana-2597	263	25	of	of	ADP
cana-2597	263	26	natural	natural	ADJ
cana-2597	263	27	disaster	disaster	NOUN
cana-2597	263	28	on	on	ADP
cana-2597	263	29	online	online	ADJ
cana-2597	263	30	news	news	NOUN
cana-2597	263	31	data	datum	NOUN
cana-2597	263	32	using	use	VERB
cana-2597	263	33	machine	machine	NOUN
cana-2597	263	34	learning	learning	NOUN
cana-2597	263	35	,	,	PUNCT
cana-2597	263	36	"	"	PUNCT
cana-2597	263	37	2021	2021	NUM
cana-2597	263	38	5th	5th	ADJ
cana-2597	263	39	international	international	ADJ
cana-2597	263	40	conference	conference	NOUN
cana-2597	263	41	on	on	ADP
cana-2597	263	42	electrical	electrical	ADJ
cana-2597	263	43	,	,	PUNCT
cana-2597	263	44	telecommunication	telecommunication	NOUN
cana-2597	263	45	and	and	CCONJ
cana-2597	263	46	computer	computer	NOUN
cana-2597	263	47	engineering	engineering	NOUN
cana-2597	263	48	(	(	PUNCT
cana-2597	263	49	elticom	elticom	NOUN
cana-2597	263	50	)	)	PUNCT
cana-2597	263	51	,	,	PUNCT
cana-2597	263	52	medan	medan	PROPN
cana-2597	263	53	,	,	PUNCT
cana-2597	263	54	indonesia	indonesia	PROPN
cana-2597	263	55	,	,	PUNCT
cana-2597	263	56	2021	2021	NUM
cana-2597	263	57	,	,	PUNCT
cana-2597	263	58	pp	pp	ADV
cana-2597	263	59	.	.	PUNCT
cana-2597	263	60	42	42	NUM
cana-2597	263	61	-	-	SYM
cana-2597	263	62	46	46	NUM
cana-2597	263	63	,	,	PUNCT
cana-2597	263	64	doi	doi	NOUN
cana-2597	263	65	:	:	PUNCT
cana-2597	263	66	10.1109	10.1109	NUM
cana-2597	263	67	/	/	SYM
cana-2597	263	68	elticom53303.2021.9590125	elticom53303.2021.9590125	NOUN
cana-2597	263	69	.	.	PUNCT
cana-2597	264	1	[	[	X
cana-2597	264	2	10	10	NUM
cana-2597	264	3	]	]	SYM
cana-2597	264	4	a.	a.	NOUN
cana-2597	264	5	,	,	PUNCT
cana-2597	264	6	aswathy	aswathy	ADJ
cana-2597	264	7	.	.	PUNCT
cana-2597	264	8	,	,	PUNCT
cana-2597	264	9	rekha	rekha	PROPN
cana-2597	264	10	,	,	PUNCT
cana-2597	264	11	prabha	prabha	NOUN
cana-2597	264	12	.	.	PUNCT
cana-2597	264	13	,	,	PUNCT
cana-2597	264	14	lakshmi	lakshmi	PROPN
cana-2597	264	15	,	,	PUNCT
cana-2597	264	16	gopal	gopal	PROPN
cana-2597	264	17	.	.	PROPN
cana-2597	264	18	,	,	PUNCT
cana-2597	264	19	divya	divya	PROPN
cana-2597	264	20	,	,	PUNCT
cana-2597	264	21	pullarkatt	pullarkatt	PROPN
cana-2597	264	22	.	.	PUNCT
cana-2597	264	23	,	,	PUNCT
cana-2597	264	24	maneesha	maneesha	PROPN
cana-2597	264	25	,	,	PUNCT
cana-2597	264	26	vinodini	vinodini	PROPN
cana-2597	264	27	,	,	PUNCT
cana-2597	264	28	ramesh	ramesh	PROPN
cana-2597	264	29	.	.	PUNCT
cana-2597	265	1	(	(	PUNCT
cana-2597	265	2	2022	2022	NUM
cana-2597	265	3	)	)	PUNCT
cana-2597	265	4	.	.	PUNCT
cana-2597	266	1	“	"	PUNCT
cana-2597	266	2	an	an	DET
cana-2597	266	3	efficient	efficient	ADJ
cana-2597	266	4	twitter	twitter	NOUN
cana-2597	266	5	data	datum	NOUN
cana-2597	266	6	collection	collection	NOUN
cana-2597	266	7	and	and	CCONJ
cana-2597	266	8	analytics	analytic	NOUN
cana-2597	266	9	framework	framework	NOUN
cana-2597	266	10	for	for	ADP
cana-2597	266	11	effective	effective	ADJ
cana-2597	266	12	disaster	disaster	NOUN
cana-2597	266	13	management	management	NOUN
cana-2597	266	14	.	.	PUNCT
cana-2597	266	15	”	"	PUNCT
cana-2597	267	1	doi	doi	NOUN
cana-2597	267	2	:	:	PUNCT
cana-2597	267	3	0.1109	0.1109	NUM
cana-2597	267	4	/	/	SYM
cana-2597	267	5	delcon54057.2022.9753627	delcon54057.2022.9753627	PROPN
cana-2597	268	1	[	[	X
cana-2597	268	2	11	11	NUM
cana-2597	268	3	]	]	SYM
cana-2597	268	4	meghna	meghna	NOUN
cana-2597	268	5	,	,	PUNCT
cana-2597	268	6	jayan	jayan	PROPN
cana-2597	268	7	.	.	PROPN
cana-2597	268	8	,	,	PUNCT
cana-2597	268	9	lija	lija	PROPN
cana-2597	268	10	,	,	PUNCT
cana-2597	268	11	jacob	jacob	PROPN
cana-2597	268	12	.	.	PUNCT
cana-2597	268	13	(	(	PUNCT
cana-2597	268	14	2023	2023	NUM
cana-2597	268	15	)	)	PUNCT
cana-2597	268	16	.	.	PUNCT
cana-2597	269	1	1	1	X
cana-2597	269	2	.	.	PUNCT
cana-2597	269	3	“	"	PUNCT
cana-2597	269	4	categorizing	categorize	VERB
cana-2597	269	5	disaster	disaster	NOUN
cana-2597	269	6	tweets	tweet	NOUN
cana-2597	269	7	using	use	VERB
cana-2597	269	8	learning	learn	VERB
cana-2597	269	9	based	base	VERB
cana-2597	269	10	models	model	NOUN
cana-2597	269	11	for	for	ADP
cana-2597	269	12	emergency	emergency	NOUN
cana-2597	269	13	crisis	crisis	NOUN
cana-2597	269	14	management	management	NOUN
cana-2597	269	15	.	.	PUNCT
cana-2597	269	16	”	"	PUNCT
cana-2597	270	1	doi	doi	NOUN
cana-2597	270	2	:	:	PUNCT
cana-2597	270	3	10.1109	10.1109	NUM
cana-2597	270	4	/	/	SYM
cana-2597	270	5	icaccs57279.2023.10113105	icaccs57279.2023.10113105	PROPN
cana-2597	271	1	[	[	X
cana-2597	271	2	12	12	NUM
cana-2597	271	3	]	]	X
cana-2597	271	4	p.	p.	NOUN
cana-2597	271	5	,	,	PUNCT
cana-2597	271	6	gnanasekaran	gnanasekaran	NOUN
cana-2597	271	7	.	.	PUNCT
cana-2597	272	1	(	(	PUNCT
cana-2597	272	2	2023	2023	NUM
cana-2597	272	3	)	)	PUNCT
cana-2597	272	4	.	.	PUNCT
cana-2597	273	1	1	1	X
cana-2597	273	2	.	.	PUNCT
cana-2597	273	3	“	"	PUNCT
cana-2597	273	4	juncture	juncture	NOUN
cana-2597	273	5	of	of	ADP
cana-2597	273	6	text	text	NOUN
cana-2597	273	7	preprocessing	preprocessing	NOUN
cana-2597	273	8	techniques	technique	NOUN
cana-2597	273	9	&	&	CCONJ
cana-2597	273	10	extracting	extract	VERB
cana-2597	273	11	sentiment	sentiment	NOUN
cana-2597	273	12	analyzing	analyze	VERB
cana-2597	273	13	of	of	ADP
cana-2597	273	14	micro	micro	NOUN
cana-2597	273	15	-	-	NOUN
cana-2597	273	16	blog	blog	NOUN
cana-2597	273	17	based	base	VERB
cana-2597	273	18	on	on	ADP
cana-2597	273	19	machine	machine	NOUN
cana-2597	273	20	learning	learning	NOUN
cana-2597	273	21	algorithms	algorithm	NOUN
cana-2597	273	22	.	.	PUNCT
cana-2597	273	23	”	"	PUNCT
cana-2597	274	1	doi	doi	NOUN
cana-2597	274	2	:	:	PUNCT
cana-2597	274	3	10.1109	10.1109	NUM
cana-2597	274	4	/	/	SYM
cana-2597	274	5	icses60034.2023.10465450	icses60034.2023.10465450	NOUN
cana-2597	275	1	[	[	PUNCT
cana-2597	275	2	13	13	NUM
cana-2597	275	3	]	]	X
cana-2597	275	4	parth	parth	NOUN
cana-2597	275	5	,	,	PUNCT
cana-2597	275	6	mahajan	mahajan	PROPN
cana-2597	275	7	.	.	PROPN
cana-2597	275	8	,	,	PUNCT
cana-2597	275	9	pranshu	pranshu	NOUN
cana-2597	275	10	,	,	PUNCT
cana-2597	275	11	raghuwanshi	raghuwanshi	NOUN
cana-2597	275	12	.	.	PROPN
cana-2597	275	13	,	,	PUNCT
cana-2597	275	14	hardik	hardik	PROPN
cana-2597	275	15	,	,	PUNCT
cana-2597	275	16	setia	setia	PROPN
cana-2597	275	17	.	.	PROPN
cana-2597	275	18	,	,	PUNCT
cana-2597	275	19	princy	princy	NOUN
cana-2597	275	20	,	,	PUNCT
cana-2597	275	21	randhawa	randhawa	PROPN
cana-2597	275	22	.	.	PUNCT
cana-2597	276	1	(	(	PUNCT
cana-2597	276	2	2024	2024	NUM
cana-2597	276	3	)	)	PUNCT
cana-2597	276	4	.	.	PUNCT
cana-2597	277	1	1	1	X
cana-2597	277	2	.	.	PUNCT
cana-2597	277	3	“	"	PUNCT
cana-2597	277	4	a	a	DET
cana-2597	277	5	multi	multi	ADJ
cana-2597	277	6	-	-	ADJ
cana-2597	277	7	model	model	ADJ
cana-2597	277	8	approach	approach	NOUN
cana-2597	277	9	for	for	ADP
cana-2597	277	10	disaster	disaster	NOUN
cana-2597	277	11	-	-	PUNCT
cana-2597	277	12	related	relate	VERB
cana-2597	277	13	tweets	tweet	NOUN
cana-2597	277	14	.	.	PUNCT
cana-2597	277	15	”	"	PUNCT
cana-2597	278	1	doi:10.57159	doi:10.57159	PROPN
cana-2597	278	2	/	/	SYM
cana-2597	278	3	gadl.jcmm.3.2.240125	gadl.jcmm.3.2.240125	PART
cana-2597	279	1	[	[	X
cana-2597	279	2	14	14	NUM
cana-2597	279	3	]	]	X
cana-2597	279	4	pardeep	pardeep	NOUN
cana-2597	279	5	,	,	PUNCT
cana-2597	279	6	singh	singh	PROPN
cana-2597	279	7	.	.	PROPN
cana-2597	279	8	,	,	PUNCT
cana-2597	279	9	monika	monika	PROPN
cana-2597	279	10	,	,	PUNCT
cana-2597	279	11	singh	singh	PROPN
cana-2597	279	12	.	.	PUNCT
cana-2597	279	13	,	,	PUNCT
cana-2597	280	1	p.	p.	PROPN
cana-2597	280	2	,	,	PUNCT
cana-2597	280	3	das	das	PROPN
cana-2597	280	4	.	.	PROPN
cana-2597	280	5	,	,	PUNCT
cana-2597	280	6	satish	satish	PROPN
cana-2597	280	7	,	,	PUNCT
cana-2597	280	8	chand	chand	PROPN
cana-2597	280	9	.	.	PUNCT
cana-2597	281	1	(	(	PUNCT
cana-2597	281	2	2024	2024	NUM
cana-2597	281	3	)	)	PUNCT
cana-2597	281	4	.	.	PUNCT
cana-2597	282	1	2	2	X
cana-2597	282	2	.	.	PUNCT
cana-2597	282	3	“	"	PUNCT
cana-2597	282	4	leveraging	leverage	VERB
cana-2597	282	5	domain	domain	NOUN
cana-2597	282	6	-	-	PUNCT
cana-2597	282	7	adapted	adapt	VERB
cana-2597	282	8	transformer	transformer	NOUN
cana-2597	282	9	based	base	VERB
cana-2597	282	10	models	model	NOUN
cana-2597	282	11	for	for	ADP
cana-2597	282	12	disaster	disaster	NOUN
cana-2597	282	13	-	-	PUNCT
cana-2597	282	14	specific	specific	ADJ
cana-2597	282	15	tweet	tweet	NOUN
cana-2597	282	16	classification	classification	NOUN
cana-2597	282	17	.	.	PUNCT
cana-2597	282	18	”	"	PUNCT
cana-2597	282	19	doi	doi	NOUN
cana-2597	282	20	:	:	PUNCT
cana-2597	282	21	10.3233	10.3233	NUM
cana-2597	282	22	/	/	SYM
cana-2597	282	23	jifs-219419	jifs-219419	PROPN
cana-2597	283	1	[	[	X
cana-2597	283	2	15	15	NUM
cana-2597	283	3	]	]	X
cana-2597	283	4	wei	wei	PROPN
cana-2597	283	5	,	,	PUNCT
cana-2597	283	6	chen	chen	PROPN
cana-2597	283	7	.	.	PROPN
cana-2597	283	8	,	,	PUNCT
cana-2597	283	9	janet	janet	PROPN
cana-2597	283	10	,	,	PUNCT
cana-2597	283	11	fang	fang	PROPN
cana-2597	283	12	.	.	PUNCT
cana-2597	283	13	(	(	PUNCT
cana-2597	283	14	2024	2024	NUM
cana-2597	283	15	)	)	PUNCT
cana-2597	283	16	.	.	PUNCT
cana-2597	284	1	1	1	X
cana-2597	284	2	.	.	PUNCT
cana-2597	284	3	“	"	PUNCT
cana-2597	284	4	optimizing	optimize	VERB
cana-2597	284	5	ai	ai	NOUN
cana-2597	284	6	-	-	PUNCT
cana-2597	284	7	driven	drive	VERB
cana-2597	284	8	disaster	disaster	NOUN
cana-2597	284	9	management	management	NOUN
cana-2597	284	10	through	through	ADP
cana-2597	284	11	llms	llm	NOUN
cana-2597	284	12	.	.	PUNCT
cana-2597	284	13	”	"	PUNCT
cana-2597	284	14	doi	doi	NOUN
cana-2597	284	15	:	:	PUNCT
cana-2597	284	16	10.20944	10.20944	NUM
cana-2597	284	17	/	/	SYM
cana-2597	285	1	preprints202407.1446.v1	preprints202407.1446.v1	NOUN
cana-2597	285	2	[	[	X
cana-2597	285	3	16	16	NUM
cana-2597	285	4	]	]	X
cana-2597	285	5	frankie	frankie	NOUN
cana-2597	285	6	,	,	PUNCT
cana-2597	285	7	frankie	frankie	NOUN
cana-2597	285	8	.	.	PUNCT
cana-2597	285	9	,	,	PUNCT
cana-2597	285	10	yeremia	yeremia	NOUN
cana-2597	285	11	,	,	PUNCT
cana-2597	285	12	alfa	alfa	NOUN
cana-2597	285	13	,	,	PUNCT
cana-2597	285	14	susetyo	susetyo	NOUN
cana-2597	285	15	.	.	PUNCT
cana-2597	285	16	(	(	PUNCT
cana-2597	285	17	2023	2023	NUM
cana-2597	285	18	)	)	PUNCT
cana-2597	285	19	.	.	PUNCT
cana-2597	286	1	2	2	X
cana-2597	286	2	.	.	X
cana-2597	286	3	“	"	PUNCT
cana-2597	286	4	implementation	implementation	NOUN
cana-2597	286	5	of	of	ADP
cana-2597	286	6	text	text	NOUN
cana-2597	286	7	indexing	indexing	NOUN
cana-2597	286	8	system	system	NOUN
cana-2597	286	9	in	in	ADP
cana-2597	286	10	web	web	NOUN
cana-2597	286	11	-	-	PUNCT
cana-2597	286	12	based	base	VERB
cana-2597	286	13	document	document	NOUN
cana-2597	286	14	search	search	NOUN
cana-2597	286	15	application	application	NOUN
cana-2597	286	16	using	use	VERB
cana-2597	286	17	mongodb	mongodb	NOUN
cana-2597	286	18	.	.	PUNCT
cana-2597	286	19	”	"	PUNCT
cana-2597	287	1	doi	doi	NOUN
cana-2597	287	2	:	:	PUNCT
cana-2597	287	3	10.52436/1.jutif.2023.4.5.959	10.52436/1.jutif.2023.4.5.959	NUM
cana-2597	287	4	[	[	SYM
cana-2597	287	5	17	17	NUM
cana-2597	287	6	]	]	PUNCT
cana-2597	287	7	siva	siva	NOUN
cana-2597	287	8	,	,	PUNCT
cana-2597	287	9	krishna	krishna	PROPN
cana-2597	287	10	,	,	PUNCT
cana-2597	287	11	dasari	dasari	NOUN
cana-2597	287	12	.	.	PUNCT
cana-2597	287	13	,	,	PUNCT
cana-2597	287	14	g.	g.	PROPN
cana-2597	287	15	,	,	PUNCT
cana-2597	287	16	srinivas	srinivas	PROPN
cana-2597	287	17	.	.	PROPN
cana-2597	287	18	,	,	PUNCT
cana-2597	288	1	p.	p.	NOUN
cana-2597	288	2	,	,	PUNCT
cana-2597	288	3	jayarama	jayarama	PROPN
cana-2597	288	4	,	,	PUNCT
cana-2597	288	5	reddy	reddy	PROPN
cana-2597	288	6	.	.	PUNCT
cana-2597	289	1	(	(	PUNCT
cana-2597	289	2	2024	2024	NUM
cana-2597	289	3	)	)	PUNCT
cana-2597	289	4	.	.	PUNCT
cana-2597	290	1	2	2	X
cana-2597	290	2	.	.	PUNCT
cana-2597	290	3	“	"	PUNCT
cana-2597	290	4	a	a	DET
cana-2597	290	5	deep	deep	ADJ
cana-2597	290	6	parallel	parallel	ADJ
cana-2597	290	7	hybrid	hybrid	ADJ
cana-2597	290	8	fusion	fusion	NOUN
cana-2597	290	9	model	model	NOUN
cana-2597	290	10	for	for	ADP
cana-2597	290	11	disaster	disaster	NOUN
cana-2597	290	12	tweet	tweet	NOUN
cana-2597	290	13	classification	classification	NOUN
cana-2597	290	14	on	on	ADP
cana-2597	290	15	twitter	twitter	PROPN
cana-2597	290	16	data	datum	NOUN
cana-2597	290	17	.	.	PUNCT
cana-2597	290	18	”	"	PUNCT
cana-2597	291	1	decision	decision	NOUN
cana-2597	291	2	analytics	analytic	NOUN
cana-2597	291	3	journal	journal	NOUN
cana-2597	291	4	,	,	PUNCT
cana-2597	291	5	doi	doi	PROPN
cana-2597	291	6	:	:	PUNCT
cana-2597	291	7	10.1016	10.1016	NUM
cana-2597	291	8	/	/	SYM
cana-2597	291	9	j.dajour.2024.100453	j.dajour.2024.100453	NOUN