id	sid	tid	token	lemma	pos
cana-5777	1	1	comparing	compare	VERB
cana-5777	1	2	machine	machine	NOUN
cana-5777	1	3	learning	learn	VERB
cana-5777	1	4	algorithms	algorithm	NOUN
cana-5777	1	5	:	:	PUNCT
cana-5777	1	6	a	a	DET
cana-5777	1	7	graph	graph	NOUN
cana-5777	1	8	theory	theory	NOUN
cana-5777	1	9	approach	approach	NOUN
cana-5777	1	10	for	for	ADP
cana-5777	1	11	improving	improve	VERB
cana-5777	1	12	accuracy	accuracy	NOUN
cana-5777	1	13	1	1	NUM
cana-5777	1	14	g.	g.	NOUN
cana-5777	1	15	keerthi	keerthi	PROPN
cana-5777	1	16	2,*m	2,*m	PROPN
cana-5777	1	17	.	.	PUNCT
cana-5777	2	1	siva	siva	PROPN
cana-5777	2	2	parvathi	parvathi	PROPN
cana-5777	2	3	3d	3d	PROPN
cana-5777	2	4	.	.	PUNCT
cana-5777	3	1	sujatha	sujatha	PROPN
cana-5777	3	2	4n.v	4n.v	NUM
cana-5777	3	3	.	.	PUNCT
cana-5777	3	4	muthu	muthu	NOUN
cana-5777	3	5	lakshmi	lakshmi	PROPN
cana-5777	3	6	1,2department	1,2department	NUM
cana-5777	3	7	of	of	ADP
cana-5777	3	8	applied	apply	VERB
cana-5777	3	9	mathematics	mathematic	NOUN
cana-5777	3	10	,	,	PUNCT
cana-5777	3	11	sri	sri	PROPN
cana-5777	3	12	padmavati	padmavati	PROPN
cana-5777	3	13	mahilavisvavidyalayam	mahilavisvavidyalayam	PROPN
cana-5777	3	14	,	,	PUNCT
cana-5777	3	15	tirupati	tirupati	PROPN
cana-5777	3	16	,	,	PUNCT
cana-5777	3	17	andhra	andhra	PROPN
cana-5777	3	18	pradesh	pradesh	PROPN
cana-5777	3	19	,	,	PUNCT
cana-5777	3	20	india	india	PROPN
cana-5777	3	21	3institute	3institute	NUM
cana-5777	3	22	of	of	ADP
cana-5777	3	23	pharmaceutical	pharmaceutical	ADJ
cana-5777	3	24	technology	technology	NOUN
cana-5777	3	25	,	,	PUNCT
cana-5777	3	26	sri	sri	PROPN
cana-5777	3	27	padmavati	padmavati	PROPN
cana-5777	3	28	mahilavisvavidyalayam	mahilavisvavidyalayam	PROPN
cana-5777	3	29	,	,	PUNCT
cana-5777	3	30	tirupati	tirupati	PROPN
cana-5777	3	31	,	,	PUNCT
cana-5777	3	32	andhra	andhra	PROPN
cana-5777	3	33	pradesh	pradesh	PROPN
cana-5777	3	34	,	,	PUNCT
cana-5777	3	35	india	india	PROPN
cana-5777	3	36	4department	4department	NUM
cana-5777	3	37	of	of	ADP
cana-5777	3	38	computer	computer	NOUN
cana-5777	3	39	science	science	NOUN
cana-5777	3	40	,	,	PUNCT
cana-5777	3	41	sri	sri	PROPN
cana-5777	3	42	padmavati	padmavati	PROPN
cana-5777	3	43	mahilavisvavidyalayam	mahilavisvavidyalayam	PROPN
cana-5777	3	44	,	,	PUNCT
cana-5777	3	45	tirupati	tirupati	PROPN
cana-5777	3	46	,	,	PUNCT
cana-5777	3	47	andhra	andhra	PROPN
cana-5777	3	48	pradesh	pradesh	PROPN
cana-5777	3	49	,	,	PUNCT
cana-5777	3	50	india	india	PROPN
cana-5777	3	51	corresponding	corresponding	PROPN
cana-5777	3	52	author	author	NOUN
cana-5777	3	53	:	:	PUNCT
cana-5777	3	54	m.siva	m.siva	PROPN
cana-5777	3	55	parvathi	parvathi	PROPN
cana-5777	3	56	article	article	PROPN
cana-5777	3	57	history	history	NOUN
cana-5777	3	58	:	:	PUNCT
cana-5777	3	59	received	receive	VERB
cana-5777	3	60	20.09.2024	20.09.2024	NUM
cana-5777	3	61	revised	revise	VERB
cana-5777	3	62	:	:	PUNCT
cana-5777	3	63	24.10.2024	24.10.2024	NUM
cana-5777	3	64	accepted	accept	VERB
cana-5777	3	65	:	:	PUNCT
cana-5777	3	66	30.11.2024	30.11.2024	NUM
cana-5777	3	67	abstract	abstract	ADJ
cana-5777	3	68	graph	graph	NOUN
cana-5777	3	69	theory	theory	NOUN
cana-5777	3	70	provides	provide	VERB
cana-5777	3	71	a	a	DET
cana-5777	3	72	robust	robust	ADJ
cana-5777	3	73	framework	framework	NOUN
cana-5777	3	74	for	for	ADP
cana-5777	3	75	modelling	model	VERB
cana-5777	3	76	complex	complex	ADJ
cana-5777	3	77	relationships	relationship	NOUN
cana-5777	3	78	in	in	ADP
cana-5777	3	79	medical	medical	ADJ
cana-5777	3	80	data	datum	NOUN
cana-5777	3	81	,	,	PUNCT
cana-5777	3	82	enhancing	enhance	VERB
cana-5777	3	83	classification	classification	NOUN
cana-5777	3	84	accuracy	accuracy	NOUN
cana-5777	3	85	through	through	ADP
cana-5777	3	86	relational	relational	ADJ
cana-5777	3	87	learning	learning	NOUN
cana-5777	3	88	.	.	PUNCT
cana-5777	4	1	unlike	unlike	ADP
cana-5777	4	2	traditional	traditional	ADJ
cana-5777	4	3	machine	machine	NOUN
cana-5777	4	4	learning	learning	NOUN
cana-5777	4	5	(	(	PUNCT
cana-5777	4	6	ml	ml	NOUN
cana-5777	4	7	)	)	PUNCT
cana-5777	4	8	models	model	NOUN
cana-5777	4	9	that	that	PRON
cana-5777	4	10	treat	treat	VERB
cana-5777	4	11	data	datum	NOUN
cana-5777	4	12	points	point	NOUN
cana-5777	4	13	independently	independently	ADV
cana-5777	4	14	,	,	PUNCT
cana-5777	4	15	graph	graph	NOUN
cana-5777	4	16	-	-	PUNCT
cana-5777	4	17	based	base	VERB
cana-5777	4	18	approaches	approach	NOUN
cana-5777	4	19	support	support	VERB
cana-5777	4	20	structural	structural	ADJ
cana-5777	4	21	dependencies	dependency	NOUN
cana-5777	4	22	to	to	PART
cana-5777	4	23	improve	improve	VERB
cana-5777	4	24	feature	feature	NOUN
cana-5777	4	25	representation	representation	NOUN
cana-5777	4	26	.	.	PUNCT
cana-5777	5	1	this	this	DET
cana-5777	5	2	study	study	NOUN
cana-5777	5	3	explores	explore	VERB
cana-5777	5	4	the	the	DET
cana-5777	5	5	application	application	NOUN
cana-5777	5	6	of	of	ADP
cana-5777	5	7	graph	graph	NOUN
cana-5777	5	8	attention	attention	NOUN
cana-5777	5	9	networks	network	NOUN
cana-5777	5	10	(	(	PUNCT
cana-5777	5	11	gat	gat	NOUN
cana-5777	5	12	)	)	PUNCT
cana-5777	5	13	,	,	PUNCT
cana-5777	5	14	graphsage	graphsage	NOUN
cana-5777	5	15	,	,	PUNCT
cana-5777	5	16	graph	graph	NOUN
cana-5777	5	17	convolutional	convolutional	ADJ
cana-5777	5	18	network(gcn	network(gcn	NOUN
cana-5777	5	19	)	)	PUNCT
cana-5777	5	20	and	and	CCONJ
cana-5777	5	21	graph	graph	VERB
cana-5777	5	22	isomorphism	isomorphism	NOUN
cana-5777	5	23	network	network	NOUN
cana-5777	5	24	(	(	PUNCT
cana-5777	5	25	gin	gin	NOUN
cana-5777	5	26	)	)	PUNCT
cana-5777	5	27	for	for	ADP
cana-5777	5	28	breast	breast	NOUN
cana-5777	5	29	cancer	cancer	NOUN
cana-5777	5	30	classification	classification	NOUN
cana-5777	5	31	,	,	PUNCT
cana-5777	5	32	employing	employ	VERB
cana-5777	5	33	k	k	NOUN
cana-5777	5	34	-	-	PUNCT
cana-5777	5	35	nearest	near	ADJ
cana-5777	5	36	neighbor	neighbor	NOUN
cana-5777	5	37	(	(	PUNCT
cana-5777	5	38	knn	knn	PROPN
cana-5777	5	39	)	)	PUNCT
cana-5777	5	40	graphs	graph	NOUN
cana-5777	5	41	to	to	PART
cana-5777	5	42	construct	construct	VERB
cana-5777	5	43	a	a	DET
cana-5777	5	44	structured	structured	ADJ
cana-5777	5	45	dataset	dataset	NOUN
cana-5777	5	46	where	where	SCONJ
cana-5777	5	47	nodes	node	NOUN
cana-5777	5	48	represent	represent	VERB
cana-5777	5	49	patients	patient	NOUN
cana-5777	5	50	and	and	CCONJ
cana-5777	5	51	edges	edge	NOUN
cana-5777	5	52	capture	capture	VERB
cana-5777	5	53	feature	feature	NOUN
cana-5777	5	54	similarities	similarity	NOUN
cana-5777	5	55	.	.	PUNCT
cana-5777	6	1	the	the	DET
cana-5777	6	2	effectiveness	effectiveness	NOUN
cana-5777	6	3	of	of	ADP
cana-5777	6	4	the	the	DET
cana-5777	6	5	graph	graph	NOUN
cana-5777	6	6	-	-	PUNCT
cana-5777	6	7	based	base	VERB
cana-5777	6	8	approaches	approach	NOUN
cana-5777	6	9	is	be	AUX
cana-5777	6	10	evaluated	evaluate	VERB
cana-5777	6	11	against	against	ADP
cana-5777	6	12	traditional	traditional	ADJ
cana-5777	6	13	ml	ml	NOUN
cana-5777	6	14	classifiers	classifier	NOUN
cana-5777	6	15	,	,	PUNCT
cana-5777	6	16	including	include	VERB
cana-5777	6	17	decision	decision	NOUN
cana-5777	6	18	trees	tree	NOUN
cana-5777	6	19	,	,	PUNCT
cana-5777	6	20	random	random	ADJ
cana-5777	6	21	forest	forest	NOUN
cana-5777	6	22	,	,	PUNCT
cana-5777	6	23	lightgbm	lightgbm	ADJ
cana-5777	6	24	,	,	PUNCT
cana-5777	6	25	and	and	CCONJ
cana-5777	6	26	xgboost	xgboost	X
cana-5777	6	27	.	.	PUNCT
cana-5777	7	1	experimental	experimental	ADJ
cana-5777	7	2	results	result	NOUN
cana-5777	7	3	indicate	indicate	VERB
cana-5777	7	4	that	that	PRON
cana-5777	7	5	gcn	gcn	NOUN
cana-5777	7	6	,	,	PUNCT
cana-5777	7	7	gin	gin	NOUN
cana-5777	7	8	,	,	PUNCT
cana-5777	7	9	gat	gat	NOUN
cana-5777	7	10	and	and	CCONJ
cana-5777	7	11	graphsage	graphsage	NOUN
cana-5777	7	12	are	be	AUX
cana-5777	7	13	beat	beat	VERB
cana-5777	7	14	conventional	conventional	ADJ
cana-5777	7	15	methods	method	NOUN
cana-5777	7	16	,	,	PUNCT
cana-5777	7	17	with	with	ADP
cana-5777	7	18	gcn	gcn	ADJ
cana-5777	7	19	,	,	PUNCT
cana-5777	7	20	gat	gat	NOUN
cana-5777	7	21	and	and	CCONJ
cana-5777	7	22	graphsage	graphsage	NOUN
cana-5777	7	23	achieved	achieve	VERB
cana-5777	7	24	100	100	NUM
cana-5777	7	25	%	%	NOUN
cana-5777	7	26	test	test	NOUN
cana-5777	7	27	-	-	PUNCT
cana-5777	7	28	accuracy	accuracy	NOUN
cana-5777	7	29	and	and	CCONJ
cana-5777	7	30	with	with	ADP
cana-5777	7	31	gin	gin	NOUN
cana-5777	7	32	achieved	achieve	VERB
cana-5777	7	33	99.42	99.42	NUM
cana-5777	7	34	%	%	NOUN
cana-5777	7	35	,	,	PUNCT
cana-5777	7	36	to	to	PART
cana-5777	7	37	confirm	confirm	VERB
cana-5777	7	38	the	the	DET
cana-5777	7	39	percentage	percentage	NOUN
cana-5777	7	40	of	of	ADP
cana-5777	7	41	accuracy	accuracy	NOUN
cana-5777	7	42	,	,	PUNCT
cana-5777	7	43	authors	author	NOUN
cana-5777	7	44	conducted	conduct	VERB
cana-5777	7	45	extensive	extensive	ADJ
cana-5777	7	46	experiments	experiment	NOUN
cana-5777	7	47	,	,	PUNCT
cana-5777	7	48	including	include	VERB
cana-5777	7	49	robustness	robustness	NOUN
cana-5777	7	50	testing	testing	NOUN
cana-5777	7	51	by	by	ADP
cana-5777	7	52	reducing	reduce	VERB
cana-5777	7	53	knn	knn	NOUN
cana-5777	7	54	connections	connection	NOUN
cana-5777	7	55	,	,	PUNCT
cana-5777	7	56	introducing	introduce	VERB
cana-5777	7	57	noise	noise	NOUN
cana-5777	7	58	,	,	PUNCT
cana-5777	7	59	and	and	CCONJ
cana-5777	7	60	shuffling	shuffle	VERB
cana-5777	7	61	train	train	NOUN
cana-5777	7	62	-	-	PUNCT
cana-5777	7	63	test	test	NOUN
cana-5777	7	64	splits	split	NOUN
cana-5777	7	65	.	.	PUNCT
cana-5777	8	1	results	result	NOUN
cana-5777	8	2	demonstrate	demonstrate	VERB
cana-5777	8	3	that	that	SCONJ
cana-5777	8	4	graph	graph	NOUN
cana-5777	8	5	-	-	PUNCT
cana-5777	8	6	based	base	VERB
cana-5777	8	7	models	model	NOUN
cana-5777	8	8	are	be	AUX
cana-5777	8	9	significantly	significantly	ADV
cana-5777	8	10	best	good	ADJ
cana-5777	8	11	than	than	ADP
cana-5777	8	12	traditional	traditional	ADJ
cana-5777	8	13	ml	ml	NOUN
cana-5777	8	14	models	model	NOUN
cana-5777	8	15	,	,	PUNCT
cana-5777	8	16	and	and	CCONJ
cana-5777	8	17	these	these	DET
cana-5777	8	18	graph	graph	NOUN
cana-5777	8	19	-	-	PUNCT
cana-5777	8	20	based	base	VERB
cana-5777	8	21	models	model	NOUN
cana-5777	8	22	maintain	maintain	VERB
cana-5777	8	23	same	same	ADJ
cana-5777	8	24	classification	classification	NOUN
cana-5777	8	25	accuracy	accuracy	NOUN
cana-5777	8	26	while	while	SCONJ
cana-5777	8	27	maintaining	maintain	VERB
cana-5777	8	28	stability	stability	NOUN
cana-5777	8	29	under	under	ADP
cana-5777	8	30	robustness	robustness	NOUN
cana-5777	8	31	tests	test	NOUN
cana-5777	8	32	.	.	PUNCT
cana-5777	9	1	the	the	DET
cana-5777	9	2	findings	finding	NOUN
cana-5777	9	3	confirm	confirm	VERB
cana-5777	9	4	that	that	SCONJ
cana-5777	9	5	graph	graph	NOUN
cana-5777	9	6	-	-	PUNCT
cana-5777	9	7	based	base	VERB
cana-5777	9	8	learning	learning	NOUN
cana-5777	9	9	provides	provide	VERB
cana-5777	9	10	a	a	DET
cana-5777	9	11	scalable	scalable	ADJ
cana-5777	9	12	,	,	PUNCT
cana-5777	9	13	interpretable	interpretable	ADJ
cana-5777	9	14	,	,	PUNCT
cana-5777	9	15	and	and	CCONJ
cana-5777	9	16	highly	highly	ADV
cana-5777	9	17	accurate	accurate	ADJ
cana-5777	9	18	alternative	alternative	NOUN
cana-5777	9	19	for	for	ADP
cana-5777	9	20	medical	medical	ADJ
cana-5777	9	21	classification	classification	NOUN
cana-5777	9	22	tasks	task	NOUN
cana-5777	9	23	,	,	PUNCT
cana-5777	9	24	proving	prove	VERB
cana-5777	9	25	its	its	PRON
cana-5777	9	26	effectiveness	effectiveness	NOUN
cana-5777	9	27	in	in	ADP
cana-5777	9	28	distinguishing	distinguish	VERB
cana-5777	9	29	between	between	ADP
cana-5777	9	30	benign	benign	ADJ
cana-5777	9	31	and	and	CCONJ
cana-5777	9	32	malignant	malignant	ADJ
cana-5777	9	33	tumors	tumor	NOUN
cana-5777	9	34	.	.	PUNCT
cana-5777	10	1	keywords	keyword	NOUN
cana-5777	10	2	:	:	PUNCT
cana-5777	10	3	breast	breast	NOUN
cana-5777	10	4	cancer	cancer	NOUN
cana-5777	10	5	classification	classification	NOUN
cana-5777	10	6	,	,	PUNCT
cana-5777	10	7	graph	graph	NOUN
cana-5777	10	8	neural	neural	ADJ
cana-5777	10	9	networks	network	NOUN
cana-5777	10	10	,	,	PUNCT
cana-5777	10	11	graph	graph	NOUN
cana-5777	10	12	attention	attention	NOUN
cana-5777	10	13	networks	network	NOUN
cana-5777	10	14	,	,	PUNCT
cana-5777	10	15	graphsage	graphsage	NOUN
cana-5777	10	16	,	,	PUNCT
cana-5777	10	17	graph	graph	NOUN
cana-5777	10	18	convolutional	convolutional	ADJ
cana-5777	10	19	network	network	NOUN
cana-5777	10	20	,	,	PUNCT
cana-5777	10	21	graph	graph	NOUN
cana-5777	10	22	isomorphism	isomorphism	NOUN
cana-5777	10	23	network	network	NOUN
cana-5777	10	24	,	,	PUNCT
cana-5777	10	25	k	k	NOUN
cana-5777	10	26	-	-	PUNCT
cana-5777	10	27	nearest	near	ADJ
cana-5777	10	28	neighbors	neighbor	NOUN
cana-5777	10	29	.	.	PUNCT
cana-5777	11	1	2020	2020	NUM
cana-5777	11	2	mathematics	mathematic	NOUN
cana-5777	11	3	subject	subject	ADJ
cana-5777	11	4	classification	classification	NOUN
cana-5777	11	5	:	:	PUNCT
cana-5777	11	6	05c90	05c90	NUM
cana-5777	11	7	,	,	PUNCT
cana-5777	11	8	05c62	05c62	NUM
cana-5777	11	9	,	,	PUNCT
cana-5777	11	10	68t07	68t07	NUM
cana-5777	11	11	,	,	PUNCT
cana-5777	11	12	90c35	90c35	NUM
cana-5777	11	13	1	1	NUM
cana-5777	11	14	.	.	PUNCT
cana-5777	12	1	introduction	introduction	NOUN
cana-5777	12	2	graph	graph	NOUN
cana-5777	12	3	theory	theory	NOUN
cana-5777	12	4	is	be	AUX
cana-5777	12	5	a	a	DET
cana-5777	12	6	powerful	powerful	ADJ
cana-5777	12	7	mathematical	mathematical	ADJ
cana-5777	12	8	approach	approach	NOUN
cana-5777	12	9	used	use	VERB
cana-5777	12	10	to	to	PART
cana-5777	12	11	model	model	VERB
cana-5777	12	12	relationships	relationship	NOUN
cana-5777	12	13	between	between	ADP
cana-5777	12	14	data	datum	NOUN
cana-5777	12	15	points	point	NOUN
cana-5777	12	16	,	,	PUNCT
cana-5777	12	17	making	make	VERB
cana-5777	12	18	it	it	PRON
cana-5777	12	19	highly	highly	ADV
cana-5777	12	20	effective	effective	ADJ
cana-5777	12	21	in	in	ADP
cana-5777	12	22	medical	medical	ADJ
cana-5777	12	23	data	datum	NOUN
cana-5777	12	24	analysis	analysis	NOUN
cana-5777	12	25	.	.	PUNCT
cana-5777	13	1	unlike	unlike	ADP
cana-5777	13	2	traditional	traditional	ADJ
cana-5777	13	3	machine	machine	NOUN
cana-5777	13	4	learning	learn	VERB
cana-5777	13	5	communications	communication	NOUN
cana-5777	13	6	on	on	ADP
cana-5777	13	7	applied	apply	VERB
cana-5777	13	8	nonlinear	nonlinear	ADJ
cana-5777	13	9	analysis	analysis	NOUN
cana-5777	13	10	issn	issn	NOUN
cana-5777	13	11	:	:	PUNCT
cana-5777	13	12	1074	1074	NUM
cana-5777	13	13	-	-	PUNCT
cana-5777	13	14	133x	133x	NUM
cana-5777	13	15	vol	vol	NOUN
cana-5777	13	16	31	31	NUM
cana-5777	13	17	no	no	NOUN
cana-5777	13	18	.	.	PUNCT
cana-5777	14	1	8s	8s	PROPN
cana-5777	14	2	(	(	PUNCT
cana-5777	14	3	2024	2024	NUM
cana-5777	14	4	)	)	PUNCT
cana-5777	14	5	https://internationalpubls.com	https://internationalpubls.com	X
cana-5777	14	6	1059	1059	NUM
cana-5777	14	7	(	(	PUNCT
cana-5777	14	8	ml	ml	NOUN
cana-5777	14	9	)	)	PUNCT
cana-5777	14	10	models	model	NOUN
cana-5777	14	11	,	,	PUNCT
cana-5777	14	12	which	which	PRON
cana-5777	14	13	treat	treat	VERB
cana-5777	14	14	each	each	DET
cana-5777	14	15	data	datum	NOUN
cana-5777	14	16	point	point	NOUN
cana-5777	14	17	as	as	ADP
cana-5777	14	18	independent	independent	ADJ
cana-5777	14	19	,	,	PUNCT
cana-5777	14	20	graph	graph	NOUN
cana-5777	14	21	-	-	PUNCT
cana-5777	14	22	based	base	VERB
cana-5777	14	23	methods	method	NOUN
cana-5777	14	24	capture	capture	VERB
cana-5777	14	25	the	the	DET
cana-5777	14	26	inherent	inherent	ADJ
cana-5777	14	27	connections	connection	NOUN
cana-5777	14	28	between	between	ADP
cana-5777	14	29	similar	similar	ADJ
cana-5777	14	30	cases	case	NOUN
cana-5777	14	31	,	,	PUNCT
cana-5777	14	32	allowing	allow	VERB
cana-5777	14	33	for	for	ADP
cana-5777	14	34	a	a	DET
cana-5777	14	35	more	more	ADV
cana-5777	14	36	structured	structured	ADJ
cana-5777	14	37	and	and	CCONJ
cana-5777	14	38	meaningful	meaningful	ADJ
cana-5777	14	39	representation	representation	NOUN
cana-5777	14	40	of	of	ADP
cana-5777	14	41	complex	complex	ADJ
cana-5777	14	42	datasets	dataset	NOUN
cana-5777	14	43	.	.	PUNCT
cana-5777	15	1	in	in	ADP
cana-5777	15	2	medical	medical	ADJ
cana-5777	15	3	diagnosis	diagnosis	NOUN
cana-5777	15	4	,	,	PUNCT
cana-5777	15	5	particularly	particularly	ADV
cana-5777	15	6	cancer	cancer	NOUN
cana-5777	15	7	classification	classification	NOUN
cana-5777	15	8	,	,	PUNCT
cana-5777	15	9	patient	patient	ADJ
cana-5777	15	10	data	datum	NOUN
cana-5777	15	11	is	be	AUX
cana-5777	15	12	often	often	ADV
cana-5777	15	13	interrelated	interrelate	VERB
cana-5777	15	14	,	,	PUNCT
cana-5777	15	15	as	as	SCONJ
cana-5777	15	16	certain	certain	ADJ
cana-5777	15	17	biological	biological	ADJ
cana-5777	15	18	markers	marker	NOUN
cana-5777	15	19	and	and	CCONJ
cana-5777	15	20	genetic	genetic	ADJ
cana-5777	15	21	features	feature	NOUN
cana-5777	15	22	exhibit	exhibit	VERB
cana-5777	15	23	strong	strong	ADJ
cana-5777	15	24	dependencies	dependency	NOUN
cana-5777	15	25	,	,	PUNCT
cana-5777	15	26	ignoring	ignore	VERB
cana-5777	15	27	these	these	DET
cana-5777	15	28	relationships	relationship	NOUN
cana-5777	15	29	can	can	AUX
cana-5777	15	30	limit	limit	VERB
cana-5777	15	31	the	the	DET
cana-5777	15	32	predictive	predictive	ADJ
cana-5777	15	33	power	power	NOUN
cana-5777	15	34	of	of	ADP
cana-5777	15	35	conventional	conventional	ADJ
cana-5777	15	36	ml	ml	NOUN
cana-5777	15	37	models	model	NOUN
cana-5777	15	38	.	.	PUNCT
cana-5777	16	1	graph	graph	NOUN
cana-5777	16	2	-	-	PUNCT
cana-5777	16	3	based	base	VERB
cana-5777	16	4	learning	learning	NOUN
cana-5777	16	5	approaches	approach	NOUN
cana-5777	16	6	offer	offer	VERB
cana-5777	16	7	a	a	DET
cana-5777	16	8	solution	solution	NOUN
cana-5777	16	9	by	by	ADP
cana-5777	16	10	structuring	structure	VERB
cana-5777	16	11	data	datum	NOUN
cana-5777	16	12	as	as	ADP
cana-5777	16	13	a	a	DET
cana-5777	16	14	graph	graph	NOUN
cana-5777	16	15	,	,	PUNCT
cana-5777	16	16	where	where	SCONJ
cana-5777	16	17	nodes	node	NOUN
cana-5777	16	18	represent	represent	VERB
cana-5777	16	19	patients	patient	NOUN
cana-5777	16	20	and	and	CCONJ
cana-5777	16	21	edges	edge	NOUN
cana-5777	16	22	signify	signify	VERB
cana-5777	16	23	feature	feature	NOUN
cana-5777	16	24	-	-	PUNCT
cana-5777	16	25	based	base	VERB
cana-5777	16	26	similarities	similarity	NOUN
cana-5777	16	27	.	.	PUNCT
cana-5777	17	1	this	this	DET
cana-5777	17	2	structure	structure	NOUN
cana-5777	17	3	enables	enable	VERB
cana-5777	17	4	relational	relational	ADJ
cana-5777	17	5	learning	learning	NOUN
cana-5777	17	6	,	,	PUNCT
cana-5777	17	7	allowing	allow	VERB
cana-5777	17	8	models	model	NOUN
cana-5777	17	9	to	to	PART
cana-5777	17	10	incorporate	incorporate	VERB
cana-5777	17	11	both	both	DET
cana-5777	17	12	individual	individual	ADJ
cana-5777	17	13	attributes	attribute	NOUN
cana-5777	17	14	and	and	CCONJ
cana-5777	17	15	the	the	DET
cana-5777	17	16	connections	connection	NOUN
cana-5777	17	17	between	between	ADP
cana-5777	17	18	patients	patient	NOUN
cana-5777	17	19	with	with	ADP
cana-5777	17	20	similar	similar	ADJ
cana-5777	17	21	characteristics	characteristic	NOUN
cana-5777	17	22	.	.	PUNCT
cana-5777	18	1	unlike	unlike	ADP
cana-5777	18	2	conventional	conventional	ADJ
cana-5777	18	3	ml	ml	NOUN
cana-5777	18	4	techniques	technique	NOUN
cana-5777	18	5	,	,	PUNCT
cana-5777	18	6	which	which	PRON
cana-5777	18	7	focus	focus	VERB
cana-5777	18	8	solely	solely	ADV
cana-5777	18	9	on	on	ADP
cana-5777	18	10	numerical	numerical	ADJ
cana-5777	18	11	patterns	pattern	NOUN
cana-5777	18	12	,	,	PUNCT
cana-5777	18	13	graph	graph	NOUN
cana-5777	18	14	-	-	PUNCT
cana-5777	18	15	based	base	VERB
cana-5777	18	16	models	model	NOUN
cana-5777	18	17	can	can	AUX
cana-5777	18	18	leverage	leverage	VERB
cana-5777	18	19	the	the	DET
cana-5777	18	20	structural	structural	ADJ
cana-5777	18	21	dependencies	dependency	NOUN
cana-5777	18	22	in	in	ADP
cana-5777	18	23	data	datum	NOUN
cana-5777	18	24	,	,	PUNCT
cana-5777	18	25	making	make	VERB
cana-5777	18	26	them	they	PRON
cana-5777	18	27	particularly	particularly	ADV
cana-5777	18	28	effective	effective	ADJ
cana-5777	18	29	for	for	ADP
cana-5777	18	30	challenging	challenge	VERB
cana-5777	18	31	classification	classification	NOUN
cana-5777	18	32	problems	problem	NOUN
cana-5777	18	33	,	,	PUNCT
cana-5777	18	34	such	such	ADJ
cana-5777	18	35	as	as	ADP
cana-5777	18	36	distinguishing	distinguish	VERB
cana-5777	18	37	between	between	ADP
cana-5777	18	38	benign	benign	ADJ
cana-5777	18	39	and	and	CCONJ
cana-5777	18	40	malignant	malignant	ADJ
cana-5777	18	41	tumours	tumour	NOUN
cana-5777	18	42	.	.	PUNCT
cana-5777	19	1	this	this	DET
cana-5777	19	2	structure	structure	NOUN
cana-5777	19	3	enables	enable	VERB
cana-5777	19	4	relational	relational	ADJ
cana-5777	19	5	learning	learning	NOUN
cana-5777	19	6	,	,	PUNCT
cana-5777	19	7	allowing	allow	VERB
cana-5777	19	8	the	the	DET
cana-5777	19	9	model	model	NOUN
cana-5777	19	10	to	to	PART
cana-5777	19	11	incorporate	incorporate	VERB
cana-5777	19	12	not	not	PART
cana-5777	19	13	only	only	ADV
cana-5777	19	14	individual	individual	ADJ
cana-5777	19	15	patient	patient	ADJ
cana-5777	19	16	characteristics	characteristic	NOUN
cana-5777	19	17	but	but	CCONJ
cana-5777	19	18	also	also	ADV
cana-5777	19	19	the	the	DET
cana-5777	19	20	relationships	relationship	NOUN
cana-5777	19	21	betweensimilar	betweensimilar	NOUN
cana-5777	19	22	cases	case	NOUN
cana-5777	19	23	.	.	PUNCT
cana-5777	20	1	recent	recent	ADJ
cana-5777	20	2	advancements	advancement	NOUN
cana-5777	20	3	in	in	ADP
cana-5777	20	4	graph	graph	NOUN
cana-5777	20	5	neural	neural	ADJ
cana-5777	20	6	networks	network	NOUN
cana-5777	20	7	(	(	PUNCT
cana-5777	20	8	gnns	gnns	NOUN
cana-5777	20	9	)	)	PUNCT
cana-5777	20	10	,	,	PUNCT
cana-5777	20	11	including	include	VERB
cana-5777	20	12	graph	graph	NOUN
cana-5777	20	13	convolutional	convolutional	ADJ
cana-5777	20	14	network	network	NOUN
cana-5777	20	15	(	(	PUNCT
cana-5777	20	16	gcn	gcn	NOUN
cana-5777	20	17	)	)	PUNCT
cana-5777	20	18	,	,	PUNCT
cana-5777	20	19	graph	graph	NOUN
cana-5777	20	20	isomorphism	isomorphism	NOUN
cana-5777	20	21	network	network	NOUN
cana-5777	20	22	(	(	PUNCT
cana-5777	20	23	gin	gin	NOUN
cana-5777	20	24	)	)	PUNCT
cana-5777	20	25	,	,	PUNCT
cana-5777	20	26	graph	graph	VERB
cana-5777	20	27	attention	attention	NOUN
cana-5777	20	28	networks	network	NOUN
cana-5777	20	29	(	(	PUNCT
cana-5777	20	30	gat	gat	NOUN
cana-5777	20	31	)	)	PUNCT
cana-5777	20	32	,	,	PUNCT
cana-5777	20	33	and	and	CCONJ
cana-5777	20	34	graphsagehave	graphsagehave	AUX
cana-5777	20	35	shown	show	VERB
cana-5777	20	36	remarkable	remarkable	ADJ
cana-5777	20	37	improvements	improvement	NOUN
cana-5777	20	38	in	in	ADP
cana-5777	20	39	classification	classification	NOUN
cana-5777	20	40	tasks	task	NOUN
cana-5777	20	41	by	by	ADP
cana-5777	20	42	effectively	effectively	ADV
cana-5777	20	43	leveraging	leverage	VERB
cana-5777	20	44	these	these	DET
cana-5777	20	45	relationships	relationship	NOUN
cana-5777	20	46	.	.	PUNCT
cana-5777	21	1	in	in	ADP
cana-5777	21	2	this	this	DET
cana-5777	21	3	study	study	NOUN
cana-5777	21	4	,	,	PUNCT
cana-5777	21	5	we	we	PRON
cana-5777	21	6	investigate	investigate	VERB
cana-5777	21	7	the	the	DET
cana-5777	21	8	potential	potential	NOUN
cana-5777	21	9	of	of	ADP
cana-5777	21	10	graph	graph	NOUN
cana-5777	21	11	-	-	PUNCT
cana-5777	21	12	based	base	VERB
cana-5777	21	13	learning	learning	NOUN
cana-5777	21	14	models	model	NOUN
cana-5777	21	15	for	for	ADP
cana-5777	21	16	breast	breast	NOUN
cana-5777	21	17	cancer	cancer	NOUN
cana-5777	21	18	classification	classification	NOUN
cana-5777	21	19	,	,	PUNCT
cana-5777	21	20	comparing	compare	VERB
cana-5777	21	21	their	their	PRON
cana-5777	21	22	performance	performance	NOUN
cana-5777	21	23	to	to	ADP
cana-5777	21	24	conventional	conventional	ADJ
cana-5777	21	25	ml	ml	NOUN
cana-5777	21	26	classifiers	classifier	NOUN
cana-5777	21	27	such	such	ADJ
cana-5777	21	28	as	as	ADP
cana-5777	21	29	decision	decision	NOUN
cana-5777	21	30	trees	tree	NOUN
cana-5777	21	31	,	,	PUNCT
cana-5777	21	32	random	random	ADJ
cana-5777	21	33	forest	forest	NOUN
cana-5777	21	34	,	,	PUNCT
cana-5777	21	35	lightgbm	lightgbm	ADJ
cana-5777	21	36	,	,	PUNCT
cana-5777	21	37	and	and	CCONJ
cana-5777	21	38	xgboost	xgboost	X
cana-5777	21	39	.	.	PUNCT
cana-5777	22	1	a	a	DET
cana-5777	22	2	graph	graph	NOUN
cana-5777	22	3	fork	fork	NOUN
cana-5777	22	4	-	-	PUNCT
cana-5777	22	5	nearest	near	ADJ
cana-5777	22	6	neighbours	neighbour	NOUN
cana-5777	22	7	(	(	PUNCT
cana-5777	22	8	knn	knn	PROPN
cana-5777	22	9	)	)	PUNCT
cana-5777	22	10	is	be	AUX
cana-5777	22	11	used	use	VERB
cana-5777	22	12	to	to	PART
cana-5777	22	13	establish	establish	VERB
cana-5777	22	14	connections	connection	NOUN
cana-5777	22	15	between	between	ADP
cana-5777	22	16	patients	patient	NOUN
cana-5777	22	17	with	with	ADP
cana-5777	22	18	similar	similar	ADJ
cana-5777	22	19	medical	medical	ADJ
cana-5777	22	20	profiles	profile	NOUN
cana-5777	22	21	,	,	PUNCT
cana-5777	22	22	enabling	enable	VERB
cana-5777	22	23	a	a	DET
cana-5777	22	24	more	more	ADV
cana-5777	22	25	structured	structured	ADJ
cana-5777	22	26	representation	representation	NOUN
cana-5777	22	27	of	of	ADP
cana-5777	22	28	the	the	DET
cana-5777	22	29	dataset	dataset	NOUN
cana-5777	22	30	.	.	PUNCT
cana-5777	23	1	to	to	PART
cana-5777	23	2	ensure	ensure	VERB
cana-5777	23	3	reliability	reliability	NOUN
cana-5777	23	4	,	,	PUNCT
cana-5777	23	5	we	we	PRON
cana-5777	23	6	conducted	conduct	VERB
cana-5777	23	7	robustness	robustness	NOUN
cana-5777	23	8	testing	testing	NOUN
cana-5777	23	9	by	by	ADP
cana-5777	23	10	varying	vary	VERB
cana-5777	23	11	knn	knn	PROPN
cana-5777	23	12	connectivity	connectivity	NOUN
cana-5777	23	13	,	,	PUNCT
cana-5777	23	14	introducing	introduce	VERB
cana-5777	23	15	feature	feature	NOUN
cana-5777	23	16	noise	noise	NOUN
cana-5777	23	17	,	,	PUNCT
cana-5777	23	18	and	and	CCONJ
cana-5777	23	19	shuffling	shuffle	VERB
cana-5777	23	20	the	the	DET
cana-5777	23	21	train	train	NOUN
cana-5777	23	22	-	-	PUNCT
cana-5777	23	23	test	test	NOUN
cana-5777	23	24	split	split	NOUN
cana-5777	23	25	.	.	PUNCT
cana-5777	24	1	experimental	experimental	ADJ
cana-5777	24	2	results	result	NOUN
cana-5777	24	3	show	show	VERB
cana-5777	24	4	that	that	SCONJ
cana-5777	24	5	gcn	gcn	NOUN
cana-5777	24	6	,	,	PUNCT
cana-5777	24	7	gin	gin	NOUN
cana-5777	24	8	,	,	PUNCT
cana-5777	24	9	gat	gat	NOUN
cana-5777	24	10	and	and	CCONJ
cana-5777	24	11	graphsage	graphsage	VERB
cana-5777	24	12	outperform	outperform	ADJ
cana-5777	24	13	traditional	traditional	ADJ
cana-5777	24	14	ml	ml	NOUN
cana-5777	24	15	models	model	NOUN
cana-5777	24	16	.	.	PUNCT
cana-5777	25	1	even	even	ADV
cana-5777	25	2	under	under	ADP
cana-5777	25	3	robustness	robustness	NOUN
cana-5777	25	4	testing	testing	NOUN
cana-5777	25	5	graph	graph	NOUN
cana-5777	25	6	-	-	PUNCT
cana-5777	25	7	based	base	VERB
cana-5777	25	8	models	model	NOUN
cana-5777	25	9	maintained	maintain	VERB
cana-5777	25	10	same	same	ADJ
cana-5777	25	11	accuracies	accuracy	NOUN
cana-5777	25	12	.	.	PUNCT
cana-5777	26	1	these	these	DET
cana-5777	26	2	findings	finding	NOUN
cana-5777	26	3	highlight	highlight	VERB
cana-5777	26	4	the	the	DET
cana-5777	26	5	potential	potential	NOUN
cana-5777	26	6	of	of	ADP
cana-5777	26	7	graph	graph	NOUN
cana-5777	26	8	-	-	PUNCT
cana-5777	26	9	based	base	VERB
cana-5777	26	10	learning	learning	NOUN
cana-5777	26	11	in	in	ADP
cana-5777	26	12	medical	medical	ADJ
cana-5777	26	13	diagnosis	diagnosis	NOUN
cana-5777	26	14	,	,	PUNCT
cana-5777	26	15	demonstrating	demonstrate	VERB
cana-5777	26	16	that	that	SCONJ
cana-5777	26	17	incorporating	incorporate	VERB
cana-5777	26	18	structural	structural	ADJ
cana-5777	26	19	dependencies	dependency	NOUN
cana-5777	26	20	enhances	enhance	VERB
cana-5777	26	21	model	model	NOUN
cana-5777	26	22	accuracy	accuracy	NOUN
cana-5777	26	23	and	and	CCONJ
cana-5777	26	24	interpretability	interpretability	NOUN
cana-5777	26	25	.	.	PUNCT
cana-5777	27	1	this	this	DET
cana-5777	27	2	research	research	NOUN
cana-5777	27	3	underscores	underscore	VERB
cana-5777	27	4	the	the	DET
cana-5777	27	5	importance	importance	NOUN
cana-5777	27	6	of	of	ADP
cana-5777	27	7	integrating	integrate	VERB
cana-5777	27	8	graph	graph	NOUN
cana-5777	27	9	theory	theory	NOUN
cana-5777	27	10	with	with	ADP
cana-5777	27	11	ml	ml	ADP
cana-5777	27	12	techniques	technique	NOUN
cana-5777	27	13	to	to	PART
cana-5777	27	14	develop	develop	VERB
cana-5777	27	15	more	more	ADV
cana-5777	27	16	reliable	reliable	ADJ
cana-5777	27	17	and	and	CCONJ
cana-5777	27	18	efficient	efficient	ADJ
cana-5777	27	19	diagnostic	diagnostic	ADJ
cana-5777	27	20	models	model	NOUN
cana-5777	27	21	for	for	ADP
cana-5777	27	22	medical	medical	ADJ
cana-5777	27	23	applications	application	NOUN
cana-5777	27	24	.	.	PUNCT
cana-5777	28	1	2	2	X
cana-5777	28	2	.	.	X
cana-5777	28	3	literature	literature	NOUN
cana-5777	28	4	review	review	NOUN
cana-5777	28	5	graph	graph	NOUN
cana-5777	28	6	-	-	PUNCT
cana-5777	28	7	based	base	VERB
cana-5777	28	8	learning	learning	NOUN
cana-5777	28	9	has	have	AUX
cana-5777	28	10	gained	gain	VERB
cana-5777	28	11	significant	significant	ADJ
cana-5777	28	12	attention	attention	NOUN
cana-5777	28	13	due	due	ADP
cana-5777	28	14	to	to	ADP
cana-5777	28	15	its	its	PRON
cana-5777	28	16	ability	ability	NOUN
cana-5777	28	17	to	to	PART
cana-5777	28	18	model	model	VERB
cana-5777	28	19	relationships	relationship	NOUN
cana-5777	28	20	and	and	CCONJ
cana-5777	28	21	dependencies	dependency	NOUN
cana-5777	28	22	between	between	ADP
cana-5777	28	23	data	datum	NOUN
cana-5777	28	24	points	point	NOUN
cana-5777	28	25	.	.	PUNCT
cana-5777	29	1	traditional	traditional	ADJ
cana-5777	29	2	machine	machine	NOUN
cana-5777	29	3	learning	learning	NOUN
cana-5777	29	4	models	model	NOUN
cana-5777	29	5	often	often	ADV
cana-5777	29	6	treat	treat	VERB
cana-5777	29	7	data	datum	NOUN
cana-5777	29	8	as	as	ADP
cana-5777	29	9	independent	independent	ADJ
cana-5777	29	10	entities	entity	NOUN
cana-5777	29	11	,	,	PUNCT
cana-5777	29	12	missing	miss	VERB
cana-5777	29	13	crucial	crucial	ADJ
cana-5777	29	14	structural	structural	ADJ
cana-5777	29	15	information	information	NOUN
cana-5777	29	16	.	.	PUNCT
cana-5777	30	1	in	in	ADP
cana-5777	30	2	contrast	contrast	NOUN
cana-5777	30	3	,	,	PUNCT
cana-5777	30	4	graph	graph	NOUN
cana-5777	30	5	-	-	PUNCT
cana-5777	30	6	based	base	VERB
cana-5777	30	7	models	model	NOUN
cana-5777	30	8	use	use	VERB
cana-5777	30	9	connectivity	connectivity	NOUN
cana-5777	30	10	patterns	pattern	NOUN
cana-5777	30	11	,	,	PUNCT
cana-5777	30	12	making	make	VERB
cana-5777	30	13	them	they	PRON
cana-5777	30	14	highly	highly	ADV
cana-5777	30	15	effective	effective	ADJ
cana-5777	30	16	in	in	ADP
cana-5777	30	17	domains	domain	NOUN
cana-5777	30	18	such	such	ADJ
cana-5777	30	19	as	as	ADP
cana-5777	30	20	social	social	ADJ
cana-5777	30	21	networks	network	NOUN
cana-5777	30	22	,	,	PUNCT
cana-5777	30	23	fraud	fraud	NOUN
cana-5777	30	24	detection	detection	NOUN
cana-5777	30	25	,	,	PUNCT
cana-5777	30	26	and	and	CCONJ
cana-5777	30	27	healthcare	healthcare	PROPN
cana-5777	30	28	.	.	PUNCT
cana-5777	31	1	these	these	DET
cana-5777	31	2	models	model	NOUN
cana-5777	31	3	leverage	leverage	NOUN
cana-5777	31	4	graph	graph	NOUN
cana-5777	31	5	structures	structure	NOUN
cana-5777	31	6	to	to	PART
cana-5777	31	7	improve	improve	VERB
cana-5777	31	8	classification	classification	NOUN
cana-5777	31	9	accuracy	accuracy	NOUN
cana-5777	31	10	,	,	PUNCT
cana-5777	31	11	link	link	NOUN
cana-5777	31	12	prediction	prediction	NOUN
cana-5777	31	13	,	,	PUNCT
cana-5777	31	14	and	and	CCONJ
cana-5777	31	15	representation	representation	NOUN
cana-5777	31	16	learning	learning	NOUN
cana-5777	31	17	.	.	PUNCT
cana-5777	32	1	the	the	DET
cana-5777	32	2	foundation	foundation	NOUN
cana-5777	32	3	of	of	ADP
cana-5777	32	4	graph	graph	NOUN
cana-5777	32	5	neural	neural	ADJ
cana-5777	32	6	networks	network	NOUN
cana-5777	32	7	(	(	PUNCT
cana-5777	32	8	gnns	gnns	NOUN
cana-5777	32	9	)	)	PUNCT
cana-5777	32	10	was	be	AUX
cana-5777	32	11	introduced	introduce	VERB
cana-5777	32	12	by	by	ADP
cana-5777	32	13	scarselli	scarselli	PROPN
cana-5777	32	14	et	et	PROPN
cana-5777	32	15	al	al	PROPN
cana-5777	32	16	.	.	PUNCT
cana-5777	33	1	[	[	X
cana-5777	33	2	1	1	NUM
cana-5777	33	3	]	]	PUNCT
cana-5777	33	4	,	,	PUNCT
cana-5777	33	5	who	who	PRON
cana-5777	33	6	proposed	propose	VERB
cana-5777	33	7	a	a	DET
cana-5777	33	8	framework	framework	NOUN
cana-5777	33	9	to	to	PART
cana-5777	33	10	extend	extend	VERB
cana-5777	33	11	traditional	traditional	ADJ
cana-5777	33	12	neural	neural	ADJ
cana-5777	33	13	networks	network	NOUN
cana-5777	33	14	to	to	ADP
cana-5777	33	15	graph	graph	NOUN
cana-5777	33	16	-	-	PUNCT
cana-5777	33	17	structured	structure	VERB
cana-5777	33	18	data	datum	NOUN
cana-5777	33	19	.	.	PUNCT
cana-5777	34	1	their	their	PRON
cana-5777	34	2	study	study	NOUN
cana-5777	34	3	introduced	introduce	VERB
cana-5777	34	4	a	a	DET
cana-5777	34	5	recursive	recursive	ADJ
cana-5777	34	6	approach	approach	NOUN
cana-5777	34	7	where	where	SCONJ
cana-5777	34	8	node	node	ADJ
cana-5777	34	9	representations	representation	NOUN
cana-5777	34	10	were	be	AUX
cana-5777	34	11	updated	update	VERB
cana-5777	34	12	iteratively	iteratively	ADV
cana-5777	34	13	based	base	VERB
cana-5777	34	14	on	on	ADP
cana-5777	34	15	communications	communication	NOUN
cana-5777	34	16	on	on	ADP
cana-5777	34	17	applied	apply	VERB
cana-5777	34	18	nonlinear	nonlinear	ADJ
cana-5777	34	19	analysis	analysis	NOUN
cana-5777	34	20	issn	issn	NOUN
cana-5777	34	21	:	:	PUNCT
cana-5777	34	22	1074	1074	NUM
cana-5777	34	23	-	-	PUNCT
cana-5777	34	24	133x	133x	NUM
cana-5777	34	25	vol	vol	NOUN
cana-5777	34	26	31	31	NUM
cana-5777	34	27	no	no	NOUN
cana-5777	34	28	.	.	PUNCT
cana-5777	35	1	8s	8s	PROPN
cana-5777	35	2	(	(	PUNCT
cana-5777	35	3	2024	2024	NUM
cana-5777	35	4	)	)	PUNCT
cana-5777	35	5	https://internationalpubls.com	https://internationalpubls.com	X
cana-5777	35	6	1060	1060	NUM
cana-5777	35	7	neighbouring	neighbouring	ADJ
cana-5777	35	8	nodes	node	NOUN
cana-5777	35	9	,	,	PUNCT
cana-5777	35	10	allowing	allow	VERB
cana-5777	35	11	for	for	ADP
cana-5777	35	12	effective	effective	ADJ
cana-5777	35	13	learning	learning	NOUN
cana-5777	35	14	from	from	ADP
cana-5777	35	15	complex	complex	ADJ
cana-5777	35	16	graph	graph	NOUN
cana-5777	35	17	relationships	relationship	NOUN
cana-5777	35	18	.	.	PUNCT
cana-5777	36	1	this	this	DET
cana-5777	36	2	pioneering	pioneer	VERB
cana-5777	36	3	work	work	NOUN
cana-5777	36	4	paved	pave	VERB
cana-5777	36	5	the	the	DET
cana-5777	36	6	way	way	NOUN
cana-5777	36	7	for	for	ADP
cana-5777	36	8	more	more	ADV
cana-5777	36	9	advanced	advanced	ADJ
cana-5777	36	10	graph	graph	NOUN
cana-5777	36	11	-	-	PUNCT
cana-5777	36	12	based	base	VERB
cana-5777	36	13	architectures	architecture	NOUN
cana-5777	36	14	that	that	PRON
cana-5777	36	15	have	have	AUX
cana-5777	36	16	since	since	ADV
cana-5777	36	17	been	be	AUX
cana-5777	36	18	widely	widely	ADV
cana-5777	36	19	applied	apply	VERB
cana-5777	36	20	in	in	ADP
cana-5777	36	21	various	various	ADJ
cana-5777	36	22	domains	domain	NOUN
cana-5777	36	23	.	.	PUNCT
cana-5777	37	1	one	one	NUM
cana-5777	37	2	of	of	ADP
cana-5777	37	3	the	the	DET
cana-5777	37	4	most	most	ADV
cana-5777	37	5	influential	influential	ADJ
cana-5777	37	6	advancements	advancement	NOUN
cana-5777	37	7	in	in	ADP
cana-5777	37	8	this	this	DET
cana-5777	37	9	field	field	NOUN
cana-5777	37	10	was	be	AUX
cana-5777	37	11	the	the	DET
cana-5777	37	12	introduction	introduction	NOUN
cana-5777	37	13	of	of	ADP
cana-5777	37	14	graph	graph	NOUN
cana-5777	37	15	convolutional	convolutional	ADJ
cana-5777	37	16	networks	network	NOUN
cana-5777	37	17	(	(	PUNCT
cana-5777	37	18	gcns	gcns	PROPN
cana-5777	37	19	)	)	PUNCT
cana-5777	37	20	by	by	ADP
cana-5777	37	21	kipf	kipf	VERB
cana-5777	37	22	and	and	CCONJ
cana-5777	37	23	welling	well	VERB
cana-5777	37	24	[	[	X
cana-5777	37	25	2	2	NUM
cana-5777	37	26	]	]	PUNCT
cana-5777	37	27	.	.	PUNCT
cana-5777	38	1	their	their	PRON
cana-5777	38	2	work	work	NOUN
cana-5777	38	3	extended	extend	VERB
cana-5777	38	4	traditional	traditional	ADJ
cana-5777	38	5	convolutional	convolutional	ADJ
cana-5777	38	6	neural	neural	ADJ
cana-5777	38	7	networks	network	NOUN
cana-5777	38	8	(	(	PUNCT
cana-5777	38	9	cnns	cnns	PROPN
cana-5777	38	10	)	)	PUNCT
cana-5777	38	11	to	to	PART
cana-5777	38	12	operate	operate	VERB
cana-5777	38	13	on	on	ADP
cana-5777	38	14	graph	graph	NOUN
cana-5777	38	15	-	-	PUNCT
cana-5777	38	16	structured	structure	VERB
cana-5777	38	17	data	datum	NOUN
cana-5777	38	18	using	use	VERB
cana-5777	38	19	spectral	spectral	ADJ
cana-5777	38	20	methods	method	NOUN
cana-5777	38	21	.	.	PUNCT
cana-5777	39	1	the	the	DET
cana-5777	39	2	key	key	ADJ
cana-5777	39	3	innovation	innovation	NOUN
cana-5777	39	4	in	in	ADP
cana-5777	39	5	gcns	gcns	PROPN
cana-5777	39	6	was	be	AUX
cana-5777	39	7	their	their	PRON
cana-5777	39	8	ability	ability	NOUN
cana-5777	39	9	to	to	PART
cana-5777	39	10	aggregate	aggregate	VERB
cana-5777	39	11	information	information	NOUN
cana-5777	39	12	from	from	ADP
cana-5777	39	13	neighbouring	neighbouring	ADJ
cana-5777	39	14	nodes	node	NOUN
cana-5777	39	15	in	in	ADP
cana-5777	39	16	a	a	DET
cana-5777	39	17	computationally	computationally	ADV
cana-5777	39	18	efficient	efficient	ADJ
cana-5777	39	19	manner	manner	NOUN
cana-5777	39	20	.	.	PUNCT
cana-5777	40	1	their	their	PRON
cana-5777	40	2	study	study	NOUN
cana-5777	40	3	demonstrated	demonstrate	VERB
cana-5777	40	4	that	that	SCONJ
cana-5777	40	5	gcns	gcns	PROPN
cana-5777	40	6	outperform	outperform	ADJ
cana-5777	40	7	traditional	traditional	ADJ
cana-5777	40	8	models	model	NOUN
cana-5777	40	9	in	in	ADP
cana-5777	40	10	semi	semi	ADJ
cana-5777	40	11	-	-	ADJ
cana-5777	40	12	supervised	supervised	ADJ
cana-5777	40	13	classification	classification	NOUN
cana-5777	40	14	tasks	task	NOUN
cana-5777	40	15	,	,	PUNCT
cana-5777	40	16	particularly	particularly	ADV
cana-5777	40	17	in	in	ADP
cana-5777	40	18	social	social	ADJ
cana-5777	40	19	networks	network	NOUN
cana-5777	40	20	and	and	CCONJ
cana-5777	40	21	citation	citation	NOUN
cana-5777	40	22	graphs	graph	NOUN
cana-5777	40	23	.	.	PUNCT
cana-5777	41	1	by	by	ADP
cana-5777	41	2	applying	apply	VERB
cana-5777	41	3	convolutional	convolutional	ADJ
cana-5777	41	4	operations	operation	NOUN
cana-5777	41	5	to	to	PART
cana-5777	41	6	graph	graph	VERB
cana-5777	41	7	data	datum	NOUN
cana-5777	41	8	,	,	PUNCT
cana-5777	41	9	they	they	PRON
cana-5777	41	10	improved	improve	VERB
cana-5777	41	11	node	node	ADJ
cana-5777	41	12	classification	classification	NOUN
cana-5777	41	13	accuracy	accuracy	NOUN
cana-5777	41	14	and	and	CCONJ
cana-5777	41	15	enhanced	enhance	VERB
cana-5777	41	16	feature	feature	NOUN
cana-5777	41	17	learning	learn	VERB
cana-5777	41	18	from	from	ADP
cana-5777	41	19	graph	graph	NOUN
cana-5777	41	20	structures	structure	NOUN
cana-5777	41	21	.	.	PUNCT
cana-5777	42	1	building	build	VERB
cana-5777	42	2	upon	upon	SCONJ
cana-5777	42	3	the	the	DET
cana-5777	42	4	limitations	limitation	NOUN
cana-5777	42	5	of	of	ADP
cana-5777	42	6	early	early	ADJ
cana-5777	42	7	gnns	gnns	NOUN
cana-5777	42	8	,	,	PUNCT
cana-5777	42	9	graph	graph	NOUN
cana-5777	42	10	isomorphism	isomorphism	NOUN
cana-5777	42	11	networks	network	NOUN
cana-5777	42	12	(	(	PUNCT
cana-5777	42	13	gin	gin	NOUN
cana-5777	42	14	)	)	PUNCT
cana-5777	42	15	were	be	AUX
cana-5777	42	16	introduced	introduce	VERB
cana-5777	42	17	by	by	ADP
cana-5777	42	18	xu	xu	PROPN
cana-5777	42	19	et	et	PROPN
cana-5777	42	20	al	al	PROPN
cana-5777	42	21	.	.	PUNCT
cana-5777	43	1	[	[	X
cana-5777	43	2	3	3	X
cana-5777	43	3	]	]	PUNCT
cana-5777	43	4	to	to	PART
cana-5777	43	5	address	address	VERB
cana-5777	43	6	the	the	DET
cana-5777	43	7	challenge	challenge	NOUN
cana-5777	43	8	of	of	ADP
cana-5777	43	9	distinguishing	distinguish	VERB
cana-5777	43	10	graph	graph	NOUN
cana-5777	43	11	structures	structure	NOUN
cana-5777	43	12	that	that	PRON
cana-5777	43	13	appear	appear	VERB
cana-5777	43	14	similar	similar	ADJ
cana-5777	43	15	but	but	CCONJ
cana-5777	43	16	have	have	VERB
cana-5777	43	17	different	different	ADJ
cana-5777	43	18	underlying	underlie	VERB
cana-5777	43	19	properties	property	NOUN
cana-5777	43	20	.	.	PUNCT
cana-5777	44	1	their	their	PRON
cana-5777	44	2	study	study	NOUN
cana-5777	44	3	proposed	propose	VERB
cana-5777	44	4	a	a	DET
cana-5777	44	5	more	more	ADV
cana-5777	44	6	powerful	powerful	ADJ
cana-5777	44	7	aggregation	aggregation	NOUN
cana-5777	44	8	function	function	NOUN
cana-5777	44	9	compared	compare	VERB
cana-5777	44	10	to	to	ADP
cana-5777	44	11	gcns	gcns	PROPN
cana-5777	44	12	.	.	PUNCT
cana-5777	45	1	another	another	DET
cana-5777	45	2	significant	significant	ADJ
cana-5777	45	3	contribution	contribution	NOUN
cana-5777	45	4	came	come	VERB
cana-5777	45	5	from	from	ADP
cana-5777	45	6	velicković	velicković	PROPN
cana-5777	45	7	et	et	PROPN
cana-5777	45	8	al	al	PROPN
cana-5777	45	9	.	.	PUNCT
cana-5777	46	1	[	[	X
cana-5777	46	2	4	4	NUM
cana-5777	46	3	]	]	PUNCT
cana-5777	46	4	,	,	PUNCT
cana-5777	46	5	who	who	PRON
cana-5777	46	6	introduced	introduce	VERB
cana-5777	46	7	graph	graph	NOUN
cana-5777	46	8	attention	attention	NOUN
cana-5777	46	9	networks	network	NOUN
cana-5777	46	10	(	(	PUNCT
cana-5777	46	11	gat	gat	NOUN
cana-5777	46	12	)	)	PUNCT
cana-5777	46	13	,	,	PUNCT
cana-5777	46	14	incorporating	incorporate	VERB
cana-5777	46	15	self	self	NOUN
cana-5777	46	16	-	-	PUNCT
cana-5777	46	17	attention	attention	NOUN
cana-5777	46	18	mechanisms	mechanism	NOUN
cana-5777	46	19	into	into	ADP
cana-5777	46	20	graph	graph	NOUN
cana-5777	46	21	-	-	PUNCT
cana-5777	46	22	based	base	VERB
cana-5777	46	23	learning	learning	NOUN
cana-5777	46	24	.	.	PUNCT
cana-5777	47	1	unlike	unlike	ADP
cana-5777	47	2	gcns	gcns	PROPN
cana-5777	47	3	and	and	CCONJ
cana-5777	47	4	gin	gin	NOUN
cana-5777	47	5	,	,	PUNCT
cana-5777	47	6	which	which	PRON
cana-5777	47	7	treat	treat	VERB
cana-5777	47	8	all	all	DET
cana-5777	47	9	neighbouring	neighbouring	ADJ
cana-5777	47	10	nodes	node	NOUN
cana-5777	47	11	equally	equally	ADV
cana-5777	47	12	,	,	PUNCT
cana-5777	47	13	gat	gat	NOUN
cana-5777	47	14	assigns	assign	VERB
cana-5777	47	15	different	different	ADJ
cana-5777	47	16	attention	attention	NOUN
cana-5777	47	17	weights	weight	NOUN
cana-5777	47	18	to	to	ADP
cana-5777	47	19	each	each	DET
cana-5777	47	20	neighbour	neighbour	NOUN
cana-5777	47	21	,	,	PUNCT
cana-5777	47	22	allowing	allow	VERB
cana-5777	47	23	the	the	DET
cana-5777	47	24	model	model	NOUN
cana-5777	47	25	to	to	PART
cana-5777	47	26	focus	focus	VERB
cana-5777	47	27	on	on	ADP
cana-5777	47	28	the	the	DET
cana-5777	47	29	most	most	ADV
cana-5777	47	30	relevant	relevant	ADJ
cana-5777	47	31	connections	connection	NOUN
cana-5777	47	32	.	.	PUNCT
cana-5777	48	1	this	this	DET
cana-5777	48	2	innovation	innovation	NOUN
cana-5777	48	3	led	lead	VERB
cana-5777	48	4	to	to	ADP
cana-5777	48	5	improved	improved	ADJ
cana-5777	48	6	performance	performance	NOUN
cana-5777	48	7	in	in	ADP
cana-5777	48	8	semi	semi	ADJ
cana-5777	48	9	-	-	ADJ
cana-5777	48	10	supervised	supervised	ADJ
cana-5777	48	11	classification	classification	NOUN
cana-5777	48	12	tasks	task	NOUN
cana-5777	48	13	,	,	PUNCT
cana-5777	48	14	as	as	SCONJ
cana-5777	48	15	it	it	PRON
cana-5777	48	16	better	well	ADV
cana-5777	48	17	captured	capture	VERB
cana-5777	48	18	the	the	DET
cana-5777	48	19	varying	vary	VERB
cana-5777	48	20	importance	importance	NOUN
cana-5777	48	21	of	of	ADP
cana-5777	48	22	different	different	ADJ
cana-5777	48	23	nodes	node	NOUN
cana-5777	48	24	.	.	PUNCT
cana-5777	49	1	in	in	ADP
cana-5777	49	2	parallel	parallel	NOUN
cana-5777	49	3	,	,	PUNCT
cana-5777	49	4	hamilton	hamilton	PROPN
cana-5777	49	5	et	et	PROPN
cana-5777	49	6	al	al	PROPN
cana-5777	49	7	.	.	PUNCT
cana-5777	50	1	[	[	X
cana-5777	50	2	5	5	NUM
cana-5777	50	3	]	]	PUNCT
cana-5777	50	4	developed	develop	VERB
cana-5777	50	5	graphsage	graphsage	NOUN
cana-5777	50	6	,	,	PUNCT
cana-5777	50	7	an	an	DET
cana-5777	50	8	inductive	inductive	ADJ
cana-5777	50	9	learning	learning	NOUN
cana-5777	50	10	framework	framework	NOUN
cana-5777	50	11	designed	design	VERB
cana-5777	50	12	to	to	PART
cana-5777	50	13	generate	generate	VERB
cana-5777	50	14	embeddings	embedding	NOUN
cana-5777	50	15	for	for	ADP
cana-5777	50	16	unseen	unseen	ADJ
cana-5777	50	17	nodes	node	NOUN
cana-5777	50	18	.	.	PUNCT
cana-5777	51	1	unlike	unlike	ADP
cana-5777	51	2	gcns	gcns	PROPN
cana-5777	51	3	,	,	PUNCT
cana-5777	51	4	which	which	PRON
cana-5777	51	5	require	require	VERB
cana-5777	51	6	the	the	DET
cana-5777	51	7	entire	entire	ADJ
cana-5777	51	8	graph	graph	NOUN
cana-5777	51	9	during	during	ADP
cana-5777	51	10	training	training	NOUN
cana-5777	51	11	,	,	PUNCT
cana-5777	51	12	graphsage	graphsage	NOUN
cana-5777	51	13	learns	learn	VERB
cana-5777	51	14	node	node	ADJ
cana-5777	51	15	representations	representation	NOUN
cana-5777	51	16	by	by	ADP
cana-5777	51	17	sampling	sample	VERB
cana-5777	51	18	and	and	CCONJ
cana-5777	51	19	aggregating	aggregate	VERB
cana-5777	51	20	information	information	NOUN
cana-5777	51	21	from	from	ADP
cana-5777	51	22	neighbouring	neighbouring	ADJ
cana-5777	51	23	nodes	node	NOUN
cana-5777	51	24	.	.	PUNCT
cana-5777	52	1	this	this	DET
cana-5777	52	2	approach	approach	NOUN
cana-5777	52	3	is	be	AUX
cana-5777	52	4	particularly	particularly	ADV
cana-5777	52	5	beneficial	beneficial	ADJ
cana-5777	52	6	for	for	ADP
cana-5777	52	7	dynamic	dynamic	ADJ
cana-5777	52	8	graphs	graph	NOUN
cana-5777	52	9	,	,	PUNCT
cana-5777	52	10	such	such	ADJ
cana-5777	52	11	as	as	ADP
cana-5777	52	12	evolving	evolve	VERB
cana-5777	52	13	social	social	ADJ
cana-5777	52	14	networks	network	NOUN
cana-5777	52	15	and	and	CCONJ
cana-5777	52	16	healthcare	healthcare	NOUN
cana-5777	52	17	records	record	NOUN
cana-5777	52	18	,	,	PUNCT
cana-5777	52	19	where	where	SCONJ
cana-5777	52	20	new	new	ADJ
cana-5777	52	21	nodes	node	NOUN
cana-5777	52	22	frequently	frequently	ADV
cana-5777	52	23	appear	appear	VERB
cana-5777	52	24	.	.	PUNCT
cana-5777	53	1	their	their	PRON
cana-5777	53	2	study	study	NOUN
cana-5777	53	3	demonstrated	demonstrate	VERB
cana-5777	53	4	that	that	SCONJ
cana-5777	53	5	graphsage	graphsage	NOUN
cana-5777	53	6	generalizes	generalize	VERB
cana-5777	53	7	well	well	ADV
cana-5777	53	8	to	to	ADP
cana-5777	53	9	unseen	unseen	ADJ
cana-5777	53	10	data	datum	NOUN
cana-5777	53	11	,	,	PUNCT
cana-5777	53	12	making	make	VERB
cana-5777	53	13	it	it	PRON
cana-5777	53	14	ideal	ideal	ADJ
cana-5777	53	15	for	for	ADP
cana-5777	53	16	large	large	ADJ
cana-5777	53	17	-	-	PUNCT
cana-5777	53	18	scale	scale	NOUN
cana-5777	53	19	applications	application	NOUN
cana-5777	53	20	requiring	require	VERB
cana-5777	53	21	real	real	ADJ
cana-5777	53	22	-	-	PUNCT
cana-5777	53	23	time	time	NOUN
cana-5777	53	24	learning	learning	NOUN
cana-5777	53	25	.	.	PUNCT
cana-5777	54	1	beyond	beyond	ADP
cana-5777	54	2	these	these	DET
cana-5777	54	3	foundational	foundational	ADJ
cana-5777	54	4	models	model	NOUN
cana-5777	54	5	,	,	PUNCT
cana-5777	54	6	researchers	researcher	NOUN
cana-5777	54	7	have	have	AUX
cana-5777	54	8	explored	explore	VERB
cana-5777	54	9	the	the	DET
cana-5777	54	10	application	application	NOUN
cana-5777	54	11	of	of	ADP
cana-5777	54	12	gnns	gnns	NOUN
cana-5777	54	13	in	in	ADP
cana-5777	54	14	healthcare	healthcare	PROPN
cana-5777	54	15	and	and	CCONJ
cana-5777	54	16	other	other	ADJ
cana-5777	54	17	fields	field	NOUN
cana-5777	54	18	.	.	PUNCT
cana-5777	55	1	zhang	zhang	PROPN
cana-5777	55	2	et	et	PROPN
cana-5777	55	3	al	al	PROPN
cana-5777	55	4	.	.	PUNCT
cana-5777	56	1	[	[	X
cana-5777	56	2	6	6	NUM
cana-5777	56	3	]	]	PUNCT
cana-5777	56	4	applied	apply	VERB
cana-5777	56	5	gnns	gnns	NOUN
cana-5777	56	6	to	to	ADP
cana-5777	56	7	personalized	personalized	ADJ
cana-5777	56	8	healthcare	healthcare	NOUN
cana-5777	56	9	by	by	ADP
cana-5777	56	10	developing	develop	VERB
cana-5777	56	11	a	a	DET
cana-5777	56	12	heterogeneous	heterogeneous	ADJ
cana-5777	56	13	graph	graph	NOUN
cana-5777	56	14	neural	neural	ADJ
cana-5777	56	15	network	network	NOUN
cana-5777	56	16	(	(	PUNCT
cana-5777	56	17	hgnn	hgnn	NOUN
cana-5777	56	18	)	)	PUNCT
cana-5777	56	19	that	that	DET
cana-5777	56	20	model	model	NOUN
cana-5777	56	21	’s	’s	PART
cana-5777	56	22	relationships	relationship	NOUN
cana-5777	56	23	between	between	ADP
cana-5777	56	24	medical	medical	ADJ
cana-5777	56	25	entities	entity	NOUN
cana-5777	56	26	such	such	ADJ
cana-5777	56	27	as	as	ADP
cana-5777	56	28	patients	patient	NOUN
cana-5777	56	29	,	,	PUNCT
cana-5777	56	30	diseases	disease	NOUN
cana-5777	56	31	,	,	PUNCT
cana-5777	56	32	and	and	CCONJ
cana-5777	56	33	treatments	treatment	NOUN
cana-5777	56	34	.	.	PUNCT
cana-5777	57	1	their	their	PRON
cana-5777	57	2	findings	finding	NOUN
cana-5777	57	3	showed	show	VERB
cana-5777	57	4	that	that	SCONJ
cana-5777	57	5	incorporating	incorporate	VERB
cana-5777	57	6	heterogeneous	heterogeneous	ADJ
cana-5777	57	7	graph	graph	NOUN
cana-5777	57	8	structures	structure	NOUN
cana-5777	57	9	significantly	significantly	ADV
cana-5777	57	10	enhances	enhance	VERB
cana-5777	57	11	diagnostic	diagnostic	ADJ
cana-5777	57	12	accuracy	accuracy	NOUN
cana-5777	57	13	.	.	PUNCT
cana-5777	58	1	similarly	similarly	ADV
cana-5777	58	2	,	,	PUNCT
cana-5777	58	3	wang	wang	PROPN
cana-5777	58	4	and	and	CCONJ
cana-5777	58	5	wang	wang	PROPN
cana-5777	59	1	[	[	X
cana-5777	59	2	7	7	NUM
cana-5777	59	3	]	]	PUNCT
cana-5777	59	4	investigated	investigate	VERB
cana-5777	59	5	the	the	DET
cana-5777	59	6	use	use	NOUN
cana-5777	59	7	of	of	ADP
cana-5777	59	8	gnns	gnns	NOUN
cana-5777	59	9	in	in	ADP
cana-5777	59	10	electronic	electronic	ADJ
cana-5777	59	11	health	health	NOUN
cana-5777	59	12	records	record	NOUN
cana-5777	59	13	,	,	PUNCT
cana-5777	59	14	demonstrating	demonstrate	VERB
cana-5777	59	15	that	that	SCONJ
cana-5777	59	16	structured	structured	ADJ
cana-5777	59	17	patient	patient	NOUN
cana-5777	59	18	relationships	relationship	NOUN
cana-5777	59	19	improve	improve	VERB
cana-5777	59	20	early	early	ADJ
cana-5777	59	21	disease	disease	NOUN
cana-5777	59	22	detection	detection	NOUN
cana-5777	59	23	and	and	CCONJ
cana-5777	59	24	risk	risk	NOUN
cana-5777	59	25	assessment	assessment	NOUN
cana-5777	59	26	.	.	PUNCT
cana-5777	60	1	the	the	DET
cana-5777	60	2	effectiveness	effectiveness	NOUN
cana-5777	60	3	of	of	ADP
cana-5777	60	4	graph	graph	NOUN
cana-5777	60	5	-	-	PUNCT
cana-5777	60	6	based	base	VERB
cana-5777	60	7	learning	learning	NOUN
cana-5777	60	8	has	have	AUX
cana-5777	60	9	also	also	ADV
cana-5777	60	10	been	be	AUX
cana-5777	60	11	explored	explore	VERB
cana-5777	60	12	in	in	ADP
cana-5777	60	13	fake	fake	ADJ
cana-5777	60	14	news	news	NOUN
cana-5777	60	15	detection	detection	NOUN
cana-5777	60	16	and	and	CCONJ
cana-5777	60	17	education	education	NOUN
cana-5777	60	18	analytics	analytic	NOUN
cana-5777	60	19	.	.	PUNCT
cana-5777	61	1	mahmud	mahmud	PROPN
cana-5777	61	2	et	et	PROPN
cana-5777	61	3	al	al	PROPN
cana-5777	61	4	.	.	PUNCT
cana-5777	62	1	[	[	X
cana-5777	62	2	8	8	NUM
cana-5777	62	3	]	]	PUNCT
cana-5777	62	4	conducted	conduct	VERB
cana-5777	62	5	a	a	DET
cana-5777	62	6	comparative	comparative	ADJ
cana-5777	62	7	study	study	NOUN
cana-5777	62	8	of	of	ADP
cana-5777	62	9	gnns	gnns	ADJ
cana-5777	62	10	versus	versus	ADP
cana-5777	62	11	traditional	traditional	ADJ
cana-5777	62	12	machine	machine	NOUN
cana-5777	62	13	learning	learning	NOUN
cana-5777	62	14	models	model	NOUN
cana-5777	62	15	for	for	ADP
cana-5777	62	16	fake	fake	ADJ
cana-5777	62	17	news	news	NOUN
cana-5777	62	18	detection	detection	NOUN
cana-5777	62	19	,	,	PUNCT
cana-5777	62	20	revealing	reveal	VERB
cana-5777	62	21	that	that	SCONJ
cana-5777	62	22	gnns	gnns	ADJ
cana-5777	62	23	outperform	outperform	ADJ
cana-5777	62	24	conventional	conventional	ADJ
cana-5777	62	25	classifiers	classifier	NOUN
cana-5777	62	26	by	by	ADP
cana-5777	62	27	leveraging	leverage	VERB
cana-5777	62	28	relationships	relationship	NOUN
cana-5777	62	29	between	between	ADP
cana-5777	62	30	articles	article	NOUN
cana-5777	62	31	and	and	CCONJ
cana-5777	62	32	sources	source	NOUN
cana-5777	62	33	.	.	PUNCT
cana-5777	63	1	likewise	likewise	ADV
cana-5777	63	2	,	,	PUNCT
cana-5777	63	3	communications	communication	NOUN
cana-5777	63	4	on	on	ADP
cana-5777	63	5	applied	apply	VERB
cana-5777	63	6	nonlinear	nonlinear	ADJ
cana-5777	63	7	analysis	analysis	NOUN
cana-5777	63	8	issn	issn	NOUN
cana-5777	63	9	:	:	PUNCT
cana-5777	63	10	1074	1074	NUM
cana-5777	63	11	-	-	PUNCT
cana-5777	63	12	133x	133x	NUM
cana-5777	63	13	vol	vol	NOUN
cana-5777	63	14	31	31	NUM
cana-5777	63	15	no	no	NOUN
cana-5777	63	16	.	.	PUNCT
cana-5777	64	1	8s	8s	PROPN
cana-5777	64	2	(	(	PUNCT
cana-5777	64	3	2024	2024	NUM
cana-5777	64	4	)	)	PUNCT
cana-5777	64	5	https://internationalpubls.com	https://internationalpubls.com	X
cana-5777	64	6	1061	1061	NUM
cana-5777	64	7	wang	wang	PROPN
cana-5777	64	8	et	et	PROPN
cana-5777	64	9	al	al	PROPN
cana-5777	64	10	.	.	PUNCT
cana-5777	65	1	[	[	X
cana-5777	65	2	9	9	NUM
cana-5777	65	3	]	]	PUNCT
cana-5777	65	4	introduced	introduce	VERB
cana-5777	65	5	a	a	DET
cana-5777	65	6	graph	graph	NOUN
cana-5777	65	7	-	-	PUNCT
cana-5777	65	8	based	base	VERB
cana-5777	65	9	ensemble	ensemble	ADJ
cana-5777	65	10	learning	learning	NOUN
cana-5777	65	11	method	method	NOUN
cana-5777	65	12	for	for	ADP
cana-5777	65	13	predicting	predict	VERB
cana-5777	65	14	student	student	NOUN
cana-5777	65	15	performance	performance	NOUN
cana-5777	65	16	,	,	PUNCT
cana-5777	65	17	showing	show	VERB
cana-5777	65	18	that	that	SCONJ
cana-5777	65	19	using	use	VERB
cana-5777	65	20	graph	graph	NOUN
cana-5777	65	21	representations	representation	NOUN
cana-5777	65	22	leads	lead	VERB
cana-5777	65	23	to	to	ADP
cana-5777	65	24	more	more	ADV
cana-5777	65	25	accurate	accurate	ADJ
cana-5777	65	26	predictions	prediction	NOUN
cana-5777	65	27	than	than	ADP
cana-5777	65	28	standalone	standalone	ADJ
cana-5777	65	29	ml	ml	NOUN
cana-5777	65	30	models	model	NOUN
cana-5777	65	31	.	.	PUNCT
cana-5777	66	1	a	a	DET
cana-5777	66	2	broader	broad	ADJ
cana-5777	66	3	discussion	discussion	NOUN
cana-5777	66	4	on	on	ADP
cana-5777	66	5	graph	graph	NOUN
cana-5777	66	6	-	-	PUNCT
cana-5777	66	7	based	base	VERB
cana-5777	66	8	learning	learning	NOUN
cana-5777	66	9	was	be	AUX
cana-5777	66	10	presented	present	VERB
cana-5777	66	11	by	by	ADP
cana-5777	66	12	shaila	shaila	PROPN
cana-5777	66	13	and	and	CCONJ
cana-5777	66	14	varsha	varsha	PROPN
cana-5777	66	15	[	[	X
cana-5777	66	16	10	10	NUM
cana-5777	66	17	]	]	PUNCT
cana-5777	66	18	,	,	PUNCT
cana-5777	66	19	who	who	PRON
cana-5777	66	20	provided	provide	VERB
cana-5777	66	21	a	a	DET
cana-5777	66	22	comprehensive	comprehensive	ADJ
cana-5777	66	23	review	review	NOUN
cana-5777	66	24	of	of	ADP
cana-5777	66	25	various	various	ADJ
cana-5777	66	26	graph	graph	NOUN
cana-5777	66	27	-	-	PUNCT
cana-5777	66	28	based	base	VERB
cana-5777	66	29	machine	machine	NOUN
cana-5777	66	30	learning	learning	NOUN
cana-5777	66	31	approaches	approach	NOUN
cana-5777	66	32	.	.	PUNCT
cana-5777	67	1	their	their	PRON
cana-5777	67	2	study	study	NOUN
cana-5777	67	3	highlighted	highlight	VERB
cana-5777	67	4	how	how	SCONJ
cana-5777	67	5	gnns	gnns	ADJ
cana-5777	67	6	,	,	PUNCT
cana-5777	67	7	including	include	VERB
cana-5777	67	8	gcn	gcn	NOUN
cana-5777	67	9	,	,	PUNCT
cana-5777	67	10	gin	gin	NOUN
cana-5777	67	11	,	,	PUNCT
cana-5777	67	12	gat	gat	NOUN
cana-5777	67	13	,	,	PUNCT
cana-5777	67	14	and	and	CCONJ
cana-5777	67	15	graphsage	graphsage	NOUN
cana-5777	67	16	,	,	PUNCT
cana-5777	67	17	are	be	AUX
cana-5777	67	18	transforming	transform	VERB
cana-5777	67	19	fields	field	NOUN
cana-5777	67	20	such	such	ADJ
cana-5777	67	21	as	as	ADP
cana-5777	67	22	recommendation	recommendation	NOUN
cana-5777	67	23	systems	system	NOUN
cana-5777	67	24	,	,	PUNCT
cana-5777	67	25	fraud	fraud	NOUN
cana-5777	67	26	detection	detection	NOUN
cana-5777	67	27	,	,	PUNCT
cana-5777	67	28	and	and	CCONJ
cana-5777	67	29	healthcare	healthcare	NOUN
cana-5777	67	30	analytics	analytic	NOUN
cana-5777	67	31	by	by	ADP
cana-5777	67	32	capturing	capture	VERB
cana-5777	67	33	complex	complex	ADJ
cana-5777	67	34	structural	structural	ADJ
cana-5777	67	35	dependencies	dependency	NOUN
cana-5777	67	36	.	.	PUNCT
cana-5777	68	1	our	our	PRON
cana-5777	68	2	study	study	NOUN
cana-5777	68	3	builds	build	VERB
cana-5777	68	4	upon	upon	SCONJ
cana-5777	68	5	these	these	DET
cana-5777	68	6	previous	previous	ADJ
cana-5777	68	7	works	work	NOUN
cana-5777	68	8	by	by	ADP
cana-5777	68	9	applying	apply	VERB
cana-5777	68	10	gin	gin	NOUN
cana-5777	68	11	,	,	PUNCT
cana-5777	68	12	gcn	gcn	NOUN
cana-5777	68	13	,	,	PUNCT
cana-5777	68	14	gat	gat	NOUN
cana-5777	68	15	,	,	PUNCT
cana-5777	68	16	and	and	CCONJ
cana-5777	68	17	graphsage	graphsage	VERB
cana-5777	68	18	to	to	ADP
cana-5777	68	19	breast	breast	NOUN
cana-5777	68	20	cancer	cancer	NOUN
cana-5777	68	21	classification	classification	NOUN
cana-5777	68	22	,	,	PUNCT
cana-5777	68	23	demonstrating	demonstrate	VERB
cana-5777	68	24	that	that	SCONJ
cana-5777	68	25	graph	graph	NOUN
cana-5777	68	26	-	-	PUNCT
cana-5777	68	27	based	base	VERB
cana-5777	68	28	models	model	NOUN
cana-5777	68	29	outperform	outperform	VERB
cana-5777	68	30	traditional	traditional	ADJ
cana-5777	68	31	classifiers	classifier	NOUN
cana-5777	68	32	.	.	PUNCT
cana-5777	69	1	we	we	PRON
cana-5777	69	2	construct	construct	VERB
cana-5777	69	3	a	a	DET
cana-5777	69	4	k	k	PROPN
cana-5777	69	5	-	-	PUNCT
cana-5777	69	6	nn	nn	NOUN
cana-5777	69	7	-	-	PUNCT
cana-5777	69	8	based	base	VERB
cana-5777	69	9	graph	graph	NOUN
cana-5777	69	10	to	to	PART
cana-5777	69	11	improve	improve	VERB
cana-5777	69	12	feature	feature	NOUN
cana-5777	69	13	learning	learning	NOUN
cana-5777	69	14	and	and	CCONJ
cana-5777	69	15	classification	classification	NOUN
cana-5777	69	16	accuracy	accuracy	NOUN
cana-5777	69	17	.	.	PUNCT
cana-5777	70	1	furthermore	furthermore	ADV
cana-5777	70	2	,	,	PUNCT
cana-5777	70	3	we	we	PRON
cana-5777	70	4	conduct	conduct	VERB
cana-5777	70	5	robustness	robustness	NOUN
cana-5777	70	6	testing	testing	NOUN
cana-5777	70	7	,	,	PUNCT
cana-5777	70	8	validating	validate	VERB
cana-5777	70	9	that	that	SCONJ
cana-5777	70	10	graph	graph	NOUN
cana-5777	70	11	-	-	PUNCT
cana-5777	70	12	based	base	VERB
cana-5777	70	13	learning	learning	NOUN
cana-5777	70	14	remains	remain	VERB
cana-5777	70	15	effective	effective	ADJ
cana-5777	70	16	under	under	ADP
cana-5777	70	17	reduced	reduced	ADJ
cana-5777	70	18	knn	knn	PROPN
cana-5777	70	19	connections	connection	NOUN
cana-5777	70	20	,	,	PUNCT
cana-5777	70	21	data	datum	NOUN
cana-5777	70	22	noise	noise	NOUN
cana-5777	70	23	,	,	PUNCT
cana-5777	70	24	and	and	CCONJ
cana-5777	70	25	varying	vary	VERB
cana-5777	70	26	train	train	NOUN
cana-5777	70	27	-	-	PUNCT
cana-5777	70	28	test	test	NOUN
cana-5777	70	29	splits	split	NOUN
cana-5777	70	30	.	.	PUNCT
cana-5777	71	1	these	these	DET
cana-5777	71	2	results	result	NOUN
cana-5777	71	3	support	support	VERB
cana-5777	71	4	the	the	DET
cana-5777	71	5	potential	potential	NOUN
cana-5777	71	6	of	of	ADP
cana-5777	71	7	graph	graph	NOUN
cana-5777	71	8	-	-	PUNCT
cana-5777	71	9	based	base	VERB
cana-5777	71	10	deep	deep	ADJ
cana-5777	71	11	learning	learning	NOUN
cana-5777	71	12	in	in	ADP
cana-5777	71	13	medical	medical	ADJ
cana-5777	71	14	diagnosis	diagnosis	NOUN
cana-5777	71	15	and	and	CCONJ
cana-5777	71	16	are	be	AUX
cana-5777	71	17	consistent	consistent	ADJ
cana-5777	71	18	with	with	ADP
cana-5777	71	19	previous	previous	ADJ
cana-5777	71	20	research	research	NOUN
cana-5777	71	21	.	.	PUNCT
cana-5777	72	1	3	3	X
cana-5777	72	2	.	.	X
cana-5777	72	3	methods	method	NOUN
cana-5777	72	4	and	and	CCONJ
cana-5777	72	5	materials	material	NOUN
cana-5777	72	6	in	in	ADP
cana-5777	72	7	this	this	DET
cana-5777	72	8	section	section	NOUN
cana-5777	72	9	,	,	PUNCT
cana-5777	72	10	the	the	DET
cana-5777	72	11	details	detail	NOUN
cana-5777	72	12	of	of	ADP
cana-5777	72	13	the	the	DET
cana-5777	72	14	data	data	NOUN
cana-5777	72	15	collection	collection	NOUN
cana-5777	72	16	and	and	CCONJ
cana-5777	72	17	methodology	methodology	NOUN
cana-5777	72	18	used	use	VERB
cana-5777	72	19	were	be	AUX
cana-5777	72	20	discussed	discuss	VERB
cana-5777	72	21	.	.	PUNCT
cana-5777	73	1	3.1	3.1	NUM
cana-5777	73	2	dataset	dataset	VERB
cana-5777	73	3	the	the	DET
cana-5777	73	4	kaggle	kaggle	ADJ
cana-5777	73	5	cancer	cancer	NOUN
cana-5777	73	6	dataset	dataset	NOUN
cana-5777	73	7	,	,	PUNCT
cana-5777	73	8	compiled	compile	VERB
cana-5777	73	9	by	by	ADP
cana-5777	73	10	erdem	erdem	PROPN
cana-5777	73	11	taha	taha	PROPN
cana-5777	74	1	[	[	X
cana-5777	74	2	11	11	NUM
cana-5777	74	3	]	]	PUNCT
cana-5777	74	4	,	,	PUNCT
cana-5777	74	5	consists	consist	VERB
cana-5777	74	6	of	of	ADP
cana-5777	74	7	570	570	NUM
cana-5777	74	8	samples	sample	NOUN
cana-5777	74	9	and	and	CCONJ
cana-5777	74	10	33	33	NUM
cana-5777	74	11	columns	column	NOUN
cana-5777	74	12	,	,	PUNCT
cana-5777	74	13	where	where	SCONJ
cana-5777	74	14	one	one	NUM
cana-5777	74	15	column	column	NOUN
cana-5777	74	16	represents	represent	VERB
cana-5777	74	17	the	the	DET
cana-5777	74	18	diagnosis	diagnosis	NOUN
cana-5777	74	19	(	(	PUNCT
cana-5777	74	20	malignant	malignant	ADJ
cana-5777	74	21	or	or	CCONJ
cana-5777	74	22	benign	benign	ADJ
cana-5777	74	23	)	)	PUNCT
cana-5777	74	24	,	,	PUNCT
cana-5777	74	25	and	and	CCONJ
cana-5777	74	26	the	the	DET
cana-5777	74	27	remaining	remain	VERB
cana-5777	74	28	32	32	NUM
cana-5777	74	29	numerical	numerical	ADJ
cana-5777	74	30	features	feature	NOUN
cana-5777	74	31	are	be	AUX
cana-5777	74	32	extracted	extract	VERB
cana-5777	74	33	from	from	ADP
cana-5777	74	34	image	image	NOUN
cana-5777	74	35	-	-	PUNCT
cana-5777	74	36	based	base	VERB
cana-5777	74	37	measurements	measurement	NOUN
cana-5777	74	38	of	of	ADP
cana-5777	74	39	cell	cell	NOUN
cana-5777	74	40	nuclei	nucleus	NOUN
cana-5777	74	41	.	.	PUNCT
cana-5777	75	1	these	these	DET
cana-5777	75	2	features	feature	NOUN
cana-5777	75	3	include	include	VERB
cana-5777	75	4	radius	radius	NOUN
cana-5777	75	5	,	,	PUNCT
cana-5777	75	6	texture	texture	ADJ
cana-5777	75	7	,	,	PUNCT
cana-5777	75	8	perimeter	perimeter	NOUN
cana-5777	75	9	,	,	PUNCT
cana-5777	75	10	area	area	NOUN
cana-5777	75	11	,	,	PUNCT
cana-5777	75	12	smoothness	smoothness	NOUN
cana-5777	75	13	,	,	PUNCT
cana-5777	75	14	compactness	compactness	NOUN
cana-5777	75	15	,	,	PUNCT
cana-5777	75	16	and	and	CCONJ
cana-5777	75	17	other	other	ADJ
cana-5777	75	18	shaperelated	shaperelate	VERB
cana-5777	75	19	metrics	metric	NOUN
cana-5777	75	20	.	.	PUNCT
cana-5777	76	1	in	in	ADP
cana-5777	76	2	this	this	DET
cana-5777	76	3	dataset	dataset	NOUN
cana-5777	76	4	,	,	PUNCT
cana-5777	76	5	the	the	DET
cana-5777	76	6	diagnosis	diagnosis	NOUN
cana-5777	76	7	column	column	NOUN
cana-5777	76	8	serves	serve	VERB
cana-5777	76	9	as	as	ADP
cana-5777	76	10	the	the	DET
cana-5777	76	11	dependent	dependent	ADJ
cana-5777	76	12	variable	variable	NOUN
cana-5777	76	13	,	,	PUNCT
cana-5777	76	14	while	while	SCONJ
cana-5777	76	15	the	the	DET
cana-5777	76	16	32	32	NUM
cana-5777	76	17	numerical	numerical	PROPN
cana-5777	76	18	features	feature	NOUN
cana-5777	76	19	act	act	VERB
cana-5777	76	20	as	as	ADP
cana-5777	76	21	independent	independent	ADJ
cana-5777	76	22	variables	variable	NOUN
cana-5777	76	23	.	.	PUNCT
cana-5777	77	1	to	to	PART
cana-5777	77	2	prepare	prepare	VERB
cana-5777	77	3	the	the	DET
cana-5777	77	4	data	datum	NOUN
cana-5777	77	5	for	for	ADP
cana-5777	77	6	machine	machine	NOUN
cana-5777	77	7	learning	learn	VERB
cana-5777	77	8	classification	classification	NOUN
cana-5777	77	9	,	,	PUNCT
cana-5777	77	10	the	the	DET
cana-5777	77	11	dataset	dataset	NOUN
cana-5777	77	12	is	be	AUX
cana-5777	77	13	split	split	VERB
cana-5777	77	14	into	into	ADP
cana-5777	77	15	training	training	NOUN
cana-5777	77	16	and	and	CCONJ
cana-5777	77	17	testing	testing	NOUN
cana-5777	77	18	sets	set	NOUN
cana-5777	77	19	using	use	VERB
cana-5777	77	20	a	a	DET
cana-5777	77	21	70	70	NUM
cana-5777	77	22	-	-	SYM
cana-5777	77	23	30	30	NUM
cana-5777	77	24	ratio	ratio	NOUN
cana-5777	77	25	.	.	PUNCT
cana-5777	78	1	this	this	DET
cana-5777	78	2	results	result	VERB
cana-5777	78	3	in	in	ADP
cana-5777	78	4	399	399	NUM
cana-5777	78	5	samples	sample	NOUN
cana-5777	78	6	(	(	PUNCT
cana-5777	78	7	70	70	NUM
cana-5777	78	8	%	%	NOUN
cana-5777	78	9	)	)	PUNCT
cana-5777	78	10	for	for	ADP
cana-5777	78	11	training	training	NOUN
cana-5777	78	12	and	and	CCONJ
cana-5777	78	13	171	171	NUM
cana-5777	78	14	samples	sample	NOUN
cana-5777	78	15	(	(	PUNCT
cana-5777	78	16	30	30	NUM
cana-5777	78	17	%	%	NOUN
cana-5777	78	18	)	)	PUNCT
cana-5777	78	19	for	for	ADP
cana-5777	78	20	testing	testing	NOUN
cana-5777	78	21	.	.	PUNCT
cana-5777	79	1	the	the	DET
cana-5777	79	2	splitting	splitting	NOUN
cana-5777	79	3	process	process	NOUN
cana-5777	79	4	ensures	ensure	VERB
cana-5777	79	5	a	a	DET
cana-5777	79	6	balanced	balanced	ADJ
cana-5777	79	7	approach	approach	NOUN
cana-5777	79	8	to	to	ADP
cana-5777	79	9	model	model	NOUN
cana-5777	79	10	evaluation	evaluation	NOUN
cana-5777	79	11	,	,	PUNCT
cana-5777	79	12	allowing	allow	VERB
cana-5777	79	13	the	the	DET
cana-5777	79	14	machine	machine	NOUN
cana-5777	79	15	learning	learn	VERB
cana-5777	79	16	algorithm	algorithm	NOUN
cana-5777	79	17	to	to	PART
cana-5777	79	18	learn	learn	VERB
cana-5777	79	19	from	from	ADP
cana-5777	79	20	a	a	DET
cana-5777	79	21	sufficient	sufficient	ADJ
cana-5777	79	22	number	number	NOUN
cana-5777	79	23	of	of	ADP
cana-5777	79	24	samples	sample	NOUN
cana-5777	79	25	while	while	SCONJ
cana-5777	79	26	also	also	ADV
cana-5777	79	27	being	be	AUX
cana-5777	79	28	tested	test	VERB
cana-5777	79	29	on	on	ADP
cana-5777	79	30	unseen	unseen	ADJ
cana-5777	79	31	data	datum	NOUN
cana-5777	79	32	.	.	PUNCT
cana-5777	80	1	this	this	DET
cana-5777	80	2	statistical	statistical	ADJ
cana-5777	80	3	arrangement	arrangement	NOUN
cana-5777	80	4	facilitates	facilitate	VERB
cana-5777	80	5	effective	effective	ADJ
cana-5777	80	6	classification	classification	NOUN
cana-5777	80	7	and	and	CCONJ
cana-5777	80	8	analysis	analysis	NOUN
cana-5777	80	9	of	of	ADP
cana-5777	80	10	breast	breast	NOUN
cana-5777	80	11	cancer	cancer	NOUN
cana-5777	80	12	biopsies	biopsy	NOUN
cana-5777	80	13	.	.	PUNCT
cana-5777	81	1	3.2	3.2	NUM
cana-5777	81	2	methods	method	NOUN
cana-5777	81	3	in	in	ADP
cana-5777	81	4	this	this	DET
cana-5777	81	5	research	research	NOUN
cana-5777	81	6	,	,	PUNCT
cana-5777	81	7	the	the	DET
cana-5777	81	8	authors	author	NOUN
cana-5777	81	9	compare	compare	VERB
cana-5777	81	10	traditional	traditional	ADJ
cana-5777	81	11	machine	machine	NOUN
cana-5777	81	12	learning	learn	VERB
cana-5777	81	13	classifiers	classifier	NOUN
cana-5777	81	14	with	with	ADP
cana-5777	81	15	graph	graph	NOUN
cana-5777	81	16	-	-	PUNCT
cana-5777	81	17	based	base	VERB
cana-5777	81	18	learning	learning	NOUN
cana-5777	81	19	models	model	NOUN
cana-5777	81	20	for	for	ADP
cana-5777	81	21	cancer	cancer	NOUN
cana-5777	81	22	classification	classification	NOUN
cana-5777	81	23	.	.	PUNCT
cana-5777	82	1	conventional	conventional	ADJ
cana-5777	82	2	ml	ml	NOUN
cana-5777	82	3	models	model	NOUN
cana-5777	82	4	operate	operate	VERB
cana-5777	82	5	on	on	ADP
cana-5777	82	6	tabular	tabular	PROPN
cana-5777	82	7	data	datum	NOUN
cana-5777	82	8	and	and	CCONJ
cana-5777	82	9	rely	rely	VERB
cana-5777	82	10	on	on	ADP
cana-5777	82	11	predefined	predefine	VERB
cana-5777	82	12	features	feature	NOUN
cana-5777	82	13	,	,	PUNCT
cana-5777	82	14	whereas	whereas	SCONJ
cana-5777	82	15	graph	graph	NOUN
cana-5777	82	16	-	-	PUNCT
cana-5777	82	17	based	base	VERB
cana-5777	82	18	models	model	NOUN
cana-5777	82	19	capture	capture	VERB
cana-5777	82	20	the	the	DET
cana-5777	82	21	structural	structural	ADJ
cana-5777	82	22	relationships	relationship	NOUN
cana-5777	82	23	between	between	ADP
cana-5777	82	24	data	datum	NOUN
cana-5777	82	25	points	point	NOUN
cana-5777	82	26	,	,	PUNCT
cana-5777	82	27	leveraging	leverage	VERB
cana-5777	82	28	graph	graph	NOUN
cana-5777	82	29	representations	representation	NOUN
cana-5777	82	30	to	to	PART
cana-5777	82	31	enhance	enhance	VERB
cana-5777	82	32	classification	classification	NOUN
cana-5777	82	33	accuracy	accuracy	NOUN
cana-5777	82	34	.	.	PUNCT
cana-5777	83	1	by	by	ADP
cana-5777	83	2	integrating	integrate	VERB
cana-5777	83	3	graph	graph	NOUN
cana-5777	83	4	learning	learn	VERB
cana-5777	83	5	techniques	technique	NOUN
cana-5777	83	6	(	(	PUNCT
cana-5777	83	7	gcn	gcn	NOUN
cana-5777	83	8	,	,	PUNCT
cana-5777	83	9	gin	gin	NOUN
cana-5777	83	10	,	,	PUNCT
cana-5777	83	11	graphsage	graphsage	NOUN
cana-5777	83	12	,	,	PUNCT
cana-5777	83	13	and	and	CCONJ
cana-5777	83	14	gat	gat	NOUN
cana-5777	83	15	)	)	PUNCT
cana-5777	83	16	we	we	PRON
cana-5777	83	17	constructed	construct	VERB
cana-5777	83	18	a	a	DET
cana-5777	83	19	knearest	knearest	NOUN
cana-5777	83	20	neighbor	neighbor	NOUN
cana-5777	83	21	(	(	PUNCT
cana-5777	83	22	knn	knn	PROPN
cana-5777	83	23	)	)	PUNCT
cana-5777	83	24	graph	graph	NOUN
cana-5777	83	25	where	where	SCONJ
cana-5777	83	26	nodes	node	NOUN
cana-5777	83	27	represent	represent	VERB
cana-5777	83	28	patients	patient	NOUN
cana-5777	83	29	and	and	CCONJ
cana-5777	83	30	edges	edge	NOUN
cana-5777	83	31	capture	capture	VERB
cana-5777	83	32	feature	feature	NOUN
cana-5777	83	33	similarities	similarity	NOUN
cana-5777	83	34	.	.	PUNCT
cana-5777	84	1	this	this	DET
cana-5777	84	2	graph	graph	NOUN
cana-5777	84	3	structure	structure	NOUN
cana-5777	84	4	enables	enable	VERB
cana-5777	84	5	deep	deep	ADJ
cana-5777	84	6	learning	learning	NOUN
cana-5777	84	7	models	model	NOUN
cana-5777	84	8	to	to	PART
cana-5777	84	9	effectively	effectively	ADV
cana-5777	84	10	propagate	propagate	VERB
cana-5777	84	11	information	information	NOUN
cana-5777	84	12	and	and	CCONJ
cana-5777	84	13	generate	generate	VERB
cana-5777	84	14	meaningful	meaningful	ADJ
cana-5777	84	15	representations	representation	NOUN
cana-5777	84	16	that	that	PRON
cana-5777	84	17	improve	improve	VERB
cana-5777	84	18	classification	classification	NOUN
cana-5777	84	19	outcomes	outcome	NOUN
cana-5777	84	20	.	.	PUNCT
cana-5777	85	1	communications	communication	NOUN
cana-5777	85	2	on	on	ADP
cana-5777	85	3	applied	apply	VERB
cana-5777	85	4	nonlinear	nonlinear	ADJ
cana-5777	85	5	analysis	analysis	NOUN
cana-5777	85	6	issn	issn	NOUN
cana-5777	85	7	:	:	PUNCT
cana-5777	85	8	1074	1074	NUM
cana-5777	85	9	-	-	PUNCT
cana-5777	85	10	133x	133x	NUM
cana-5777	85	11	vol	vol	NOUN
cana-5777	85	12	31	31	NUM
cana-5777	85	13	no	no	NOUN
cana-5777	85	14	.	.	PUNCT
cana-5777	86	1	8s	8s	PROPN
cana-5777	86	2	(	(	PUNCT
cana-5777	86	3	2024	2024	NUM
cana-5777	86	4	)	)	PUNCT
cana-5777	86	5	https://internationalpubls.com	https://internationalpubls.com	X
cana-5777	86	6	1062	1062	NUM
cana-5777	86	7	graph	graph	VERB
cana-5777	86	8	convolutional	convolutional	ADJ
cana-5777	86	9	networks	network	NOUN
cana-5777	86	10	(	(	PUNCT
cana-5777	86	11	gcns	gcns	PROPN
cana-5777	86	12	)	)	PUNCT
cana-5777	86	13	work	work	NOUN
cana-5777	86	14	by	by	ADP
cana-5777	86	15	aggregating	aggregate	VERB
cana-5777	86	16	information	information	NOUN
cana-5777	86	17	from	from	ADP
cana-5777	86	18	a	a	DET
cana-5777	86	19	node	node	NOUN
cana-5777	86	20	’s	’s	PART
cana-5777	86	21	neighbors	neighbor	NOUN
cana-5777	86	22	to	to	PART
cana-5777	86	23	learn	learn	VERB
cana-5777	86	24	meaningful	meaningful	ADJ
cana-5777	86	25	representations	representation	NOUN
cana-5777	86	26	.	.	PUNCT
cana-5777	87	1	the	the	DET
cana-5777	87	2	key	key	ADJ
cana-5777	87	3	mathematical	mathematical	ADJ
cana-5777	87	4	operation	operation	NOUN
cana-5777	87	5	in	in	ADP
cana-5777	87	6	gcns	gcns	PROPN
cana-5777	87	7	is	be	AUX
cana-5777	87	8	based	base	VERB
cana-5777	87	9	on	on	ADP
cana-5777	87	10	spectral	spectral	ADJ
cana-5777	87	11	graph	graph	NOUN
cana-5777	87	12	theory	theory	NOUN
cana-5777	87	13	,	,	PUNCT
cana-5777	87	14	where	where	SCONJ
cana-5777	87	15	the	the	DET
cana-5777	87	16	graph	graph	NOUN
cana-5777	87	17	structure	structure	NOUN
cana-5777	87	18	is	be	AUX
cana-5777	87	19	represented	represent	VERB
cana-5777	87	20	using	use	VERB
cana-5777	87	21	an	an	DET
cana-5777	87	22	adjacency	adjacency	NOUN
cana-5777	87	23	matrix	matrix	NOUN
cana-5777	87	24	(	(	PUNCT
cana-5777	87	25	a	a	NOUN
cana-5777	87	26	)	)	PUNCT
cana-5777	87	27	and	and	CCONJ
cana-5777	87	28	a	a	DET
cana-5777	87	29	degree	degree	NOUN
cana-5777	87	30	matrix	matrix	NOUN
cana-5777	87	31	(	(	PUNCT
cana-5777	87	32	d).the	d).the	DET
cana-5777	87	33	fundamental	fundamental	ADJ
cana-5777	87	34	formula	formula	NOUN
cana-5777	87	35	for	for	ADP
cana-5777	87	36	a	a	DET
cana-5777	87	37	single	single	ADJ
cana-5777	87	38	gcn	gcn	NOUN
cana-5777	87	39	layer	layer	NOUN
cana-5777	87	40	is	be	AUX
cana-5777	87	41	:	:	PUNCT
cana-5777	87	42	𝐻(𝑙+1	𝐻(𝑙+1	ADJ
cana-5777	87	43	)	)	PUNCT
cana-5777	87	44	=	=	SYM
cana-5777	87	45	𝜎(	𝜎(	NUM
cana-5777	87	46	�	�	NOUN
cana-5777	87	47	̂	̂	SYM
cana-5777	87	48	�	�	NOUN
cana-5777	87	49	−1/2	−1/2	ADJ
cana-5777	87	50	�	�	PROPN
cana-5777	87	51	̂	̂	SYM
cana-5777	87	52	�	�	PROPN
cana-5777	87	53	�	�	PROPN
cana-5777	87	54	̂	̂	NOUN
cana-5777	87	55	�	�	NOUN
cana-5777	87	56	−1/2𝐻(𝑙)𝑊(𝑙	−1/2𝐻(𝑙)𝑊(𝑙	NOUN
cana-5777	87	57	)	)	PUNCT
cana-5777	87	58	)	)	PUNCT
cana-5777	87	59	where	where	SCONJ
cana-5777	87	60	𝐻(𝑙	𝐻(𝑙	NUM
cana-5777	87	61	)	)	PUNCT
cana-5777	87	62	is	be	AUX
cana-5777	87	63	the	the	DET
cana-5777	87	64	feature	feature	NOUN
cana-5777	87	65	matrix	matrix	NOUN
cana-5777	87	66	at	at	ADP
cana-5777	87	67	layer	layer	NOUN
cana-5777	87	68	𝑙	𝑙	NOUN
cana-5777	87	69	,	,	PUNCT
cana-5777	87	70	with	with	ADP
cana-5777	87	71	each	each	DET
cana-5777	87	72	row	row	NOUN
cana-5777	87	73	representing	represent	VERB
cana-5777	87	74	a	a	DET
cana-5777	87	75	node	node	NOUN
cana-5777	87	76	’s	’s	PART
cana-5777	87	77	feature	feature	NOUN
cana-5777	87	78	,	,	PUNCT
cana-5777	87	79	�	�	PROPN
cana-5777	87	80	̂	̂	VERB
cana-5777	87	81	�	�	PROPN
cana-5777	87	82	=	=	SYM
cana-5777	87	83	𝐴	𝐴	PROPN
cana-5777	87	84	+	+	CCONJ
cana-5777	87	85	𝐼	𝐼	PROPN
cana-5777	87	86	is	be	AUX
cana-5777	87	87	the	the	DET
cana-5777	87	88	adjacency	adjacency	NOUN
cana-5777	87	89	matrix	matrix	NOUN
cana-5777	87	90	with	with	ADP
cana-5777	87	91	self	self	NOUN
cana-5777	87	92	-	-	PUNCT
cana-5777	87	93	loops	loop	NOUN
cana-5777	88	1	added	add	VERB
cana-5777	88	2	,	,	PUNCT
cana-5777	88	3	�	�	PROPN
cana-5777	88	4	̂	̂	VERB
cana-5777	88	5	�	�	NOUN
cana-5777	88	6	is	be	AUX
cana-5777	88	7	the	the	DET
cana-5777	88	8	degree	degree	NOUN
cana-5777	88	9	matrix	matrix	NOUN
cana-5777	88	10	of	of	ADP
cana-5777	88	11	�	�	PROPN
cana-5777	88	12	̂	̂	SYM
cana-5777	88	13	�	�	PROPN
cana-5777	88	14	,	,	PUNCT
cana-5777	88	15	which	which	PRON
cana-5777	88	16	helps	help	VERB
cana-5777	88	17	normalize	normalize	VERB
cana-5777	88	18	the	the	DET
cana-5777	88	19	aggregation	aggregation	NOUN
cana-5777	88	20	,	,	PUNCT
cana-5777	88	21	𝑊(𝑙	𝑊(𝑙	PROPN
cana-5777	88	22	)	)	PUNCT
cana-5777	88	23	is	be	AUX
cana-5777	88	24	the	the	DET
cana-5777	88	25	learnable	learnable	ADJ
cana-5777	88	26	weight	weight	NOUN
cana-5777	88	27	matrix	matrix	NOUN
cana-5777	88	28	for	for	ADP
cana-5777	88	29	layer	layer	NOUN
cana-5777	88	30	𝑙	𝑙	NOUN
cana-5777	88	31	,	,	PUNCT
cana-5777	88	32	and	and	CCONJ
cana-5777	88	33	𝜎	𝜎	PROPN
cana-5777	88	34	is	be	AUX
cana-5777	88	35	an	an	DET
cana-5777	88	36	activation	activation	NOUN
cana-5777	88	37	function	function	NOUN
cana-5777	88	38	.	.	PUNCT
cana-5777	89	1	graph	graph	NOUN
cana-5777	89	2	isomorphism	isomorphism	NOUN
cana-5777	89	3	networks	network	NOUN
cana-5777	89	4	(	(	PUNCT
cana-5777	89	5	gin	gin	NOUN
cana-5777	89	6	)	)	PUNCT
cana-5777	89	7	are	be	AUX
cana-5777	89	8	designed	design	VERB
cana-5777	89	9	to	to	PART
cana-5777	89	10	effectively	effectively	ADV
cana-5777	89	11	capture	capture	VERB
cana-5777	89	12	graph	graph	NOUN
cana-5777	89	13	structures	structure	NOUN
cana-5777	89	14	by	by	ADP
cana-5777	89	15	applying	apply	VERB
cana-5777	89	16	a	a	DET
cana-5777	89	17	powerful	powerful	ADJ
cana-5777	89	18	aggregation	aggregation	NOUN
cana-5777	89	19	function	function	NOUN
cana-5777	89	20	to	to	PART
cana-5777	89	21	node	node	NOUN
cana-5777	89	22	features	feature	NOUN
cana-5777	89	23	.	.	PUNCT
cana-5777	90	1	the	the	DET
cana-5777	90	2	mathematical	mathematical	ADJ
cana-5777	90	3	formulation	formulation	NOUN
cana-5777	90	4	of	of	ADP
cana-5777	90	5	gin	gin	NOUN
cana-5777	90	6	can	can	AUX
cana-5777	90	7	be	be	AUX
cana-5777	90	8	written	write	VERB
cana-5777	90	9	as	as	ADP
cana-5777	90	10	:	:	PUNCT
cana-5777	90	11	ℎ𝑣	ℎ𝑣	PROPN
cana-5777	90	12	(	(	PUNCT
cana-5777	90	13	𝑘	𝑘	NOUN
cana-5777	90	14	)	)	PUNCT
cana-5777	90	15	=	=	SYM
cana-5777	90	16	𝑀𝐿𝑃(𝑘)((1+∈	𝑀𝐿𝑃(𝑘)((1+∈	NUM
cana-5777	90	17	)	)	PUNCT
cana-5777	90	18	.	.	PUNCT
cana-5777	91	1	ℎ𝑣	ℎ𝑣	PROPN
cana-5777	91	2	(	(	PUNCT
cana-5777	91	3	𝑘−1	𝑘−1	PROPN
cana-5777	91	4	)	)	PUNCT
cana-5777	92	1	+	+	CCONJ
cana-5777	92	2	∑	∑	PUNCT
cana-5777	92	3	ℎ𝑢	ℎ𝑢	PROPN
cana-5777	92	4	(	(	PUNCT
cana-5777	92	5	𝑘−1	𝑘−1	PROPN
cana-5777	92	6	)	)	PUNCT
cana-5777	92	7	𝑢∈𝑁(𝑣	𝑢∈𝑁(𝑣	NOUN
cana-5777	92	8	)	)	PUNCT
cana-5777	92	9	where	where	SCONJ
cana-5777	92	10	ℎ𝑣	ℎ𝑣	ADV
cana-5777	92	11	(	(	PUNCT
cana-5777	92	12	𝑘	𝑘	NOUN
cana-5777	92	13	)	)	PUNCT
cana-5777	92	14	represents	represent	VERB
cana-5777	92	15	the	the	DET
cana-5777	92	16	node	node	NOUN
cana-5777	92	17	embedding	embed	VERB
cana-5777	92	18	at	at	ADP
cana-5777	92	19	layer	layer	NOUN
cana-5777	92	20	𝑘	𝑘	ADP
cana-5777	92	21	,	,	PUNCT
cana-5777	92	22	𝑀𝐿𝑃(𝑘)is	𝑀𝐿𝑃(𝑘)is	ADV
cana-5777	92	23	a	a	DET
cana-5777	92	24	multi	multi	ADJ
cana-5777	92	25	-	-	ADJ
cana-5777	92	26	layer	layer	ADJ
cana-5777	92	27	perceptron	perceptron	NOUN
cana-5777	92	28	that	that	PRON
cana-5777	92	29	transforms	transform	VERB
cana-5777	92	30	the	the	DET
cana-5777	92	31	aggregated	aggregated	ADJ
cana-5777	92	32	features,∈is	features,∈is	PROPN
cana-5777	92	33	a	a	DET
cana-5777	92	34	learnable	learnable	ADJ
cana-5777	92	35	parameter	parameter	NOUN
cana-5777	92	36	that	that	PRON
cana-5777	92	37	adjusts	adjust	VERB
cana-5777	92	38	the	the	DET
cana-5777	92	39	influence	influence	NOUN
cana-5777	92	40	of	of	ADP
cana-5777	92	41	the	the	DET
cana-5777	92	42	node	node	NOUN
cana-5777	92	43	's	's	PART
cana-5777	92	44	previous	previous	ADJ
cana-5777	92	45	state	state	NOUN
cana-5777	92	46	,	,	PUNCT
cana-5777	92	47	𝑁(𝑣)denotes	𝑁(𝑣)denote	VERB
cana-5777	92	48	the	the	DET
cana-5777	92	49	set	set	NOUN
cana-5777	92	50	of	of	ADP
cana-5777	92	51	neighboring	neighboring	NOUN
cana-5777	92	52	nodes	node	NOUN
cana-5777	92	53	of	of	ADP
cana-5777	92	54	node	node	ADJ
cana-5777	92	55	𝑣	𝑣	PROPN
cana-5777	92	56	and	and	CCONJ
cana-5777	92	57	the	the	DET
cana-5777	92	58	sum	sum	NOUN
cana-5777	92	59	operation(+	operation(+	ADV
cana-5777	92	60	)	)	PUNCT
cana-5777	92	61	ensures	ensure	VERB
cana-5777	92	62	that	that	SCONJ
cana-5777	92	63	the	the	DET
cana-5777	92	64	model	model	NOUN
cana-5777	92	65	captures	capture	VERB
cana-5777	92	66	structural	structural	ADJ
cana-5777	92	67	information	information	NOUN
cana-5777	92	68	efficiently	efficiently	ADV
cana-5777	92	69	.	.	PUNCT
cana-5777	93	1	graphsage	graphsage	NOUN
cana-5777	93	2	learns	learn	VERB
cana-5777	93	3	node	node	ADJ
cana-5777	93	4	embeddings	embedding	NOUN
cana-5777	93	5	by	by	ADP
cana-5777	93	6	aggregating	aggregate	VERB
cana-5777	93	7	information	information	NOUN
cana-5777	93	8	from	from	ADP
cana-5777	93	9	a	a	DET
cana-5777	93	10	node	node	NOUN
cana-5777	93	11	’s	’s	PART
cana-5777	93	12	neighbourhood	neighbourhood	NOUN
cana-5777	93	13	using	use	VERB
cana-5777	93	14	an	an	DET
cana-5777	93	15	inductive	inductive	ADJ
cana-5777	93	16	learning	learning	NOUN
cana-5777	93	17	approach	approach	NOUN
cana-5777	93	18	.	.	PUNCT
cana-5777	94	1	unlike	unlike	ADP
cana-5777	94	2	transductive	transductive	ADJ
cana-5777	94	3	methods	method	NOUN
cana-5777	94	4	that	that	PRON
cana-5777	94	5	require	require	VERB
cana-5777	94	6	the	the	DET
cana-5777	94	7	entire	entire	ADJ
cana-5777	94	8	graph	graph	NOUN
cana-5777	94	9	during	during	ADP
cana-5777	94	10	training	training	NOUN
cana-5777	94	11	,	,	PUNCT
cana-5777	94	12	graphsage	graphsage	NOUN
cana-5777	94	13	generalizes	generalize	VERB
cana-5777	94	14	to	to	ADP
cana-5777	94	15	unseen	unseen	ADJ
cana-5777	94	16	nodes	node	NOUN
cana-5777	94	17	by	by	ADP
cana-5777	94	18	iteratively	iteratively	ADV
cana-5777	94	19	updating	update	VERB
cana-5777	94	20	representations	representation	NOUN
cana-5777	94	21	based	base	VERB
cana-5777	94	22	on	on	ADP
cana-5777	94	23	sampled	sampled	ADJ
cana-5777	94	24	neighbours	neighbour	NOUN
cana-5777	94	25	.	.	PUNCT
cana-5777	95	1	the	the	DET
cana-5777	95	2	feature	feature	NOUN
cana-5777	95	3	update	update	NOUN
cana-5777	95	4	process	process	NOUN
cana-5777	95	5	follows	follow	VERB
cana-5777	95	6	the	the	DET
cana-5777	95	7	equation	equation	NOUN
cana-5777	95	8	ℎ𝑣	ℎ𝑣	VERB
cana-5777	95	9	(	(	PUNCT
cana-5777	95	10	𝑘	𝑘	NOUN
cana-5777	95	11	)	)	PUNCT
cana-5777	95	12	=	=	SYM
cana-5777	95	13	𝜎	𝜎	PROPN
cana-5777	95	14	(	(	PUNCT
cana-5777	95	15	𝑊(𝑘	𝑊(𝑘	NOUN
cana-5777	95	16	)	)	PUNCT
cana-5777	95	17	.	.	PUNCT
cana-5777	96	1	𝑓	𝑓	X
cana-5777	96	2	(	(	PUNCT
cana-5777	96	3	{	{	PUNCT
cana-5777	96	4	ℎ𝑢	ℎ𝑢	PROPN
cana-5777	96	5	(	(	PUNCT
cana-5777	96	6	𝑘−1	𝑘−1	PROPN
cana-5777	96	7	)	)	PUNCT
cana-5777	96	8	,	,	PUNCT
cana-5777	96	9	∀𝑢	∀𝑢	DET
cana-5777	96	10	∈	∈	PROPN
cana-5777	96	11	𝑁(𝑣	𝑁(𝑣	NUM
cana-5777	96	12	)	)	PUNCT
cana-5777	96	13	}	}	PUNCT
cana-5777	96	14	)	)	PUNCT
cana-5777	96	15	)	)	PUNCT
cana-5777	96	16	where	where	SCONJ
cana-5777	96	17	ℎ𝑣	ℎ𝑣	ADV
cana-5777	96	18	(	(	PUNCT
cana-5777	96	19	𝑘	𝑘	NOUN
cana-5777	96	20	)	)	PUNCT
cana-5777	96	21	is	be	AUX
cana-5777	96	22	the	the	DET
cana-5777	96	23	feature	feature	NOUN
cana-5777	96	24	embedding	embed	VERB
cana-5777	96	25	of	of	ADP
cana-5777	96	26	node	node	ADJ
cana-5777	96	27	𝒗	𝒗	PROPN
cana-5777	96	28	at	at	ADP
cana-5777	96	29	layer	layer	NOUN
cana-5777	96	30	𝑘	𝑘	ADP
cana-5777	96	31	,	,	PUNCT
cana-5777	96	32	𝑊(𝑘	𝑊(𝑘	VERB
cana-5777	96	33	)	)	PUNCT
cana-5777	96	34	is	be	AUX
cana-5777	96	35	the	the	DET
cana-5777	96	36	learnable	learnable	ADJ
cana-5777	96	37	weight	weight	NOUN
cana-5777	96	38	matrix	matrix	NOUN
cana-5777	96	39	,	,	PUNCT
cana-5777	96	40	𝜎	𝜎	PROPN
cana-5777	96	41	is	be	AUX
cana-5777	96	42	an	an	DET
cana-5777	96	43	activation	activation	NOUN
cana-5777	96	44	function	function	NOUN
cana-5777	96	45	,	,	PUNCT
cana-5777	96	46	and	and	CCONJ
cana-5777	96	47	𝑁(𝑣	𝑁(𝑣	X
cana-5777	96	48	)	)	PUNCT
cana-5777	96	49	denotes	denote	VERB
cana-5777	96	50	the	the	DET
cana-5777	96	51	neighbours	neighbour	NOUN
cana-5777	96	52	of	of	ADP
cana-5777	96	53	node	node	ADJ
cana-5777	96	54	𝑣.	𝑣.	NOUN
cana-5777	96	55	where	where	SCONJ
cana-5777	96	56	aggregation	aggregation	NOUN
cana-5777	96	57	function	function	NOUN
cana-5777	96	58	denoted	denote	VERB
cana-5777	96	59	as	as	ADP
cana-5777	96	60	𝑓	𝑓	PRON
cana-5777	96	61	and	and	CCONJ
cana-5777	96	62	it	it	PRON
cana-5777	96	63	can	can	AUX
cana-5777	96	64	take	take	VERB
cana-5777	96	65	different	different	ADJ
cana-5777	96	66	forms	form	NOUN
cana-5777	96	67	,	,	PUNCT
cana-5777	96	68	such	such	ADJ
cana-5777	96	69	as	as	ADP
cana-5777	96	70	mean	mean	ADJ
cana-5777	96	71	pooling	pooling	NOUN
cana-5777	96	72	,	,	PUNCT
cana-5777	96	73	long	long	ADJ
cana-5777	96	74	short	short	ADJ
cana-5777	96	75	-	-	PUNCT
cana-5777	96	76	term	term	NOUN
cana-5777	96	77	memory	memory	NOUN
cana-5777	96	78	(	(	PUNCT
cana-5777	96	79	lstm)-based	lstm)-based	ADJ
cana-5777	96	80	aggregation	aggregation	NOUN
cana-5777	96	81	,	,	PUNCT
cana-5777	96	82	or	or	CCONJ
cana-5777	96	83	max	max	PROPN
cana-5777	96	84	pooling	pooling	NOUN
cana-5777	96	85	,	,	PUNCT
cana-5777	96	86	ensuring	ensure	VERB
cana-5777	96	87	adaptability	adaptability	NOUN
cana-5777	96	88	in	in	ADP
cana-5777	96	89	different	different	ADJ
cana-5777	96	90	graph	graph	NOUN
cana-5777	96	91	structures	structure	NOUN
cana-5777	96	92	.	.	PUNCT
cana-5777	97	1	this	this	DET
cana-5777	97	2	methodology	methodology	NOUN
cana-5777	97	3	allows	allow	VERB
cana-5777	97	4	graphsage	graphsage	NOUN
cana-5777	97	5	to	to	PART
cana-5777	97	6	retain	retain	VERB
cana-5777	97	7	localized	localized	ADJ
cana-5777	97	8	feature	feature	NOUN
cana-5777	97	9	interactions	interaction	NOUN
cana-5777	97	10	while	while	SCONJ
cana-5777	97	11	capturing	capture	VERB
cana-5777	97	12	hierarchical	hierarchical	ADJ
cana-5777	97	13	information	information	NOUN
cana-5777	97	14	,	,	PUNCT
cana-5777	97	15	leading	lead	VERB
cana-5777	97	16	to	to	ADP
cana-5777	97	17	improved	improve	VERB
cana-5777	97	18	classification	classification	NOUN
cana-5777	97	19	accuracy	accuracy	NOUN
cana-5777	97	20	.	.	PUNCT
cana-5777	98	1	graph	graph	NOUN
cana-5777	98	2	attention	attention	NOUN
cana-5777	98	3	networks	network	NOUN
cana-5777	98	4	(	(	PUNCT
cana-5777	98	5	gat	gat	NOUN
cana-5777	98	6	)	)	PUNCT
cana-5777	98	7	extend	extend	VERB
cana-5777	98	8	traditional	traditional	ADJ
cana-5777	98	9	graph	graph	NOUN
cana-5777	98	10	convolution	convolution	NOUN
cana-5777	98	11	techniques	technique	NOUN
cana-5777	98	12	by	by	ADP
cana-5777	98	13	introducing	introduce	VERB
cana-5777	98	14	an	an	DET
cana-5777	98	15	attention	attention	NOUN
cana-5777	98	16	mechanism	mechanism	NOUN
cana-5777	98	17	that	that	PRON
cana-5777	98	18	assigns	assign	VERB
cana-5777	98	19	different	different	ADJ
cana-5777	98	20	importance	importance	NOUN
cana-5777	98	21	weights	weight	VERB
cana-5777	98	22	to	to	ADP
cana-5777	98	23	neighbouring	neighbouring	NOUN
cana-5777	98	24	nodes	node	NOUN
cana-5777	98	25	.	.	PUNCT
cana-5777	99	1	unlike	unlike	ADP
cana-5777	99	2	conventional	conventional	ADJ
cana-5777	99	3	aggregation	aggregation	NOUN
cana-5777	99	4	methods	method	NOUN
cana-5777	99	5	that	that	PRON
cana-5777	99	6	treat	treat	VERB
cana-5777	99	7	all	all	DET
cana-5777	99	8	neighbours	neighbour	NOUN
cana-5777	99	9	equally	equally	ADV
cana-5777	99	10	,	,	PUNCT
cana-5777	99	11	gat	gat	NOUN
cana-5777	99	12	dynamically	dynamically	ADV
cana-5777	99	13	determines	determine	VERB
cana-5777	99	14	the	the	DET
cana-5777	99	15	contribution	contribution	NOUN
cana-5777	99	16	of	of	ADP
cana-5777	99	17	each	each	DET
cana-5777	99	18	node	node	NOUN
cana-5777	99	19	through	through	ADP
cana-5777	99	20	a	a	DET
cana-5777	99	21	self	self	NOUN
cana-5777	99	22	-	-	PUNCT
cana-5777	99	23	attention	attention	NOUN
cana-5777	99	24	mechanism	mechanism	NOUN
cana-5777	99	25	.	.	PUNCT
cana-5777	100	1	the	the	DET
cana-5777	100	2	node	node	NOUN
cana-5777	100	3	embedding	embed	VERB
cana-5777	100	4	update	update	NOUN
cana-5777	100	5	follows	follow	VERB
cana-5777	100	6	ℎ𝑣	ℎ𝑣	PROPN
cana-5777	100	7	(	(	PUNCT
cana-5777	100	8	𝑘	𝑘	NOUN
cana-5777	100	9	)	)	PUNCT
cana-5777	100	10	=	=	SYM
cana-5777	100	11	𝜎	𝜎	PROPN
cana-5777	100	12	(	(	PUNCT
cana-5777	100	13	∑	∑	ADV
cana-5777	100	14	𝛼𝑣𝑢	𝛼𝑣𝑢	INTJ
cana-5777	100	15	(	(	PUNCT
cana-5777	100	16	𝑘	𝑘	NOUN
cana-5777	100	17	)	)	PUNCT
cana-5777	100	18	𝑊(𝑘)ℎ𝑢	𝑊(𝑘)ℎ𝑢	ADJ
cana-5777	100	19	(	(	PUNCT
cana-5777	100	20	𝑘−1	𝑘−1	PROPN
cana-5777	100	21	)	)	PUNCT
cana-5777	100	22	𝑢∈𝑁(𝑣	𝑢∈𝑁(𝑣	NOUN
cana-5777	100	23	)	)	PUNCT
cana-5777	100	24	)	)	PUNCT
cana-5777	101	1	communications	communication	NOUN
cana-5777	101	2	on	on	ADP
cana-5777	101	3	applied	apply	VERB
cana-5777	101	4	nonlinear	nonlinear	ADJ
cana-5777	101	5	analysis	analysis	NOUN
cana-5777	101	6	issn	issn	NOUN
cana-5777	101	7	:	:	PUNCT
cana-5777	101	8	1074	1074	NUM
cana-5777	101	9	-	-	PUNCT
cana-5777	101	10	133x	133x	NUM
cana-5777	101	11	vol	vol	NOUN
cana-5777	101	12	31	31	NUM
cana-5777	101	13	no	no	NOUN
cana-5777	101	14	.	.	PUNCT
cana-5777	102	1	8s	8s	PROPN
cana-5777	102	2	(	(	PUNCT
cana-5777	102	3	2024	2024	NUM
cana-5777	102	4	)	)	PUNCT
cana-5777	102	5	https://internationalpubls.com	https://internationalpubls.com	X
cana-5777	102	6	1063	1063	NUM
cana-5777	102	7	where	where	SCONJ
cana-5777	102	8	𝛼𝑣𝑢	𝛼𝑣𝑢	ADV
cana-5777	102	9	is	be	AUX
cana-5777	102	10	the	the	DET
cana-5777	102	11	attention	attention	NOUN
cana-5777	102	12	coefficient	coefficient	NOUN
cana-5777	102	13	computed	compute	VERB
cana-5777	102	14	as	as	ADP
cana-5777	102	15	𝛼𝑣𝑢	𝛼𝑣𝑢	ADV
cana-5777	102	16	=	=	SYM
cana-5777	102	17	e(ℒℛ(𝑎𝑇[𝑊ℎ𝑣‖𝑊ℎ𝑢	e(ℒℛ(𝑎𝑇[𝑊ℎ𝑣‖𝑊ℎ𝑢	PROPN
cana-5777	102	18	]	]	NOUN
cana-5777	102	19	)	)	PUNCT
cana-5777	102	20	)	)	PUNCT
cana-5777	103	1	∑	∑	PUNCT
cana-5777	103	2	𝑒(ℒℛ(𝑎𝑇[𝑊ℎ𝑣‖𝑊ℎ𝑗	𝑒(ℒℛ(𝑎𝑇[𝑊ℎ𝑣‖𝑊ℎ𝑗	NOUN
cana-5777	103	3	]	]	X
cana-5777	103	4	)	)	PUNCT
cana-5777	103	5	)	)	PUNCT
cana-5777	103	6	𝑗∈𝑁(𝑣	𝑗∈𝑁(𝑣	X
cana-5777	103	7	)	)	PUNCT
cana-5777	103	8	,	,	PUNCT
cana-5777	103	9	where	where	SCONJ
cana-5777	103	10	𝑎	𝑎	NOUN
cana-5777	103	11	is	be	AUX
cana-5777	103	12	a	a	DET
cana-5777	103	13	learnable	learnable	ADJ
cana-5777	103	14	attention	attention	NOUN
cana-5777	103	15	vector,𝑊	vector,𝑊	NOUN
cana-5777	103	16	is	be	AUX
cana-5777	103	17	a	a	DET
cana-5777	103	18	learnable	learnable	ADJ
cana-5777	103	19	weight	weight	NOUN
cana-5777	103	20	matrix	matrix	NOUN
cana-5777	103	21	applied	apply	VERB
cana-5777	103	22	to	to	ADP
cana-5777	103	23	the	the	DET
cana-5777	103	24	node	node	NOUN
cana-5777	103	25	features	feature	NOUN
cana-5777	103	26	,	,	PUNCT
cana-5777	103	27	||	||	PROPN
cana-5777	103	28	denotes	denotes	PROPN
cana-5777	103	29	concatenation	concatenation	PROPN
cana-5777	103	30	,	,	PUNCT
cana-5777	103	31	𝑁(𝑖	𝑁(𝑖	NOUN
cana-5777	103	32	)	)	PUNCT
cana-5777	103	33	is	be	AUX
cana-5777	103	34	the	the	DET
cana-5777	103	35	set	set	NOUN
cana-5777	103	36	of	of	ADP
cana-5777	103	37	neighbors	neighbor	NOUN
cana-5777	103	38	of	of	ADP
cana-5777	103	39	node	node	PROPN
cana-5777	103	40	𝑖	𝑖	SYM
cana-5777	103	41	,	,	PUNCT
cana-5777	103	42	ℒℛ	ℒℛ	PROPN
cana-5777	103	43	representsleakyreluit	representsleakyreluit	NOUN
cana-5777	103	44	is	be	AUX
cana-5777	103	45	a	a	DET
cana-5777	103	46	variant	variant	NOUN
cana-5777	103	47	of	of	ADP
cana-5777	103	48	the	the	DET
cana-5777	103	49	rectified	rectified	ADJ
cana-5777	103	50	linear	linear	NOUN
cana-5777	103	51	unit	unit	NOUN
cana-5777	103	52	(	(	PUNCT
cana-5777	103	53	relu	relu	NOUN
cana-5777	103	54	or	or	CCONJ
cana-5777	103	55	ℛ	ℛ	NOUN
cana-5777	103	56	)	)	PUNCT
cana-5777	103	57	activation	activation	NOUN
cana-5777	103	58	function	function	NOUN
cana-5777	103	59	used	use	VERB
cana-5777	103	60	in	in	ADP
cana-5777	103	61	neural	neural	ADJ
cana-5777	103	62	networks	network	NOUN
cana-5777	103	63	.	.	PUNCT
cana-5777	104	1	the	the	DET
cana-5777	104	2	ℒℛ	ℒℛ	PROPN
cana-5777	104	3	function	function	NOUN
cana-5777	104	4	is	be	AUX
cana-5777	104	5	defined	define	VERB
cana-5777	104	6	as	as	ADP
cana-5777	104	7	:	:	PUNCT
cana-5777	104	8	𝑓(𝑥	𝑓(𝑥	NOUN
cana-5777	104	9	)	)	PUNCT
cana-5777	105	1	=	=	PRON
cana-5777	105	2	{	{	PUNCT
cana-5777	105	3	𝑥	𝑥	X
cana-5777	105	4	,	,	PUNCT
cana-5777	105	5	𝑖𝑓	𝑖𝑓	ADP
cana-5777	105	6	𝑥	𝑥	NOUN
cana-5777	105	7	>	>	PUNCT
cana-5777	105	8	0	0	PUNCT
cana-5777	106	1	𝛼𝑥	𝛼𝑥	ADV
cana-5777	106	2	,	,	PUNCT
cana-5777	106	3	𝑖𝑓	𝑖𝑓	ADP
cana-5777	106	4	𝑥	𝑥	NOUN
cana-5777	106	5	≤	≤	NUM
cana-5777	106	6	0	0	NUM
cana-5777	106	7	where	where	SCONJ
cana-5777	106	8	,	,	PUNCT
cana-5777	106	9	𝑥	𝑥	PROPN
cana-5777	106	10	is	be	AUX
cana-5777	106	11	the	the	DET
cana-5777	106	12	input	input	NOUN
cana-5777	106	13	value	value	NOUN
cana-5777	106	14	and	and	CCONJ
cana-5777	106	15	𝛼is	𝛼is	PROPN
cana-5777	106	16	a	a	DET
cana-5777	106	17	small	small	ADJ
cana-5777	106	18	leak	leak	NOUN
cana-5777	106	19	factor	factor	NOUN
cana-5777	106	20	(	(	PUNCT
cana-5777	106	21	e.g.	e.g.	ADV
cana-5777	106	22	,	,	PUNCT
cana-5777	106	23	0.01	0.01	NUM
cana-5777	106	24	)	)	PUNCT
cana-5777	106	25	,	,	PUNCT
cana-5777	106	26	which	which	PRON
cana-5777	106	27	determines	determine	VERB
cana-5777	106	28	how	how	SCONJ
cana-5777	106	29	much	much	ADJ
cana-5777	106	30	negative	negative	ADJ
cana-5777	106	31	values	value	NOUN
cana-5777	106	32	are	be	AUX
cana-5777	106	33	allowed	allow	VERB
cana-5777	106	34	to	to	PART
cana-5777	106	35	pass	pass	VERB
cana-5777	106	36	through	through	ADP
cana-5777	106	37	instead	instead	ADV
cana-5777	106	38	of	of	ADP
cana-5777	106	39	being	be	AUX
cana-5777	106	40	completely	completely	ADV
cana-5777	106	41	zeroed	zero	VERB
cana-5777	106	42	out	out	ADP
cana-5777	106	43	.	.	PUNCT
cana-5777	107	1	the	the	DET
cana-5777	107	2	attention	attention	NOUN
cana-5777	107	3	mechanism	mechanism	NOUN
cana-5777	107	4	enhances	enhance	VERB
cana-5777	107	5	learning	learn	VERB
cana-5777	107	6	by	by	ADP
cana-5777	107	7	selectively	selectively	ADV
cana-5777	107	8	focusing	focus	VERB
cana-5777	107	9	on	on	ADP
cana-5777	107	10	the	the	DET
cana-5777	107	11	most	most	ADV
cana-5777	107	12	relevant	relevant	ADJ
cana-5777	107	13	neighbors	neighbor	NOUN
cana-5777	107	14	,	,	PUNCT
cana-5777	107	15	improving	improve	VERB
cana-5777	107	16	node	node	ADJ
cana-5777	107	17	classification	classification	NOUN
cana-5777	107	18	by	by	ADP
cana-5777	107	19	preserving	preserve	VERB
cana-5777	107	20	critical	critical	ADJ
cana-5777	107	21	relationships	relationship	NOUN
cana-5777	107	22	within	within	ADP
cana-5777	107	23	the	the	DET
cana-5777	107	24	graph	graph	NOUN
cana-5777	107	25	.	.	PUNCT
cana-5777	108	1	for	for	ADP
cana-5777	108	2	comparison	comparison	NOUN
cana-5777	108	3	,	,	PUNCT
cana-5777	108	4	we	we	PRON
cana-5777	108	5	also	also	ADV
cana-5777	108	6	implemented	implement	VERB
cana-5777	108	7	traditional	traditional	ADJ
cana-5777	108	8	ml	ml	NOUN
cana-5777	108	9	classifiers	classifier	NOUN
cana-5777	108	10	,	,	PUNCT
cana-5777	108	11	including	include	VERB
cana-5777	108	12	decision	decision	NOUN
cana-5777	108	13	trees	tree	NOUN
cana-5777	108	14	,	,	PUNCT
cana-5777	108	15	random	random	ADJ
cana-5777	108	16	forest	forest	NOUN
cana-5777	108	17	,	,	PUNCT
cana-5777	108	18	adaboost	adaboost	ADV
cana-5777	108	19	,	,	PUNCT
cana-5777	108	20	xgboost	xgboost	ADV
cana-5777	108	21	,	,	PUNCT
cana-5777	108	22	and	and	CCONJ
cana-5777	108	23	lightgbm	lightgbm	VERB
cana-5777	108	24	etcetera	etcetera	NOUN
cana-5777	108	25	.	.	PUNCT
cana-5777	109	1	among	among	ADP
cana-5777	109	2	these	these	DET
cana-5777	109	3	models	model	NOUN
cana-5777	109	4	,	,	PUNCT
cana-5777	109	5	adaboost	adaboost	ADV
cana-5777	109	6	,	,	PUNCT
cana-5777	109	7	and	and	CCONJ
cana-5777	109	8	lightgbm	lightgbm	NOUN
cana-5777	109	9	demonstrated	demonstrate	VERB
cana-5777	109	10	the	the	DET
cana-5777	109	11	highest	high	ADJ
cana-5777	109	12	accuracy	accuracy	NOUN
cana-5777	109	13	due	due	ADP
cana-5777	109	14	to	to	ADP
cana-5777	109	15	its	its	PRON
cana-5777	109	16	gradient	gradient	NOUN
cana-5777	109	17	boosting	boost	VERB
cana-5777	109	18	framework	framework	NOUN
cana-5777	109	19	,	,	PUNCT
cana-5777	109	20	which	which	PRON
cana-5777	109	21	iteratively	iteratively	ADV
cana-5777	109	22	improves	improve	VERB
cana-5777	109	23	weak	weak	ADJ
cana-5777	109	24	learners	learner	NOUN
cana-5777	109	25	.	.	PUNCT
cana-5777	110	1	the	the	DET
cana-5777	110	2	key	key	ADJ
cana-5777	110	3	factors	factor	NOUN
cana-5777	110	4	that	that	PRON
cana-5777	110	5	contributed	contribute	VERB
cana-5777	110	6	to	to	ADP
cana-5777	110	7	achieving	achieve	VERB
cana-5777	110	8	100	100	NUM
cana-5777	110	9	%	%	NOUN
cana-5777	110	10	accuracy	accuracy	NOUN
cana-5777	110	11	in	in	ADP
cana-5777	110	12	this	this	DET
cana-5777	110	13	study	study	NOUN
cana-5777	110	14	include	include	VERB
cana-5777	110	15	the	the	DET
cana-5777	110	16	structural	structural	ADJ
cana-5777	110	17	information	information	NOUN
cana-5777	110	18	captured	capture	VERB
cana-5777	110	19	by	by	ADP
cana-5777	110	20	the	the	DET
cana-5777	110	21	graph	graph	NOUN
cana-5777	110	22	,	,	PUNCT
cana-5777	110	23	graph	graph	NOUN
cana-5777	110	24	-	-	PUNCT
cana-5777	110	25	based	base	VERB
cana-5777	110	26	models	model	NOUN
cana-5777	110	27	provided	provide	VERB
cana-5777	110	28	a	a	DET
cana-5777	110	29	deeper	deep	ADJ
cana-5777	110	30	understanding	understanding	NOUN
cana-5777	110	31	of	of	ADP
cana-5777	110	32	the	the	DET
cana-5777	110	33	dataset	dataset	NOUN
cana-5777	110	34	by	by	ADP
cana-5777	110	35	representing	represent	VERB
cana-5777	110	36	relationships	relationship	NOUN
cana-5777	110	37	that	that	PRON
cana-5777	110	38	traditional	traditional	ADJ
cana-5777	110	39	ml	ml	NOUN
cana-5777	110	40	models	model	NOUN
cana-5777	110	41	could	could	AUX
cana-5777	110	42	not	not	PART
cana-5777	110	43	capture	capture	VERB
cana-5777	110	44	.	.	PUNCT
cana-5777	111	1	4	4	NUM
cana-5777	111	2	results	result	NOUN
cana-5777	111	3	and	and	CCONJ
cana-5777	111	4	discussions	discussion	NOUN
cana-5777	111	5	the	the	DET
cana-5777	111	6	classification	classification	NOUN
cana-5777	111	7	results	result	NOUN
cana-5777	111	8	obtained	obtain	VERB
cana-5777	111	9	using	use	VERB
cana-5777	111	10	graph	graph	NOUN
cana-5777	111	11	-	-	PUNCT
cana-5777	111	12	based	base	VERB
cana-5777	111	13	models	model	NOUN
cana-5777	111	14	,	,	PUNCT
cana-5777	111	15	particularly	particularly	ADV
cana-5777	111	16	graph	graph	VERB
cana-5777	111	17	neural	neural	ADJ
cana-5777	111	18	networks	network	NOUN
cana-5777	111	19	(	(	PUNCT
cana-5777	111	20	gnns	gnns	NOUN
cana-5777	111	21	)	)	PUNCT
cana-5777	111	22	,	,	PUNCT
cana-5777	111	23	were	be	AUX
cana-5777	111	24	compared	compare	VERB
cana-5777	111	25	with	with	ADP
cana-5777	111	26	traditional	traditional	ADJ
cana-5777	111	27	machine	machine	NOUN
cana-5777	111	28	learning	learning	NOUN
cana-5777	111	29	(	(	PUNCT
cana-5777	111	30	ml	ml	NOUN
cana-5777	111	31	)	)	PUNCT
cana-5777	111	32	approaches	approach	NOUN
cana-5777	111	33	to	to	PART
cana-5777	111	34	evaluate	evaluate	VERB
cana-5777	111	35	their	their	PRON
cana-5777	111	36	effectiveness	effectiveness	NOUN
cana-5777	111	37	in	in	ADP
cana-5777	111	38	distinguishing	distinguish	VERB
cana-5777	111	39	between	between	ADP
cana-5777	111	40	benign	benign	ADJ
cana-5777	111	41	and	and	CCONJ
cana-5777	111	42	malignant	malignant	ADJ
cana-5777	111	43	tumour	tumour	NOUN
cana-5777	111	44	data	datum	NOUN
cana-5777	111	45	.	.	PUNCT
cana-5777	112	1	the	the	DET
cana-5777	112	2	dataset	dataset	NOUN
cana-5777	112	3	was	be	AUX
cana-5777	112	4	first	first	ADV
cana-5777	112	5	converted	convert	VERB
cana-5777	112	6	into	into	ADP
cana-5777	112	7	a	a	DET
cana-5777	112	8	graph	graph	NOUN
cana-5777	112	9	structure	structure	NOUN
cana-5777	112	10	,	,	PUNCT
cana-5777	112	11	where	where	SCONJ
cana-5777	112	12	nodes	node	NOUN
cana-5777	112	13	represented	represent	VERB
cana-5777	112	14	individual	individual	ADJ
cana-5777	112	15	tumour	tumour	NOUN
cana-5777	112	16	samples	sample	NOUN
cana-5777	112	17	,	,	PUNCT
cana-5777	112	18	and	and	CCONJ
cana-5777	112	19	edges	edge	NOUN
cana-5777	112	20	were	be	AUX
cana-5777	112	21	created	create	VERB
cana-5777	112	22	based	base	VERB
cana-5777	112	23	on	on	ADP
cana-5777	112	24	feature	feature	NOUN
cana-5777	112	25	similarity	similarity	NOUN
cana-5777	112	26	.	.	PUNCT
cana-5777	113	1	this	this	DET
cana-5777	113	2	transformation	transformation	NOUN
cana-5777	113	3	is	be	AUX
cana-5777	113	4	visually	visually	ADV
cana-5777	113	5	represented	represent	VERB
cana-5777	113	6	in	in	ADP
cana-5777	113	7	fig.1	fig.1	ADJ
cana-5777	113	8	illustrating	illustrate	VERB
cana-5777	113	9	the	the	DET
cana-5777	113	10	connectivity	connectivity	NOUN
cana-5777	113	11	within	within	ADP
cana-5777	113	12	the	the	DET
cana-5777	113	13	dataset	dataset	NOUN
cana-5777	113	14	communications	communication	NOUN
cana-5777	113	15	on	on	ADP
cana-5777	113	16	applied	apply	VERB
cana-5777	113	17	nonlinear	nonlinear	ADJ
cana-5777	113	18	analysis	analysis	NOUN
cana-5777	113	19	issn	issn	NOUN
cana-5777	113	20	:	:	PUNCT
cana-5777	113	21	1074	1074	NUM
cana-5777	113	22	-	-	PUNCT
cana-5777	113	23	133x	133x	NUM
cana-5777	113	24	vol	vol	NOUN
cana-5777	113	25	31	31	NUM
cana-5777	113	26	no	no	NOUN
cana-5777	113	27	.	.	PUNCT
cana-5777	114	1	8s	8s	PROPN
cana-5777	114	2	(	(	PUNCT
cana-5777	114	3	2024	2024	NUM
cana-5777	114	4	)	)	PUNCT
cana-5777	114	5	https://internationalpubls.com	https://internationalpubls.com	X
cana-5777	114	6	1064	1064	NUM
cana-5777	114	7	figure	figure	NOUN
cana-5777	114	8	1	1	NUM
cana-5777	114	9	:	:	PUNCT
cana-5777	114	10	graph	graph	NOUN
cana-5777	114	11	representation	representation	NOUN
cana-5777	114	12	of	of	ADP
cana-5777	114	13	cancer	cancer	NOUN
cana-5777	114	14	data	datum	NOUN
cana-5777	114	15	the	the	DET
cana-5777	114	16	images	image	NOUN
cana-5777	114	17	show	show	VERB
cana-5777	114	18	how	how	SCONJ
cana-5777	114	19	different	different	ADJ
cana-5777	114	20	graph	graph	NOUN
cana-5777	114	21	neural	neural	ADJ
cana-5777	114	22	network	network	NOUN
cana-5777	114	23	(	(	PUNCT
cana-5777	114	24	gnn	gnn	PROPN
cana-5777	114	25	)	)	PUNCT
cana-5777	114	26	modelsthat	modelsthat	PRON
cana-5777	114	27	isgin	isgin	NOUN
cana-5777	114	28	,	,	PUNCT
cana-5777	114	29	gcn	gcn	ADJ
cana-5777	114	30	,	,	PUNCT
cana-5777	114	31	graphsage	graphsage	NOUN
cana-5777	114	32	,	,	PUNCT
cana-5777	114	33	gatand	gatand	NOUN
cana-5777	114	34	a	a	DET
cana-5777	114	35	general	general	ADJ
cana-5777	114	36	gnn	gnn	NOUN
cana-5777	114	37	.	.	PUNCT
cana-5777	115	1	the	the	DET
cana-5777	115	2	authors	author	NOUN
cana-5777	115	3	separate	separate	VERB
cana-5777	115	4	benign	benign	ADJ
cana-5777	115	5	(	(	PUNCT
cana-5777	115	6	blue	blue	ADJ
cana-5777	115	7	)	)	PUNCT
cana-5777	115	8	and	and	CCONJ
cana-5777	115	9	malignant	malignant	ADJ
cana-5777	115	10	(	(	PUNCT
cana-5777	115	11	red	red	ADJ
cana-5777	115	12	)	)	PUNCT
cana-5777	115	13	tumors	tumor	NOUN
cana-5777	115	14	based	base	VERB
cana-5777	115	15	on	on	ADP
cana-5777	115	16	learned	learn	VERB
cana-5777	115	17	embeddings	embedding	NOUN
cana-5777	115	18	.	.	PUNCT
cana-5777	116	1	each	each	DET
cana-5777	116	2	plot	plot	NOUN
cana-5777	116	3	represents	represent	VERB
cana-5777	116	4	the	the	DET
cana-5777	116	5	2d	2d	NUM
cana-5777	116	6	projection	projection	NOUN
cana-5777	116	7	of	of	ADP
cana-5777	116	8	node	node	ADJ
cana-5777	116	9	embeddings	embedding	NOUN
cana-5777	116	10	using	use	VERB
cana-5777	116	11	t	t	PROPN
cana-5777	116	12	-	-	PUNCT
cana-5777	116	13	sne	sne	NOUN
cana-5777	116	14	,	,	PUNCT
cana-5777	116	15	highlighting	highlight	VERB
cana-5777	116	16	how	how	SCONJ
cana-5777	116	17	well	well	ADV
cana-5777	116	18	each	each	DET
cana-5777	116	19	model	model	NOUN
cana-5777	116	20	distinguishes	distinguish	VERB
cana-5777	116	21	between	between	ADP
cana-5777	116	22	the	the	DET
cana-5777	116	23	two	two	NUM
cana-5777	116	24	tumor	tumor	NOUN
cana-5777	116	25	types	type	NOUN
cana-5777	116	26	.	.	PUNCT
cana-5777	117	1	the	the	DET
cana-5777	117	2	gcn	gcn	NOUN
cana-5777	117	3	and	and	CCONJ
cana-5777	117	4	general	general	ADJ
cana-5777	117	5	gnn	gnn	PROPN
cana-5777	117	6	models	model	NOUN
cana-5777	117	7	create	create	VERB
cana-5777	117	8	a	a	DET
cana-5777	117	9	clear	clear	ADJ
cana-5777	117	10	boundary	boundary	NOUN
cana-5777	117	11	between	between	ADP
cana-5777	117	12	benign	benign	ADJ
cana-5777	117	13	and	and	CCONJ
cana-5777	117	14	malignant	malignant	ADJ
cana-5777	117	15	clusters	cluster	NOUN
cana-5777	117	16	,	,	PUNCT
cana-5777	117	17	while	while	SCONJ
cana-5777	117	18	gat	gat	NOUN
cana-5777	117	19	,	,	PUNCT
cana-5777	117	20	graphsage	graphsage	NOUN
cana-5777	117	21	,	,	PUNCT
cana-5777	117	22	and	and	CCONJ
cana-5777	117	23	gin	gin	NOUN
cana-5777	117	24	also	also	ADV
cana-5777	117	25	show	show	VERB
cana-5777	117	26	good	good	ADJ
cana-5777	117	27	separation	separation	NOUN
cana-5777	117	28	but	but	CCONJ
cana-5777	117	29	with	with	ADP
cana-5777	117	30	some	some	DET
cana-5777	117	31	overlap	overlap	NOUN
cana-5777	117	32	.	.	PUNCT
cana-5777	118	1	the	the	DET
cana-5777	118	2	color	color	NOUN
cana-5777	118	3	bar	bar	NOUN
cana-5777	118	4	on	on	ADP
cana-5777	118	5	the	the	DET
cana-5777	118	6	right	right	NOUN
cana-5777	118	7	indicates	indicate	VERB
cana-5777	118	8	the	the	DET
cana-5777	118	9	classification	classification	NOUN
cana-5777	118	10	labels	label	NOUN
cana-5777	118	11	,	,	PUNCT
cana-5777	118	12	where	where	SCONJ
cana-5777	118	13	0	0	NUM
cana-5777	118	14	represents	represent	VERB
cana-5777	118	15	benign	benign	ADJ
cana-5777	118	16	tumors	tumor	NOUN
cana-5777	118	17	and	and	CCONJ
cana-5777	118	18	1	1	NUM
cana-5777	118	19	represents	represent	VERB
cana-5777	118	20	malignant	malignant	ADJ
cana-5777	118	21	tumors	tumor	NOUN
cana-5777	118	22	.	.	PUNCT
cana-5777	119	1	these	these	DET
cana-5777	119	2	visualizations	visualization	NOUN
cana-5777	119	3	help	help	VERB
cana-5777	119	4	in	in	ADP
cana-5777	119	5	understanding	understand	VERB
cana-5777	119	6	how	how	SCONJ
cana-5777	119	7	different	different	ADJ
cana-5777	119	8	gnn	gnn	PROPN
cana-5777	119	9	architectures	architecture	NOUN
cana-5777	119	10	learn	learn	VERB
cana-5777	119	11	feature	feature	NOUN
cana-5777	119	12	representations	representation	NOUN
cana-5777	119	13	for	for	ADP
cana-5777	119	14	medical	medical	ADJ
cana-5777	119	15	tumor	tumor	NOUN
cana-5777	119	16	classification	classification	NOUN
cana-5777	119	17	.	.	PUNCT
cana-5777	120	1	figure2	figure2	ADJ
cana-5777	120	2	:	:	PUNCT
cana-5777	120	3	gin	gin	NOUN
cana-5777	120	4	embeddings	embedding	NOUN
cana-5777	120	5	in	in	ADP
cana-5777	120	6	2d	2d	NUM
cana-5777	120	7	figure	figure	NOUN
cana-5777	120	8	3	3	NUM
cana-5777	120	9	:	:	PUNCT
cana-5777	120	10	gcn	gcn	NOUN
cana-5777	120	11	embeddings	embedding	NOUN
cana-5777	120	12	in	in	ADP
cana-5777	120	13	2d	2d	NOUN
cana-5777	120	14	communications	communication	NOUN
cana-5777	120	15	on	on	ADP
cana-5777	120	16	applied	apply	VERB
cana-5777	120	17	nonlinear	nonlinear	ADJ
cana-5777	120	18	analysis	analysis	NOUN
cana-5777	120	19	issn	issn	NOUN
cana-5777	120	20	:	:	PUNCT
cana-5777	120	21	1074	1074	NUM
cana-5777	120	22	-	-	PUNCT
cana-5777	120	23	133x	133x	NUM
cana-5777	120	24	vol	vol	NOUN
cana-5777	120	25	31	31	NUM
cana-5777	120	26	no	no	NOUN
cana-5777	120	27	.	.	PUNCT
cana-5777	121	1	8s	8s	PROPN
cana-5777	121	2	(	(	PUNCT
cana-5777	121	3	2024	2024	NUM
cana-5777	121	4	)	)	PUNCT
cana-5777	121	5	https://internationalpubls.com	https://internationalpubls.com	X
cana-5777	121	6	1065	1065	NUM
cana-5777	121	7	figure	figure	NOUN
cana-5777	121	8	4	4	NUM
cana-5777	121	9	:	:	PUNCT
cana-5777	121	10	graphsageembeddings	graphsageembedding	NOUN
cana-5777	121	11	in	in	ADP
cana-5777	121	12	2d	2d	NUM
cana-5777	121	13	figure	figure	NOUN
cana-5777	121	14	5	5	NUM
cana-5777	121	15	:	:	PUNCT
cana-5777	121	16	gatembeddings	gatembedding	NOUN
cana-5777	121	17	in	in	ADP
cana-5777	121	18	2d	2d	NUM
cana-5777	121	19	to	to	PART
cana-5777	121	20	improve	improve	VERB
cana-5777	121	21	classification	classification	NOUN
cana-5777	121	22	,	,	PUNCT
cana-5777	121	23	gnn	gnn	PROPN
cana-5777	121	24	model	model	NOUN
cana-5777	121	25	was	be	AUX
cana-5777	121	26	trained	train	VERB
cana-5777	121	27	on	on	ADP
cana-5777	121	28	the	the	DET
cana-5777	121	29	same	same	ADJ
cana-5777	121	30	dataset	dataset	NOUN
cana-5777	121	31	.	.	PUNCT
cana-5777	122	1	the	the	DET
cana-5777	122	2	effectiveness	effectiveness	NOUN
cana-5777	122	3	of	of	ADP
cana-5777	122	4	this	this	DET
cana-5777	122	5	approach	approach	NOUN
cana-5777	122	6	is	be	AUX
cana-5777	122	7	evident	evident	ADJ
cana-5777	122	8	in	in	ADP
cana-5777	122	9	fig.6	fig.6	PROPN
cana-5777	122	10	,	,	PUNCT
cana-5777	122	11	which	which	PRON
cana-5777	122	12	demonstrates	demonstrate	VERB
cana-5777	122	13	a	a	DET
cana-5777	122	14	clearer	clear	ADJ
cana-5777	122	15	separation	separation	NOUN
cana-5777	122	16	between	between	ADP
cana-5777	122	17	the	the	DET
cana-5777	122	18	two	two	NUM
cana-5777	122	19	classes	class	NOUN
cana-5777	122	20	compared	compare	VERB
cana-5777	122	21	to	to	ADP
cana-5777	122	22	the	the	DET
cana-5777	122	23	previous	previous	ADJ
cana-5777	122	24	embedding	embed	VERB
cana-5777	122	25	method	method	NOUN
cana-5777	122	26	.	.	PUNCT
cana-5777	123	1	figure	figure	VERB
cana-5777	123	2	6	6	NUM
cana-5777	123	3	:	:	PUNCT
cana-5777	123	4	gnn	gnn	NOUN
cana-5777	123	5	classification	classification	NOUN
cana-5777	123	6	boundaries	boundary	NOUN
cana-5777	123	7	the	the	DET
cana-5777	123	8	classification	classification	NOUN
cana-5777	123	9	performance	performance	NOUN
cana-5777	123	10	was	be	AUX
cana-5777	123	11	further	far	ADV
cana-5777	123	12	analysed	analyse	VERB
cana-5777	123	13	through	through	ADP
cana-5777	123	14	accuracy	accuracy	NOUN
cana-5777	123	15	,	,	PUNCT
cana-5777	123	16	precision	precision	NOUN
cana-5777	123	17	,	,	PUNCT
cana-5777	123	18	recall	recall	NOUN
cana-5777	123	19	,	,	PUNCT
cana-5777	123	20	f1score	f1score	NOUN
cana-5777	123	21	,	,	PUNCT
cana-5777	123	22	confusion	confusion	NOUN
cana-5777	123	23	matrix	matrix	NOUN
cana-5777	123	24	and	and	CCONJ
cana-5777	123	25	loss	loss	NOUN
cana-5777	123	26	trends	trend	NOUN
cana-5777	123	27	over	over	ADP
cana-5777	123	28	training	training	NOUN
cana-5777	123	29	epochs	epoch	NOUN
cana-5777	123	30	.	.	PUNCT
cana-5777	124	1	the	the	DET
cana-5777	124	2	model	model	NOUN
cana-5777	124	3	has	have	AUX
cana-5777	124	4	achieved	achieve	VERB
cana-5777	124	5	100	100	NUM
cana-5777	124	6	%	%	NOUN
cana-5777	124	7	accuracy	accuracy	NOUN
cana-5777	124	8	for	for	ADP
cana-5777	124	9	gcn	gcn	NOUN
cana-5777	124	10	,	,	PUNCT
cana-5777	124	11	gat	gat	NOUN
cana-5777	124	12	and	and	CCONJ
cana-5777	124	13	graphsage	graphsage	NOUN
cana-5777	124	14	,	,	PUNCT
cana-5777	124	15	and	and	CCONJ
cana-5777	124	16	99.42	99.42	NUM
cana-5777	124	17	%	%	NOUN
cana-5777	124	18	accuracy	accuracy	NOUN
cana-5777	124	19	for	for	ADP
cana-5777	124	20	gin	gin	NOUN
cana-5777	124	21	which	which	PRON
cana-5777	124	22	indicates	indicate	VERB
cana-5777	124	23	that	that	SCONJ
cana-5777	124	24	this	this	DET
cana-5777	124	25	graph	graph	NOUN
cana-5777	124	26	-	-	PUNCT
cana-5777	124	27	based	base	VERB
cana-5777	124	28	learning	learning	NOUN
cana-5777	124	29	approach	approach	NOUN
cana-5777	124	30	is	be	AUX
cana-5777	124	31	perfectly	perfectly	ADV
cana-5777	124	32	classifying	classify	VERB
cana-5777	124	33	the	the	DET
cana-5777	124	34	dataset	dataset	NOUN
cana-5777	124	35	as	as	SCONJ
cana-5777	124	36	shown	show	VERB
cana-5777	124	37	below	below	ADP
cana-5777	124	38	.	.	PUNCT
cana-5777	125	1	communications	communication	NOUN
cana-5777	125	2	on	on	ADP
cana-5777	125	3	applied	apply	VERB
cana-5777	125	4	nonlinear	nonlinear	ADJ
cana-5777	125	5	analysis	analysis	NOUN
cana-5777	125	6	issn	issn	NOUN
cana-5777	125	7	:	:	PUNCT
cana-5777	125	8	1074	1074	NUM
cana-5777	125	9	-	-	PUNCT
cana-5777	125	10	133x	133x	NUM
cana-5777	125	11	vol	vol	NOUN
cana-5777	125	12	31	31	NUM
cana-5777	125	13	no	no	NOUN
cana-5777	125	14	.	.	PUNCT
cana-5777	126	1	8s	8s	PROPN
cana-5777	126	2	(	(	PUNCT
cana-5777	126	3	2024	2024	NUM
cana-5777	126	4	)	)	PUNCT
cana-5777	126	5	https://internationalpubls.com	https://internationalpubls.com	X
cana-5777	126	6	1066	1066	NUM
cana-5777	126	7	the	the	DET
cana-5777	126	8	output	output	NOUN
cana-5777	126	9	shows	show	VERB
cana-5777	126	10	the	the	DET
cana-5777	126	11	performance	performance	NOUN
cana-5777	126	12	of	of	ADP
cana-5777	126	13	a	a	DET
cana-5777	126	14	graph	graph	NOUN
cana-5777	126	15	convolutional	convolutional	ADJ
cana-5777	126	16	network	network	NOUN
cana-5777	126	17	(	(	PUNCT
cana-5777	126	18	gcn	gcn	NOUN
cana-5777	126	19	)	)	PUNCT
cana-5777	126	20	in	in	ADP
cana-5777	126	21	classifying	classify	VERB
cana-5777	126	22	benign	benign	ADJ
cana-5777	126	23	and	and	CCONJ
cana-5777	126	24	malignant	malignant	ADJ
cana-5777	126	25	tumors	tumor	NOUN
cana-5777	126	26	.	.	PUNCT
cana-5777	127	1	the	the	DET
cana-5777	127	2	confusion	confusion	NOUN
cana-5777	127	3	matrix	matrix	NOUN
cana-5777	127	4	indicates	indicate	VERB
cana-5777	127	5	that	that	SCONJ
cana-5777	127	6	the	the	DET
cana-5777	127	7	model	model	NOUN
cana-5777	127	8	achieved	achieve	VERB
cana-5777	127	9	perfect	perfect	ADJ
cana-5777	127	10	classification	classification	NOUN
cana-5777	127	11	,	,	PUNCT
cana-5777	127	12	correctly	correctly	ADV
cana-5777	127	13	identifying	identify	VERB
cana-5777	127	14	all	all	DET
cana-5777	127	15	107	107	NUM
cana-5777	127	16	benign	benign	ADJ
cana-5777	127	17	and	and	CCONJ
cana-5777	127	18	64	64	NUM
cana-5777	127	19	malignant	malignant	ADJ
cana-5777	127	20	samples	sample	NOUN
cana-5777	127	21	without	without	ADP
cana-5777	127	22	any	any	DET
cana-5777	127	23	misclassifications	misclassification	NOUN
cana-5777	127	24	.	.	PUNCT
cana-5777	128	1	the	the	DET
cana-5777	128	2	performance	performance	NOUN
cana-5777	128	3	metrics	metric	NOUN
cana-5777	128	4	confirm	confirm	VERB
cana-5777	128	5	this	this	PRON
cana-5777	128	6	,	,	PUNCT
cana-5777	128	7	with	with	ADP
cana-5777	128	8	a	a	DET
cana-5777	128	9	test	test	NOUN
cana-5777	128	10	accuracy	accuracy	NOUN
cana-5777	128	11	of	of	ADP
cana-5777	128	12	100	100	NUM
cana-5777	128	13	%	%	NOUN
cana-5777	128	14	,	,	PUNCT
cana-5777	128	15	along	along	ADP
cana-5777	128	16	with	with	ADP
cana-5777	128	17	precision	precision	NOUN
cana-5777	128	18	,	,	PUNCT
cana-5777	128	19	recall	recall	NOUN
cana-5777	128	20	,	,	PUNCT
cana-5777	128	21	and	and	CCONJ
cana-5777	128	22	f1	f1	NOUN
cana-5777	128	23	-	-	PUNCT
cana-5777	128	24	score	score	NOUN
cana-5777	128	25	all	all	PRON
cana-5777	128	26	being	be	AUX
cana-5777	128	27	1.000	1.000	NUM
cana-5777	128	28	.	.	PUNCT
cana-5777	129	1	the	the	DET
cana-5777	129	2	loss	loss	NOUN
cana-5777	129	3	values	value	NOUN
cana-5777	129	4	remain	remain	VERB
cana-5777	129	5	low	low	ADJ
cana-5777	129	6	,	,	PUNCT
cana-5777	129	7	and	and	CCONJ
cana-5777	129	8	early	early	ADJ
cana-5777	129	9	stopping	stopping	NOUN
cana-5777	129	10	was	be	AUX
cana-5777	129	11	applied	apply	VERB
cana-5777	129	12	at	at	ADP
cana-5777	129	13	epoch	epoch	PROPN
cana-5777	129	14	45	45	NUM
cana-5777	129	15	,	,	PUNCT
cana-5777	129	16	ensuring	ensure	VERB
cana-5777	129	17	optimal	optimal	ADJ
cana-5777	129	18	training	training	NOUN
cana-5777	129	19	without	without	ADP
cana-5777	129	20	overfitting	overfitte	VERB
cana-5777	129	21	.	.	PUNCT
cana-5777	130	1	the	the	DET
cana-5777	130	2	runtime	runtime	NOUN
cana-5777	130	3	for	for	ADP
cana-5777	130	4	the	the	DET
cana-5777	130	5	model	model	NOUN
cana-5777	130	6	was	be	AUX
cana-5777	130	7	just	just	ADV
cana-5777	130	8	1	1	NUM
cana-5777	130	9	second	second	ADJ
cana-5777	130	10	,	,	PUNCT
cana-5777	130	11	highlighting	highlight	VERB
cana-5777	130	12	its	its	PRON
cana-5777	130	13	efficiency	efficiency	NOUN
cana-5777	130	14	.	.	PUNCT
cana-5777	131	1	this	this	DET
cana-5777	131	2	result	result	NOUN
cana-5777	131	3	demonstrates	demonstrate	VERB
cana-5777	131	4	that	that	SCONJ
cana-5777	131	5	the	the	DET
cana-5777	131	6	gcn	gcn	NOUN
cana-5777	131	7	successfully	successfully	ADV
cana-5777	131	8	learned	learn	VERB
cana-5777	131	9	meaningful	meaningful	ADJ
cana-5777	131	10	representations	representation	NOUN
cana-5777	131	11	for	for	ADP
cana-5777	131	12	distinguishing	distinguish	VERB
cana-5777	131	13	between	between	ADP
cana-5777	131	14	benign	benign	ADJ
cana-5777	131	15	and	and	CCONJ
cana-5777	131	16	malignant	malignant	ADJ
cana-5777	131	17	tumor	tumor	NOUN
cana-5777	131	18	samples	sample	NOUN
cana-5777	131	19	.	.	PUNCT
cana-5777	132	1	communications	communication	NOUN
cana-5777	132	2	on	on	ADP
cana-5777	132	3	applied	apply	VERB
cana-5777	132	4	nonlinear	nonlinear	ADJ
cana-5777	132	5	analysis	analysis	NOUN
cana-5777	132	6	issn	issn	NOUN
cana-5777	132	7	:	:	PUNCT
cana-5777	132	8	1074	1074	NUM
cana-5777	132	9	-	-	PUNCT
cana-5777	132	10	133x	133x	NUM
cana-5777	132	11	vol	vol	NOUN
cana-5777	132	12	31	31	NUM
cana-5777	132	13	no	no	NOUN
cana-5777	132	14	.	.	PUNCT
cana-5777	133	1	8s	8s	PROPN
cana-5777	133	2	(	(	PUNCT
cana-5777	133	3	2024	2024	NUM
cana-5777	133	4	)	)	PUNCT
cana-5777	133	5	https://internationalpubls.com	https://internationalpubls.com	X
cana-5777	133	6	1067	1067	NUM
cana-5777	133	7	the	the	DET
cana-5777	133	8	confusion	confusion	NOUN
cana-5777	133	9	matrix	matrix	NOUN
cana-5777	133	10	and	and	CCONJ
cana-5777	133	11	performance	performance	NOUN
cana-5777	133	12	metrics	metric	NOUN
cana-5777	133	13	illustrate	illustrate	VERB
cana-5777	133	14	the	the	DET
cana-5777	133	15	effectiveness	effectiveness	NOUN
cana-5777	133	16	of	of	ADP
cana-5777	133	17	the	the	DET
cana-5777	133	18	graph	graph	NOUN
cana-5777	133	19	isomorphism	isomorphism	NOUN
cana-5777	133	20	network	network	NOUN
cana-5777	133	21	(	(	PUNCT
cana-5777	133	22	gin	gin	NOUN
cana-5777	133	23	)	)	PUNCT
cana-5777	133	24	in	in	ADP
cana-5777	133	25	classifying	classify	VERB
cana-5777	133	26	benign	benign	ADJ
cana-5777	133	27	and	and	CCONJ
cana-5777	133	28	malignant	malignant	ADJ
cana-5777	133	29	tumors	tumor	NOUN
cana-5777	133	30	.	.	PUNCT
cana-5777	134	1	the	the	DET
cana-5777	134	2	model	model	NOUN
cana-5777	134	3	achieved	achieve	VERB
cana-5777	134	4	a	a	DET
cana-5777	134	5	high	high	ADJ
cana-5777	134	6	-	-	PUNCT
cana-5777	134	7	test	test	NOUN
cana-5777	134	8	accuracy	accuracy	NOUN
cana-5777	134	9	of	of	ADP
cana-5777	134	10	99.42	99.42	NUM
cana-5777	134	11	%	%	NOUN
cana-5777	134	12	,	,	PUNCT
cana-5777	134	13	with	with	ADP
cana-5777	134	14	precision	precision	NOUN
cana-5777	134	15	,	,	PUNCT
cana-5777	134	16	recall	recall	NOUN
cana-5777	134	17	,	,	PUNCT
cana-5777	134	18	and	and	CCONJ
cana-5777	134	19	f1	f1	NOUN
cana-5777	134	20	-	-	PUNCT
cana-5777	134	21	score	score	NOUN
cana-5777	134	22	all	all	PRON
cana-5777	134	23	reaching	reach	VERB
cana-5777	134	24	approximately	approximately	ADV
cana-5777	134	25	0.994	0.994	NUM
cana-5777	134	26	.	.	PUNCT
cana-5777	135	1	out	out	ADP
cana-5777	135	2	of	of	ADP
cana-5777	135	3	107	107	NUM
cana-5777	135	4	benign	benign	ADJ
cana-5777	135	5	samples	sample	NOUN
cana-5777	135	6	,	,	PUNCT
cana-5777	135	7	the	the	DET
cana-5777	135	8	model	model	NOUN
cana-5777	135	9	classified	classify	VERB
cana-5777	135	10	all	all	ADV
cana-5777	135	11	correctly	correctly	ADV
cana-5777	135	12	,	,	PUNCT
cana-5777	135	13	while	while	SCONJ
cana-5777	135	14	for	for	ADP
cana-5777	135	15	64	64	NUM
cana-5777	135	16	malignant	malignant	ADJ
cana-5777	135	17	cases	case	NOUN
cana-5777	135	18	,	,	PUNCT
cana-5777	135	19	it	it	PRON
cana-5777	135	20	correctly	correctly	ADV
cana-5777	135	21	identified	identify	VERB
cana-5777	135	22	63	63	NUM
cana-5777	135	23	but	but	CCONJ
cana-5777	135	24	misclassified	misclassifie	VERB
cana-5777	135	25	1	1	NUM
cana-5777	135	26	as	as	ADP
cana-5777	135	27	benign	benign	ADJ
cana-5777	135	28	.	.	PUNCT
cana-5777	136	1	the	the	DET
cana-5777	136	2	training	training	NOUN
cana-5777	136	3	process	process	NOUN
cana-5777	136	4	stopped	stop	VERB
cana-5777	136	5	early	early	ADV
cana-5777	136	6	at	at	ADP
cana-5777	136	7	epoch	epoch	PROPN
cana-5777	136	8	59	59	NUM
cana-5777	136	9	,	,	PUNCT
cana-5777	136	10	indicating	indicate	VERB
cana-5777	136	11	stability	stability	NOUN
cana-5777	136	12	in	in	ADP
cana-5777	136	13	performance	performance	NOUN
cana-5777	136	14	.	.	PUNCT
cana-5777	137	1	although	although	SCONJ
cana-5777	137	2	slightly	slightly	ADV
cana-5777	137	3	less	less	ADV
cana-5777	137	4	accurate	accurate	ADJ
cana-5777	137	5	than	than	ADP
cana-5777	137	6	gcn	gcn	ADJ
cana-5777	137	7	,	,	PUNCT
cana-5777	137	8	gin	gin	NOUN
cana-5777	137	9	still	still	ADV
cana-5777	137	10	demonstrates	demonstrate	VERB
cana-5777	137	11	strong	strong	ADJ
cana-5777	137	12	classification	classification	NOUN
cana-5777	137	13	capability	capability	NOUN
cana-5777	137	14	,	,	PUNCT
cana-5777	137	15	efficiently	efficiently	ADV
cana-5777	137	16	distinguishing	distinguish	VERB
cana-5777	137	17	between	between	ADP
cana-5777	137	18	benign	benign	ADJ
cana-5777	137	19	and	and	CCONJ
cana-5777	137	20	malignant	malignant	ADJ
cana-5777	137	21	tumors	tumor	NOUN
cana-5777	137	22	with	with	ADP
cana-5777	137	23	minimal	minimal	ADJ
cana-5777	137	24	errors	error	NOUN
cana-5777	137	25	.	.	PUNCT
cana-5777	138	1	communications	communication	NOUN
cana-5777	138	2	on	on	ADP
cana-5777	138	3	applied	apply	VERB
cana-5777	138	4	nonlinear	nonlinear	ADJ
cana-5777	138	5	analysis	analysis	NOUN
cana-5777	138	6	issn	issn	NOUN
cana-5777	138	7	:	:	PUNCT
cana-5777	138	8	1074	1074	NUM
cana-5777	138	9	-	-	PUNCT
cana-5777	138	10	133x	133x	NUM
cana-5777	138	11	vol	vol	NOUN
cana-5777	138	12	31	31	NUM
cana-5777	138	13	no	no	NOUN
cana-5777	138	14	.	.	PUNCT
cana-5777	139	1	8s	8s	PROPN
cana-5777	139	2	(	(	PUNCT
cana-5777	139	3	2024	2024	NUM
cana-5777	139	4	)	)	PUNCT
cana-5777	139	5	https://internationalpubls.com	https://internationalpubls.com	X
cana-5777	139	6	1068	1068	NUM
cana-5777	139	7	the	the	DET
cana-5777	139	8	graphsage	graphsage	NOUN
cana-5777	139	9	model	model	NOUN
cana-5777	139	10	demonstrated	demonstrate	VERB
cana-5777	139	11	exceptional	exceptional	ADJ
cana-5777	139	12	performance	performance	NOUN
cana-5777	139	13	in	in	ADP
cana-5777	139	14	classifying	classify	VERB
cana-5777	139	15	medical	medical	ADJ
cana-5777	139	16	graph	graph	NOUN
cana-5777	139	17	data	datum	NOUN
cana-5777	139	18	,	,	PUNCT
cana-5777	139	19	achieving	achieve	VERB
cana-5777	139	20	100	100	NUM
cana-5777	139	21	%	%	NOUN
cana-5777	139	22	accuracy	accuracy	NOUN
cana-5777	139	23	,	,	PUNCT
cana-5777	139	24	precision	precision	NOUN
cana-5777	139	25	,	,	PUNCT
cana-5777	139	26	recall	recall	NOUN
cana-5777	139	27	,	,	PUNCT
cana-5777	139	28	and	and	CCONJ
cana-5777	139	29	f1	f1	NOUN
cana-5777	139	30	-	-	PUNCT
cana-5777	139	31	score	score	NOUN
cana-5777	139	32	after	after	ADP
cana-5777	139	33	only	only	ADV
cana-5777	139	34	22	22	NUM
cana-5777	139	35	epochs	epoch	NOUN
cana-5777	139	36	.	.	PUNCT
cana-5777	140	1	as	as	SCONJ
cana-5777	140	2	shown	show	VERB
cana-5777	140	3	in	in	ADP
cana-5777	140	4	the	the	DET
cana-5777	140	5	confusion	confusion	NOUN
cana-5777	140	6	matrix	matrix	NOUN
cana-5777	140	7	,	,	PUNCT
cana-5777	140	8	the	the	DET
cana-5777	140	9	model	model	NOUN
cana-5777	140	10	correctly	correctly	ADV
cana-5777	140	11	classified	classify	VERB
cana-5777	140	12	all	all	ADV
cana-5777	140	13	benign	benign	ADJ
cana-5777	140	14	(	(	PUNCT
cana-5777	140	15	72	72	NUM
cana-5777	140	16	instances	instance	NOUN
cana-5777	140	17	)	)	PUNCT
cana-5777	140	18	and	and	CCONJ
cana-5777	140	19	malignant	malignant	ADJ
cana-5777	140	20	(	(	PUNCT
cana-5777	140	21	42	42	NUM
cana-5777	140	22	instances	instance	NOUN
cana-5777	140	23	)	)	PUNCT
cana-5777	140	24	tumors	tumor	NOUN
cana-5777	140	25	without	without	ADP
cana-5777	140	26	any	any	DET
cana-5777	140	27	false	false	ADJ
cana-5777	140	28	positives	positive	NOUN
cana-5777	140	29	or	or	CCONJ
cana-5777	140	30	false	false	ADJ
cana-5777	140	31	negatives	negative	NOUN
cana-5777	140	32	.	.	PUNCT
cana-5777	141	1	the	the	DET
cana-5777	141	2	early	early	ADJ
cana-5777	141	3	stopping	stopping	NOUN
cana-5777	141	4	mechanism	mechanism	NOUN
cana-5777	141	5	ensured	ensure	VERB
cana-5777	141	6	optimal	optimal	ADJ
cana-5777	141	7	training	training	NOUN
cana-5777	141	8	efficiency	efficiency	NOUN
cana-5777	141	9	,	,	PUNCT
cana-5777	141	10	preventing	prevent	VERB
cana-5777	141	11	overfitting	overfitting	NOUN
cana-5777	141	12	while	while	SCONJ
cana-5777	141	13	maintaining	maintain	VERB
cana-5777	141	14	perfect	perfect	ADJ
cana-5777	141	15	classification	classification	NOUN
cana-5777	141	16	within	within	ADP
cana-5777	141	17	zero	zero	NUM
cana-5777	141	18	seconds	second	NOUN
cana-5777	141	19	runtime	runtime	NOUN
cana-5777	141	20	.	.	PUNCT
cana-5777	142	1	the	the	DET
cana-5777	142	2	graph	graph	NOUN
cana-5777	142	3	attention	attention	NOUN
cana-5777	142	4	network	network	NOUN
cana-5777	142	5	(	(	PUNCT
cana-5777	142	6	gat	gat	NOUN
cana-5777	142	7	)	)	PUNCT
cana-5777	142	8	model	model	NOUN
cana-5777	142	9	exhibited	exhibit	VERB
cana-5777	142	10	outstanding	outstanding	ADJ
cana-5777	142	11	performance	performance	NOUN
cana-5777	142	12	in	in	ADP
cana-5777	142	13	classifying	classify	VERB
cana-5777	142	14	medical	medical	ADJ
cana-5777	142	15	graph	graph	NOUN
cana-5777	142	16	data	datum	NOUN
cana-5777	142	17	,	,	PUNCT
cana-5777	142	18	achieving	achieve	VERB
cana-5777	142	19	100	100	NUM
cana-5777	142	20	%	%	NOUN
cana-5777	142	21	accuracy	accuracy	NOUN
cana-5777	142	22	,	,	PUNCT
cana-5777	142	23	precision	precision	NOUN
cana-5777	142	24	,	,	PUNCT
cana-5777	142	25	recall	recall	NOUN
cana-5777	142	26	,	,	PUNCT
cana-5777	142	27	and	and	CCONJ
cana-5777	142	28	f1	f1	NOUN
cana-5777	142	29	-	-	PUNCT
cana-5777	142	30	score	score	NOUN
cana-5777	142	31	by	by	ADP
cana-5777	142	32	epoch	epoch	PROPN
cana-5777	142	33	22	22	NUM
cana-5777	142	34	.	.	PUNCT
cana-5777	143	1	initially	initially	ADV
cana-5777	143	2	,	,	PUNCT
cana-5777	143	3	at	at	ADP
cana-5777	143	4	epoch	epoch	NOUN
cana-5777	143	5	0	0	NUM
cana-5777	143	6	,	,	PUNCT
cana-5777	143	7	the	the	DET
cana-5777	143	8	model	model	NOUN
cana-5777	143	9	had	have	VERB
cana-5777	143	10	a	a	DET
cana-5777	143	11	loss	loss	NOUN
cana-5777	143	12	of	of	ADP
cana-5777	143	13	0.6213	0.6213	NUM
cana-5777	143	14	and	and	CCONJ
cana-5777	143	15	an	an	DET
cana-5777	143	16	accuracy	accuracy	NOUN
cana-5777	143	17	of	of	ADP
cana-5777	143	18	99.12	99.12	NUM
cana-5777	143	19	%	%	NOUN
cana-5777	143	20	,	,	PUNCT
cana-5777	143	21	which	which	PRON
cana-5777	143	22	improved	improve	VERB
cana-5777	143	23	rapidly	rapidly	ADV
cana-5777	143	24	as	as	SCONJ
cana-5777	143	25	training	training	NOUN
cana-5777	143	26	progressed	progress	VERB
cana-5777	143	27	.	.	PUNCT
cana-5777	144	1	by	by	ADP
cana-5777	144	2	epoch	epoch	PROPN
cana-5777	144	3	10	10	NUM
cana-5777	144	4	,	,	PUNCT
cana-5777	144	5	the	the	DET
cana-5777	144	6	model	model	NOUN
cana-5777	144	7	reached	reach	VERB
cana-5777	144	8	perfect	perfect	ADJ
cana-5777	144	9	classification	classification	NOUN
cana-5777	144	10	(	(	PUNCT
cana-5777	144	11	100	100	NUM
cana-5777	144	12	%	%	NOUN
cana-5777	144	13	accuracy	accuracy	NOUN
cana-5777	144	14	)	)	PUNCT
cana-5777	144	15	,	,	PUNCT
cana-5777	144	16	effectively	effectively	ADV
cana-5777	144	17	distinguishing	distinguish	VERB
cana-5777	144	18	between	between	ADP
cana-5777	144	19	benign	benign	ADJ
cana-5777	144	20	(	(	PUNCT
cana-5777	144	21	72	72	NUM
cana-5777	144	22	instances	instance	NOUN
cana-5777	144	23	)	)	PUNCT
cana-5777	144	24	and	and	CCONJ
cana-5777	144	25	malignant	malignant	ADJ
cana-5777	144	26	(	(	PUNCT
cana-5777	144	27	42	42	NUM
cana-5777	144	28	instances	instance	NOUN
cana-5777	144	29	)	)	PUNCT
cana-5777	144	30	tumors	tumor	NOUN
cana-5777	144	31	without	without	ADP
cana-5777	144	32	any	any	DET
cana-5777	144	33	false	false	ADJ
cana-5777	144	34	positives	positive	NOUN
cana-5777	144	35	or	or	CCONJ
cana-5777	144	36	false	false	ADJ
cana-5777	144	37	negatives	negative	NOUN
cana-5777	144	38	.	.	PUNCT
cana-5777	145	1	the	the	DET
cana-5777	145	2	early	early	ADJ
cana-5777	145	3	stopping	stopping	NOUN
cana-5777	145	4	mechanism	mechanism	NOUN
cana-5777	145	5	halted	halt	VERB
cana-5777	145	6	training	training	NOUN
cana-5777	145	7	at	at	ADP
cana-5777	145	8	epoch	epoch	NOUN
cana-5777	145	9	22	22	NUM
cana-5777	145	10	,	,	PUNCT
cana-5777	145	11	preventing	prevent	VERB
cana-5777	145	12	overfitting	overfitting	NOUN
cana-5777	145	13	while	while	SCONJ
cana-5777	145	14	ensuring	ensure	VERB
cana-5777	145	15	maximum	maximum	ADJ
cana-5777	145	16	performance	performance	NOUN
cana-5777	145	17	.	.	PUNCT
cana-5777	146	1	the	the	DET
cana-5777	146	2	confusion	confusion	NOUN
cana-5777	146	3	matrix	matrix	NOUN
cana-5777	146	4	confirms	confirm	VERB
cana-5777	146	5	this	this	PRON
cana-5777	146	6	,	,	PUNCT
cana-5777	146	7	showing	show	VERB
cana-5777	146	8	that	that	SCONJ
cana-5777	146	9	all	all	DET
cana-5777	146	10	instances	instance	NOUN
cana-5777	146	11	were	be	AUX
cana-5777	146	12	correctly	correctly	ADV
cana-5777	146	13	classified	classified	ADJ
cana-5777	146	14	.	.	PUNCT
cana-5777	147	1	the	the	DET
cana-5777	147	2	gat	gat	PROPN
cana-5777	147	3	model	model	NOUN
cana-5777	147	4	's	's	PART
cana-5777	147	5	ability	ability	NOUN
cana-5777	147	6	to	to	PART
cana-5777	147	7	leverage	leverage	VERB
cana-5777	147	8	attention	attention	NOUN
cana-5777	147	9	-	-	PUNCT
cana-5777	147	10	based	base	VERB
cana-5777	147	11	message	message	NOUN
cana-5777	147	12	passing	pass	VERB
cana-5777	147	13	further	far	ADV
cana-5777	147	14	supports	support	VERB
cana-5777	147	15	its	its	PRON
cana-5777	147	16	effectiveness	effectiveness	NOUN
cana-5777	147	17	in	in	ADP
cana-5777	147	18	handling	handle	VERB
cana-5777	147	19	complex	complex	ADJ
cana-5777	147	20	graph	graph	NOUN
cana-5777	147	21	-	-	PUNCT
cana-5777	147	22	structured	structure	VERB
cana-5777	147	23	medical	medical	ADJ
cana-5777	147	24	data	datum	NOUN
cana-5777	147	25	,	,	PUNCT
cana-5777	147	26	making	make	VERB
cana-5777	147	27	it	it	PRON
cana-5777	147	28	a	a	DET
cana-5777	147	29	powerful	powerful	ADJ
cana-5777	147	30	tool	tool	NOUN
cana-5777	147	31	for	for	ADP
cana-5777	147	32	cancer	cancer	NOUN
cana-5777	147	33	detection	detection	NOUN
cana-5777	147	34	and	and	CCONJ
cana-5777	147	35	automated	automate	VERB
cana-5777	147	36	medical	medical	ADJ
cana-5777	147	37	drug	drug	NOUN
cana-5777	147	38	analysis	analysis	NOUN
cana-5777	147	39	.	.	PUNCT
cana-5777	148	1	the	the	DET
cana-5777	148	2	authors	author	NOUN
cana-5777	148	3	tested	test	VERB
cana-5777	148	4	the	the	DET
cana-5777	148	5	robustness	robustness	NOUN
cana-5777	148	6	by	by	ADP
cana-5777	148	7	making	make	VERB
cana-5777	148	8	small	small	ADJ
cana-5777	148	9	changes	change	NOUN
cana-5777	148	10	to	to	ADP
cana-5777	148	11	the	the	DET
cana-5777	148	12	graph	graph	NOUN
cana-5777	148	13	structure	structure	NOUN
cana-5777	148	14	and	and	CCONJ
cana-5777	148	15	node	node	NOUN
cana-5777	148	16	features	feature	NOUN
cana-5777	148	17	and	and	CCONJ
cana-5777	148	18	measured	measure	VERB
cana-5777	148	19	accuracy	accuracy	NOUN
cana-5777	148	20	,	,	PUNCT
cana-5777	148	21	precision	precision	NOUN
cana-5777	148	22	,	,	PUNCT
cana-5777	148	23	recall	recall	NOUN
cana-5777	148	24	,	,	PUNCT
cana-5777	148	25	and	and	CCONJ
cana-5777	148	26	f1	f1	NOUN
cana-5777	148	27	-	-	PUNCT
cana-5777	148	28	score	score	NOUN
cana-5777	148	29	before	before	ADV
cana-5777	148	30	and	and	CCONJ
cana-5777	148	31	after	after	ADP
cana-5777	148	32	these	these	DET
cana-5777	148	33	changes	change	NOUN
cana-5777	148	34	.	.	PUNCT
cana-5777	149	1	after	after	ADP
cana-5777	149	2	modifications	modification	NOUN
cana-5777	149	3	also	also	ADV
cana-5777	149	4	all	all	DET
cana-5777	149	5	four	four	NUM
cana-5777	149	6	models	model	NOUN
cana-5777	149	7	(	(	PUNCT
cana-5777	149	8	gcn	gcn	NOUN
cana-5777	149	9	,	,	PUNCT
cana-5777	149	10	gin	gin	NOUN
cana-5777	149	11	,	,	PUNCT
cana-5777	149	12	gat	gat	NOUN
cana-5777	149	13	,	,	PUNCT
cana-5777	149	14	and	and	CCONJ
cana-5777	149	15	graphsage	graphsage	NOUN
cana-5777	149	16	)	)	PUNCT
cana-5777	149	17	performed	perform	VERB
cana-5777	149	18	well	well	ADV
cana-5777	149	19	and	and	CCONJ
cana-5777	149	20	maintained	maintain	VERB
cana-5777	149	21	the	the	DET
cana-5777	149	22	same	same	ADJ
cana-5777	149	23	accuracy	accuracy	NOUN
cana-5777	149	24	.	.	PUNCT
cana-5777	150	1	the	the	DET
cana-5777	150	2	findings	finding	NOUN
cana-5777	150	3	confirmed	confirm	VERB
cana-5777	150	4	that	that	SCONJ
cana-5777	150	5	gcn	gcn	NOUN
cana-5777	150	6	,	,	PUNCT
cana-5777	150	7	gin	gin	NOUN
cana-5777	150	8	,	,	PUNCT
cana-5777	150	9	gat	gat	NOUN
cana-5777	150	10	,	,	PUNCT
cana-5777	150	11	and	and	CCONJ
cana-5777	150	12	graphsage	graphsage	NOUN
cana-5777	150	13	can	can	AUX
cana-5777	150	14	extract	extract	VERB
cana-5777	150	15	useful	useful	ADJ
cana-5777	150	16	graph	graph	NOUN
cana-5777	150	17	-	-	PUNCT
cana-5777	150	18	based	base	VERB
cana-5777	150	19	features	feature	NOUN
cana-5777	150	20	while	while	SCONJ
cana-5777	150	21	staying	stay	VERB
cana-5777	150	22	reliable	reliable	ADJ
cana-5777	150	23	even	even	ADV
cana-5777	150	24	when	when	SCONJ
cana-5777	150	25	the	the	DET
cana-5777	150	26	data	datum	NOUN
cana-5777	150	27	changes	change	VERB
cana-5777	150	28	slightly	slightly	ADV
cana-5777	150	29	.	.	PUNCT
cana-5777	151	1	the	the	DET
cana-5777	151	2	performance	performance	NOUN
cana-5777	151	3	comparison	comparison	NOUN
cana-5777	151	4	of	of	ADP
cana-5777	151	5	traditional	traditional	ADJ
cana-5777	151	6	ml	ml	NOUN
cana-5777	151	7	classifiers	classifier	NOUN
cana-5777	151	8	and	and	CCONJ
cana-5777	151	9	graph	graph	NOUN
cana-5777	151	10	-	-	PUNCT
cana-5777	151	11	based	base	VERB
cana-5777	151	12	models	model	NOUN
cana-5777	151	13	is	be	AUX
cana-5777	151	14	presented	present	VERB
cana-5777	151	15	in	in	ADP
cana-5777	151	16	the	the	DET
cana-5777	151	17	tables	table	NOUN
cana-5777	151	18	below	below	ADV
cana-5777	151	19	:	:	PUNCT
cana-5777	151	20	communications	communication	NOUN
cana-5777	151	21	on	on	ADP
cana-5777	151	22	applied	apply	VERB
cana-5777	151	23	nonlinear	nonlinear	ADJ
cana-5777	151	24	analysis	analysis	NOUN
cana-5777	151	25	issn	issn	NOUN
cana-5777	151	26	:	:	PUNCT
cana-5777	151	27	1074	1074	NUM
cana-5777	151	28	-	-	PUNCT
cana-5777	151	29	133x	133x	NUM
cana-5777	151	30	vol	vol	NOUN
cana-5777	151	31	31	31	NUM
cana-5777	151	32	no	no	NOUN
cana-5777	151	33	.	.	PUNCT
cana-5777	152	1	8s	8s	PROPN
cana-5777	152	2	(	(	PUNCT
cana-5777	152	3	2024	2024	NUM
cana-5777	152	4	)	)	PUNCT
cana-5777	152	5	https://internationalpubls.com	https://internationalpubls.com	X
cana-5777	152	6	1069	1069	NUM
cana-5777	152	7	the	the	DET
cana-5777	152	8	test	test	NOUN
cana-5777	152	9	accuracy	accuracy	NOUN
cana-5777	152	10	of	of	ADP
cana-5777	152	11	various	various	ADJ
cana-5777	152	12	machine	machine	NOUN
cana-5777	152	13	learning	learning	NOUN
cana-5777	152	14	models	model	NOUN
cana-5777	152	15	indicates	indicate	VERB
cana-5777	152	16	that	that	SCONJ
cana-5777	152	17	collective	collective	ADJ
cana-5777	152	18	methods	method	NOUN
cana-5777	152	19	such	such	ADJ
cana-5777	152	20	as	as	ADP
cana-5777	152	21	adaboost	adaboost	ADV
cana-5777	152	22	and	and	CCONJ
cana-5777	152	23	lightgbm	lightgbm	ADJ
cana-5777	152	24	achieved	achieve	VERB
cana-5777	152	25	98.8	98.8	NUM
cana-5777	152	26	%	%	NOUN
cana-5777	152	27	accuracy	accuracy	NOUN
cana-5777	152	28	,	,	PUNCT
cana-5777	152	29	while	while	SCONJ
cana-5777	152	30	xgboost	xgboost	ADV
cana-5777	152	31	and	and	CCONJ
cana-5777	152	32	random	random	ADJ
cana-5777	152	33	forest	forest	NOUN
cana-5777	152	34	performed	perform	VERB
cana-5777	152	35	slightly	slightly	ADV
cana-5777	152	36	lower	low	ADJ
cana-5777	152	37	.	.	PUNCT
cana-5777	153	1	however	however	ADV
cana-5777	153	2	,	,	PUNCT
cana-5777	153	3	when	when	SCONJ
cana-5777	153	4	transitioning	transition	VERB
cana-5777	153	5	to	to	ADP
cana-5777	153	6	graph	graph	NOUN
cana-5777	153	7	-	-	PUNCT
cana-5777	153	8	based	base	VERB
cana-5777	153	9	learning	learning	NOUN
cana-5777	153	10	,	,	PUNCT
cana-5777	153	11	the	the	DET
cana-5777	153	12	classification	classification	NOUN
cana-5777	153	13	accuracy	accuracy	NOUN
cana-5777	153	14	further	far	ADV
cana-5777	153	15	improved	improve	VERB
cana-5777	153	16	,	,	PUNCT
cana-5777	153	17	with	with	ADP
cana-5777	153	18	gin	gin	NOUN
cana-5777	153	19	reaching	reach	VERB
cana-5777	153	20	99.42	99.42	NUM
cana-5777	153	21	%	%	NOUN
cana-5777	153	22	and	and	CCONJ
cana-5777	153	23	gcn	gcn	NOUN
cana-5777	153	24	,	,	PUNCT
cana-5777	153	25	gat	gat	NOUN
cana-5777	153	26	and	and	CCONJ
cana-5777	153	27	graphsage	graphsage	NOUN
cana-5777	153	28	are	be	AUX
cana-5777	153	29	achieving	achieve	VERB
cana-5777	153	30	100	100	NUM
cana-5777	153	31	%	%	NOUN
cana-5777	153	32	accuracy	accuracy	NOUN
cana-5777	153	33	,	,	PUNCT
cana-5777	153	34	demonstrating	demonstrate	VERB
cana-5777	153	35	the	the	DET
cana-5777	153	36	power	power	NOUN
cana-5777	153	37	of	of	ADP
cana-5777	153	38	relational	relational	ADJ
cana-5777	153	39	learning	learning	NOUN
cana-5777	153	40	.	.	PUNCT
cana-5777	154	1	communications	communication	NOUN
cana-5777	154	2	on	on	ADP
cana-5777	154	3	applied	apply	VERB
cana-5777	154	4	nonlinear	nonlinear	ADJ
cana-5777	154	5	analysis	analysis	NOUN
cana-5777	154	6	issn	issn	NOUN
cana-5777	154	7	:	:	PUNCT
cana-5777	154	8	1074	1074	NUM
cana-5777	154	9	-	-	PUNCT
cana-5777	154	10	133x	133x	NUM
cana-5777	154	11	vol	vol	NOUN
cana-5777	154	12	31	31	NUM
cana-5777	154	13	no	no	NOUN
cana-5777	154	14	.	.	PUNCT
cana-5777	155	1	8s	8s	PROPN
cana-5777	155	2	(	(	PUNCT
cana-5777	155	3	2024	2024	NUM
cana-5777	155	4	)	)	PUNCT
cana-5777	155	5	https://internationalpubls.com	https://internationalpubls.com	X
cana-5777	155	6	1070	1070	NUM
cana-5777	155	7	these	these	DET
cana-5777	155	8	differences	difference	NOUN
cana-5777	155	9	are	be	AUX
cana-5777	155	10	visually	visually	ADV
cana-5777	155	11	represented	represent	VERB
cana-5777	155	12	in	in	ADP
cana-5777	155	13	fig.7where	fig.7where	ADP
cana-5777	155	14	this	this	DET
cana-5777	155	15	bar	bar	NOUN
cana-5777	155	16	chart	chart	NOUN
cana-5777	155	17	compares	compare	VERB
cana-5777	155	18	the	the	DET
cana-5777	155	19	accuracy	accuracy	NOUN
cana-5777	155	20	of	of	ADP
cana-5777	155	21	traditional	traditional	ADJ
cana-5777	155	22	machine	machine	NOUN
cana-5777	155	23	learning	learning	NOUN
cana-5777	155	24	models	model	NOUN
cana-5777	155	25	(	(	PUNCT
cana-5777	155	26	in	in	ADP
cana-5777	155	27	blue	blue	ADJ
cana-5777	155	28	)	)	PUNCT
cana-5777	155	29	and	and	CCONJ
cana-5777	155	30	graph	graph	NOUN
cana-5777	155	31	-	-	PUNCT
cana-5777	155	32	based	base	VERB
cana-5777	155	33	learning	learning	NOUN
cana-5777	155	34	models	model	NOUN
cana-5777	155	35	(	(	PUNCT
cana-5777	155	36	in	in	ADP
cana-5777	155	37	red	red	NOUN
cana-5777	155	38	)	)	PUNCT
cana-5777	155	39	which	which	PRON
cana-5777	155	40	clearly	clearly	ADV
cana-5777	155	41	illustrates	illustrate	VERB
cana-5777	155	42	the	the	DET
cana-5777	155	43	superior	superior	ADJ
cana-5777	155	44	performance	performance	NOUN
cana-5777	155	45	of	of	ADP
cana-5777	155	46	graph	graph	NOUN
cana-5777	155	47	-	-	PUNCT
cana-5777	155	48	based	base	VERB
cana-5777	155	49	models	model	NOUN
cana-5777	155	50	.	.	PUNCT
cana-5777	156	1	figure	figure	VERB
cana-5777	156	2	7	7	NUM
cana-5777	156	3	:	:	PUNCT
cana-5777	156	4	accuracy	accuracy	NOUN
cana-5777	156	5	comparison	comparison	NOUN
cana-5777	156	6	:	:	PUNCT
cana-5777	156	7	traditional	traditional	ADJ
cana-5777	156	8	ml	ml	X
cana-5777	156	9	versus	versus	ADP
cana-5777	156	10	graph	graph	NOUN
cana-5777	156	11	-	-	PUNCT
cana-5777	156	12	based	base	VERB
cana-5777	156	13	models	model	NOUN
cana-5777	156	14	conclusion	conclusion	NOUN
cana-5777	156	15	:	:	PUNCT
cana-5777	156	16	this	this	DET
cana-5777	156	17	study	study	NOUN
cana-5777	156	18	shows	show	VERB
cana-5777	156	19	that	that	SCONJ
cana-5777	156	20	graph	graph	NOUN
cana-5777	156	21	-	-	PUNCT
cana-5777	156	22	based	base	VERB
cana-5777	156	23	models	model	NOUN
cana-5777	156	24	better	well	ADV
cana-5777	156	25	than	than	ADP
cana-5777	156	26	traditional	traditional	ADJ
cana-5777	156	27	machine	machine	NOUN
cana-5777	156	28	learning	learn	VERB
cana-5777	156	29	techniques	technique	NOUN
cana-5777	156	30	in	in	ADP
cana-5777	156	31	medical	medical	ADJ
cana-5777	156	32	classification	classification	NOUN
cana-5777	156	33	problems	problem	NOUN
cana-5777	156	34	.	.	PUNCT
cana-5777	157	1	even	even	ADV
cana-5777	157	2	while	while	SCONJ
cana-5777	157	3	advanced	advanced	ADJ
cana-5777	157	4	machine	machine	NOUN
cana-5777	157	5	learning	learning	NOUN
cana-5777	157	6	models	model	NOUN
cana-5777	157	7	like	like	ADP
cana-5777	157	8	adaboost	adaboost	ADV
cana-5777	157	9	and	and	CCONJ
cana-5777	157	10	lightgbm	lightgbm	ADJ
cana-5777	157	11	obtained	obtain	VERB
cana-5777	157	12	a	a	DET
cana-5777	157	13	high	high	ADJ
cana-5777	157	14	accuracy	accuracy	NOUN
cana-5777	157	15	of	of	ADP
cana-5777	157	16	98.83	98.83	NUM
cana-5777	157	17	%	%	NOUN
cana-5777	157	18	where	where	SCONJ
cana-5777	157	19	graph	graph	NOUN
cana-5777	157	20	-	-	PUNCT
cana-5777	157	21	based	base	VERB
cana-5777	157	22	models	model	NOUN
cana-5777	157	23	performed	perform	VERB
cana-5777	157	24	significantly	significantly	ADV
cana-5777	157	25	better	well	ADV
cana-5777	157	26	.	.	PUNCT
cana-5777	158	1	even	even	ADV
cana-5777	158	2	the	the	DET
cana-5777	158	3	most	most	ADV
cana-5777	158	4	advanced	advanced	ADJ
cana-5777	158	5	conventional	conventional	ADJ
cana-5777	158	6	ml	ml	NOUN
cana-5777	158	7	models	model	NOUN
cana-5777	158	8	were	be	AUX
cana-5777	158	9	outperforming	outperform	VERB
cana-5777	158	10	by	by	ADP
cana-5777	158	11	gin	gin	NOUN
cana-5777	158	12	,	,	PUNCT
cana-5777	158	13	which	which	PRON
cana-5777	158	14	obtained	obtain	VERB
cana-5777	158	15	99.42	99.42	NUM
cana-5777	158	16	%	%	NOUN
cana-5777	158	17	accuracy	accuracy	NOUN
cana-5777	158	18	,	,	PUNCT
cana-5777	158	19	while	while	SCONJ
cana-5777	158	20	gcn	gcn	ADJ
cana-5777	158	21	,	,	PUNCT
cana-5777	158	22	gat	gat	NOUN
cana-5777	158	23	and	and	CCONJ
cana-5777	158	24	graphsage	graphsage	NOUN
cana-5777	158	25	achieved	achieve	VERB
cana-5777	158	26	100	100	NUM
cana-5777	158	27	%	%	NOUN
cana-5777	158	28	accuracy	accuracy	NOUN
cana-5777	158	29	.	.	PUNCT
cana-5777	159	1	the	the	DET
cana-5777	159	2	effect	effect	NOUN
cana-5777	159	3	is	be	AUX
cana-5777	159	4	further	far	ADV
cana-5777	159	5	shown	show	VERB
cana-5777	159	6	by	by	ADP
cana-5777	159	7	the	the	DET
cana-5777	159	8	confusion	confusion	NOUN
cana-5777	159	9	matrices	matrix	NOUN
cana-5777	159	10	,	,	PUNCT
cana-5777	159	11	which	which	PRON
cana-5777	159	12	show	show	VERB
cana-5777	159	13	that	that	SCONJ
cana-5777	159	14	graph	graph	NOUN
cana-5777	159	15	-	-	PUNCT
cana-5777	159	16	based	base	VERB
cana-5777	159	17	models	model	NOUN
cana-5777	159	18	achieved	achieve	VERB
cana-5777	159	19	very	very	ADV
cana-5777	159	20	reliable	reliable	ADJ
cana-5777	159	21	classification	classification	NOUN
cana-5777	159	22	by	by	ADP
cana-5777	159	23	consistently	consistently	ADV
cana-5777	159	24	minimizing	minimize	VERB
cana-5777	159	25	both	both	DET
cana-5777	159	26	false	false	ADJ
cana-5777	159	27	positives	positive	NOUN
cana-5777	159	28	and	and	CCONJ
cana-5777	159	29	false	false	ADJ
cana-5777	159	30	negatives	negative	NOUN
cana-5777	159	31	.	.	PUNCT
cana-5777	160	1	on	on	ADP
cana-5777	160	2	the	the	DET
cana-5777	160	3	other	other	ADJ
cana-5777	160	4	hand	hand	NOUN
cana-5777	160	5	,	,	PUNCT
cana-5777	160	6	traditional	traditional	ADJ
cana-5777	160	7	machine	machine	NOUN
cana-5777	160	8	learning	learn	VERB
cana-5777	160	9	algorithms	algorithm	NOUN
cana-5777	160	10	were	be	AUX
cana-5777	160	11	unable	unable	ADJ
cana-5777	160	12	to	to	PART
cana-5777	160	13	achieve	achieve	VERB
cana-5777	160	14	an	an	DET
cana-5777	160	15	appropriate	appropriate	ADJ
cana-5777	160	16	balance	balance	NOUN
cana-5777	160	17	between	between	ADP
cana-5777	160	18	recall	recall	NOUN
cana-5777	160	19	and	and	CCONJ
cana-5777	160	20	precision	precision	NOUN
cana-5777	160	21	,	,	PUNCT
cana-5777	160	22	especially	especially	ADV
cana-5777	160	23	models	model	NOUN
cana-5777	160	24	that	that	PRON
cana-5777	160	25	utilize	utilize	VERB
cana-5777	160	26	linear	linear	ADJ
cana-5777	160	27	decision	decision	NOUN
cana-5777	160	28	boundaries	boundary	NOUN
cana-5777	160	29	(	(	PUNCT
cana-5777	160	30	such	such	ADJ
cana-5777	160	31	as	as	ADP
cana-5777	160	32	sgd	sgd	PROPN
cana-5777	160	33	,	,	PUNCT
cana-5777	160	34	which	which	PRON
cana-5777	160	35	had	have	VERB
cana-5777	160	36	an	an	DET
cana-5777	160	37	accuracy	accuracy	NOUN
cana-5777	160	38	of	of	ADP
cana-5777	160	39	85.38	85.38	NUM
cana-5777	160	40	%	%	NOUN
cana-5777	160	41	)	)	PUNCT
cana-5777	160	42	.	.	PUNCT
cana-5777	161	1	even	even	ADV
cana-5777	161	2	the	the	DET
cana-5777	161	3	best	well	ADV
cana-5777	161	4	-	-	PUNCT
cana-5777	161	5	performing	perform	VERB
cana-5777	161	6	ml	ml	NOUN
cana-5777	161	7	models	model	NOUN
cana-5777	161	8	showed	show	VERB
cana-5777	161	9	occasional	occasional	ADJ
cana-5777	161	10	misclassifications	misclassification	NOUN
cana-5777	161	11	,	,	PUNCT
cana-5777	161	12	whereas	whereas	SCONJ
cana-5777	161	13	gnns	gnns	PROPN
cana-5777	161	14	handled	handle	VERB
cana-5777	161	15	complex	complex	ADJ
cana-5777	161	16	structural	structural	ADJ
cana-5777	161	17	dependencies	dependency	NOUN
cana-5777	161	18	in	in	ADP
cana-5777	161	19	the	the	DET
cana-5777	161	20	dataset	dataset	NOUN
cana-5777	161	21	with	with	ADP
cana-5777	161	22	remarkable	remarkable	ADJ
cana-5777	161	23	efficiency	efficiency	NOUN
cana-5777	161	24	.	.	PUNCT
cana-5777	162	1	these	these	DET
cana-5777	162	2	results	result	NOUN
cana-5777	162	3	highlight	highlight	VERB
cana-5777	162	4	the	the	DET
cana-5777	162	5	ability	ability	NOUN
cana-5777	162	6	of	of	ADP
cana-5777	162	7	gnns	gnns	NOUN
cana-5777	162	8	to	to	PART
cana-5777	162	9	capture	capture	VERB
cana-5777	162	10	complex	complex	ADJ
cana-5777	162	11	relationships	relationship	NOUN
cana-5777	162	12	within	within	ADP
cana-5777	162	13	medical	medical	ADJ
cana-5777	162	14	datasets	dataset	NOUN
cana-5777	162	15	,	,	PUNCT
cana-5777	162	16	making	make	VERB
cana-5777	162	17	them	they	PRON
cana-5777	162	18	a	a	DET
cana-5777	162	19	highly	highly	ADV
cana-5777	162	20	promising	promising	ADJ
cana-5777	162	21	approach	approach	NOUN
cana-5777	162	22	for	for	ADP
cana-5777	162	23	real	real	ADJ
cana-5777	162	24	-	-	PUNCT
cana-5777	162	25	world	world	NOUN
cana-5777	162	26	applications	application	NOUN
cana-5777	162	27	.	.	PUNCT
cana-5777	163	1	their	their	PRON
cana-5777	163	2	efficacy	efficacy	NOUN
cana-5777	163	3	lies	lie	VERB
cana-5777	163	4	on	on	ADP
cana-5777	163	5	their	their	PRON
cana-5777	163	6	capacity	capacity	NOUN
cana-5777	163	7	to	to	PART
cana-5777	163	8	represent	represent	VERB
cana-5777	163	9	complex	complex	ADJ
cana-5777	163	10	structural	structural	ADJ
cana-5777	163	11	dependencies	dependency	NOUN
cana-5777	163	12	in	in	ADP
cana-5777	163	13	data	datum	NOUN
cana-5777	163	14	,	,	PUNCT
cana-5777	163	15	which	which	PRON
cana-5777	163	16	are	be	AUX
cana-5777	163	17	difficult	difficult	ADJ
cana-5777	163	18	for	for	SCONJ
cana-5777	163	19	traditional	traditional	ADJ
cana-5777	163	20	machine	machine	NOUN
cana-5777	163	21	learning	learning	NOUN
cana-5777	163	22	models	model	NOUN
cana-5777	163	23	to	to	PART
cana-5777	163	24	capture	capture	VERB
cana-5777	163	25	.	.	PUNCT
cana-5777	164	1	future	future	ADJ
cana-5777	164	2	work	work	NOUN
cana-5777	164	3	will	will	AUX
cana-5777	164	4	focus	focus	VERB
cana-5777	164	5	on	on	ADP
cana-5777	164	6	exploring	explore	VERB
cana-5777	164	7	more	more	ADV
cana-5777	164	8	advanced	advanced	ADJ
cana-5777	164	9	gnn	gnn	PROPN
cana-5777	164	10	architectures	architecture	NOUN
cana-5777	164	11	,	,	PUNCT
cana-5777	164	12	optimizing	optimize	VERB
cana-5777	164	13	computational	computational	ADJ
cana-5777	164	14	efficiency	efficiency	NOUN
cana-5777	164	15	,	,	PUNCT
cana-5777	164	16	and	and	CCONJ
cana-5777	164	17	expanding	expand	VERB
cana-5777	164	18	the	the	DET
cana-5777	164	19	study	study	NOUN
cana-5777	164	20	to	to	ADP
cana-5777	164	21	larger	large	ADJ
cana-5777	164	22	and	and	CCONJ
cana-5777	164	23	more	more	ADV
cana-5777	164	24	diverse	diverse	ADJ
cana-5777	164	25	medical	medical	ADJ
cana-5777	164	26	datasets	dataset	NOUN
cana-5777	164	27	to	to	PART
cana-5777	164	28	further	far	ADV
cana-5777	164	29	validate	validate	VERB
cana-5777	164	30	these	these	DET
cana-5777	164	31	findings	finding	NOUN
cana-5777	164	32	.	.	PUNCT
cana-5777	165	1	communications	communication	NOUN
cana-5777	165	2	on	on	ADP
cana-5777	165	3	applied	apply	VERB
cana-5777	165	4	nonlinear	nonlinear	ADJ
cana-5777	165	5	analysis	analysis	NOUN
cana-5777	165	6	issn	issn	NOUN
cana-5777	165	7	:	:	PUNCT
cana-5777	165	8	1074	1074	NUM
cana-5777	165	9	-	-	PUNCT
cana-5777	165	10	133x	133x	NUM
cana-5777	165	11	vol	vol	NOUN
cana-5777	165	12	31	31	NUM
cana-5777	165	13	no	no	NOUN
cana-5777	165	14	.	.	PUNCT
cana-5777	166	1	8s	8s	PROPN
cana-5777	166	2	(	(	PUNCT
cana-5777	166	3	2024	2024	NUM
cana-5777	166	4	)	)	PUNCT
cana-5777	166	5	https://internationalpubls.com	https://internationalpubls.com	X
cana-5777	166	6	1071	1071	NUM
cana-5777	166	7	references	reference	NOUN
cana-5777	166	8	[	[	X
cana-5777	166	9	1	1	NUM
cana-5777	166	10	]	]	PUNCT
cana-5777	166	11	scarselli	scarselli	NOUN
cana-5777	166	12	,	,	PUNCT
cana-5777	166	13	f.	f.	PROPN
cana-5777	166	14	,	,	PUNCT
cana-5777	166	15	gori	gori	PROPN
cana-5777	166	16	,	,	PUNCT
cana-5777	166	17	m.	m.	NOUN
cana-5777	166	18	,	,	PUNCT
cana-5777	166	19	tsoi	tsoi	PROPN
cana-5777	166	20	,	,	PUNCT
cana-5777	166	21	a.	a.	PROPN
cana-5777	166	22	c.	c.	PROPN
cana-5777	166	23	,	,	PUNCT
cana-5777	166	24	hagenbuchner	hagenbuchner	ADJ
cana-5777	166	25	,	,	PUNCT
cana-5777	166	26	m.	m.	NOUN
cana-5777	166	27	,	,	PUNCT
cana-5777	166	28	&	&	CCONJ
cana-5777	166	29	monfardini	monfardini	ADJ
cana-5777	166	30	,	,	PUNCT
cana-5777	166	31	g.	g.	PROPN
cana-5777	166	32	(	(	PUNCT
cana-5777	166	33	2009	2009	NUM
cana-5777	166	34	)	)	PUNCT
cana-5777	166	35	.	.	PUNCT
cana-5777	167	1	the	the	DET
cana-5777	167	2	graph	graph	NOUN
cana-5777	167	3	neural	neural	ADJ
cana-5777	167	4	network	network	NOUN
cana-5777	167	5	model	model	NOUN
cana-5777	167	6	.	.	PUNCT
cana-5777	168	1	ieee	ieee	NOUN
cana-5777	168	2	transactions	transaction	NOUN
cana-5777	168	3	on	on	ADP
cana-5777	168	4	neural	neural	ADJ
cana-5777	168	5	networks.https://ieeexplore.ieee.org/document/4700287	networks.https://ieeexplore.ieee.org/document/4700287	PROPN
cana-5777	168	6	[	[	X
cana-5777	168	7	2	2	NUM
cana-5777	168	8	]	]	X
cana-5777	168	9	kipf	kipf	NOUN
cana-5777	168	10	,	,	PUNCT
cana-5777	168	11	t.	t.	PROPN
cana-5777	168	12	n.	n.	PROPN
cana-5777	168	13	,	,	PUNCT
cana-5777	168	14	&	&	CCONJ
cana-5777	168	15	welling	well	VERB
cana-5777	168	16	,	,	PUNCT
cana-5777	168	17	m.	m.	NOUN
cana-5777	168	18	(	(	PUNCT
cana-5777	168	19	2017	2017	NUM
cana-5777	168	20	)	)	PUNCT
cana-5777	168	21	.	.	PUNCT
cana-5777	169	1	semi	semi	ADJ
cana-5777	169	2	-	-	ADJ
cana-5777	169	3	supervised	supervised	ADJ
cana-5777	169	4	classification	classification	NOUN
cana-5777	169	5	with	with	ADP
cana-5777	169	6	graph	graph	NOUN
cana-5777	169	7	convolutional	convolutional	ADJ
cana-5777	169	8	networks.international	networks.international	ADJ
cana-5777	169	9	conference	conference	NOUN
cana-5777	169	10	on	on	ADP
cana-5777	169	11	learning	learn	VERB
cana-5777	169	12	representations	representation	NOUN
cana-5777	169	13	(	(	PUNCT
cana-5777	169	14	iclr).https://arxiv.org	iclr).https://arxiv.org	X
cana-5777	169	15	/	/	SYM
cana-5777	169	16	pdf/1609.02907	pdf/1609.02907	NOUN
cana-5777	170	1	[	[	X
cana-5777	170	2	3	3	NUM
cana-5777	170	3	]	]	X
cana-5777	170	4	xu	xu	PROPN
cana-5777	170	5	,	,	PUNCT
cana-5777	170	6	k.	k.	PROPN
cana-5777	170	7	,	,	PUNCT
cana-5777	170	8	hu	hu	PROPN
cana-5777	170	9	,	,	PUNCT
cana-5777	170	10	w.	w.	PROPN
cana-5777	170	11	,	,	PUNCT
cana-5777	170	12	leskovec	leskovec	PROPN
cana-5777	170	13	,	,	PUNCT
cana-5777	170	14	j.	j.	PROPN
cana-5777	170	15	,	,	PUNCT
cana-5777	170	16	&	&	CCONJ
cana-5777	170	17	jegelka	jegelka	PROPN
cana-5777	170	18	,	,	PUNCT
cana-5777	170	19	s.	s.	PROPN
cana-5777	170	20	(	(	PUNCT
cana-5777	170	21	2019	2019	NUM
cana-5777	170	22	)	)	PUNCT
cana-5777	170	23	.	.	PUNCT
cana-5777	171	1	how	how	SCONJ
cana-5777	171	2	powerful	powerful	ADJ
cana-5777	171	3	are	be	AUX
cana-5777	171	4	graph	graph	VERB
cana-5777	171	5	neural	neural	ADJ
cana-5777	171	6	networks	network	NOUN
cana-5777	171	7	?	?	PUNCT
cana-5777	172	1	iclr.https://arxiv.org/abs/1810.00826	iclr.https://arxiv.org/abs/1810.00826	PROPN
cana-5777	173	1	[	[	X
cana-5777	173	2	4	4	NUM
cana-5777	173	3	]	]	X
cana-5777	173	4	veličković	veličković	ADJ
cana-5777	173	5	,	,	PUNCT
cana-5777	173	6	p.	p.	NOUN
cana-5777	173	7	,	,	PUNCT
cana-5777	173	8	cucurull	cucurull	NOUN
cana-5777	173	9	,	,	PUNCT
cana-5777	173	10	g.	g.	PROPN
cana-5777	173	11	,	,	PUNCT
cana-5777	173	12	casanova	casanova	PROPN
cana-5777	173	13	,	,	PUNCT
cana-5777	173	14	a.	a.	PROPN
cana-5777	173	15	,	,	PUNCT
cana-5777	173	16	romero	romero	PROPN
cana-5777	173	17	,	,	PUNCT
cana-5777	173	18	a.	a.	PROPN
cana-5777	173	19	,	,	PUNCT
cana-5777	173	20	lio	lio	PROPN
cana-5777	173	21	,	,	PUNCT
cana-5777	173	22	p.	p.	PROPN
cana-5777	173	23	,	,	PUNCT
cana-5777	173	24	&	&	CCONJ
cana-5777	173	25	bengio	bengio	PROPN
cana-5777	173	26	,	,	PUNCT
cana-5777	173	27	y.	y.	PROPN
cana-5777	173	28	(	(	PUNCT
cana-5777	173	29	2018	2018	NUM
cana-5777	173	30	)	)	PUNCT
cana-5777	173	31	.	.	PUNCT
cana-5777	174	1	graph	graph	VERB
cana-5777	174	2	attention	attention	NOUN
cana-5777	174	3	networks.international	networks.international	ADJ
cana-5777	174	4	conference	conference	NOUN
cana-5777	174	5	on	on	ADP
cana-5777	174	6	learning	learn	VERB
cana-5777	174	7	representations	representation	NOUN
cana-5777	174	8	(	(	PUNCT
cana-5777	174	9	iclr).https://arxiv.org	iclr).https://arxiv.org	X
cana-5777	174	10	/	/	SYM
cana-5777	174	11	pdf/1710.10903	pdf/1710.10903	PROPN
cana-5777	174	12	[	[	X
cana-5777	174	13	5	5	NUM
cana-5777	174	14	]	]	X
cana-5777	174	15	hamilton	hamilton	PROPN
cana-5777	174	16	,	,	PUNCT
cana-5777	174	17	w.	w.	PROPN
cana-5777	174	18	,	,	PUNCT
cana-5777	174	19	ying	ying	PROPN
cana-5777	174	20	,	,	PUNCT
cana-5777	174	21	r.	r.	PROPN
cana-5777	174	22	,	,	PUNCT
cana-5777	174	23	&	&	CCONJ
cana-5777	174	24	leskovec	leskovec	PROPN
cana-5777	174	25	,	,	PUNCT
cana-5777	174	26	j.	j.	PROPN
cana-5777	174	27	(	(	PUNCT
cana-5777	174	28	2017	2017	NUM
cana-5777	174	29	)	)	PUNCT
cana-5777	174	30	.	.	PUNCT
cana-5777	175	1	inductive	inductive	ADJ
cana-5777	175	2	representation	representation	NOUN
cana-5777	175	3	learning	learn	VERB
cana-5777	175	4	on	on	ADP
cana-5777	175	5	large	large	ADJ
cana-5777	175	6	graphs.advances	graphs.advance	NOUN
cana-5777	175	7	in	in	ADP
cana-5777	175	8	neural	neural	ADJ
cana-5777	175	9	information	information	NOUN
cana-5777	175	10	processing	processing	NOUN
cana-5777	175	11	systems	system	NOUN
cana-5777	175	12	(	(	PUNCT
cana-5777	175	13	neurips	neurip	NOUN
cana-5777	175	14	)	)	PUNCT
cana-5777	175	15	,	,	PUNCT
cana-5777	175	16	30	30	NUM
cana-5777	175	17	.	.	PUNCT
cana-5777	175	18	https://arxiv.org/pdf/1706.02216	https://arxiv.org/pdf/1706.02216	PROPN
cana-5777	175	19	[	[	X
cana-5777	175	20	6	6	NUM
cana-5777	175	21	]	]	X
cana-5777	175	22	zhang	zhang	PROPN
cana-5777	175	23	,	,	PUNCT
cana-5777	175	24	c.	c.	PROPN
cana-5777	175	25	,	,	PUNCT
cana-5777	175	26	song	song	NOUN
cana-5777	175	27	,	,	PUNCT
cana-5777	175	28	d.	d.	PROPN
cana-5777	175	29	,	,	PUNCT
cana-5777	175	30	huang	huang	PROPN
cana-5777	175	31	,	,	PUNCT
cana-5777	175	32	c.	c.	PROPN
cana-5777	175	33	,	,	PUNCT
cana-5777	175	34	swami	swami	PROPN
cana-5777	175	35	,	,	PUNCT
cana-5777	175	36	a.	a.	PROPN
cana-5777	175	37	,	,	PUNCT
cana-5777	175	38	&	&	CCONJ
cana-5777	175	39	chawla	chawla	PROPN
cana-5777	175	40	,	,	PUNCT
cana-5777	175	41	n.	n.	PROPN
cana-5777	175	42	v.	v.	PROPN
cana-5777	175	43	(	(	PUNCT
cana-5777	175	44	2019	2019	NUM
cana-5777	175	45	)	)	PUNCT
cana-5777	175	46	.	.	PUNCT
cana-5777	176	1	heterogeneous	heterogeneous	ADJ
cana-5777	176	2	graph	graph	NOUN
cana-5777	176	3	neural	neural	ADJ
cana-5777	176	4	network	network	NOUN
cana-5777	176	5	for	for	ADP
cana-5777	176	6	personalized	personalized	ADJ
cana-5777	176	7	healthcare.proceedings	healthcare.proceeding	NOUN
cana-5777	176	8	of	of	ADP
cana-5777	176	9	the	the	DET
cana-5777	176	10	25th	25th	ADJ
cana-5777	176	11	acm	acm	PROPN
cana-5777	176	12	sigkdd	sigkdd	NOUN
cana-5777	176	13	international	international	ADJ
cana-5777	176	14	conference	conference	NOUN
cana-5777	176	15	on	on	ADP
cana-5777	176	16	knowledge	knowledge	PROPN
cana-5777	176	17	discovery	discovery	PROPN
cana-5777	176	18	&	&	CCONJ
cana-5777	176	19	data	data	PROPN
cana-5777	176	20	mining	mining	NOUN
cana-5777	176	21	(	(	PUNCT
cana-5777	176	22	pp	pp	ADJ
cana-5777	176	23	.	.	PUNCT
cana-5777	177	1	1308–1316	1308–1316	NUM
cana-5777	177	2	)	)	PUNCT
cana-5777	177	3	.	.	PUNCT
cana-5777	178	1	acm	acm	PROPN
cana-5777	178	2	.	.	PUNCT
cana-5777	179	1	https://dl.acm.org/doi/pdf/10.1145/3292500.3330961	https://dl.acm.org/doi/pdf/10.1145/3292500.3330961	PUNCT
cana-5777	179	2	[	[	X
cana-5777	179	3	7	7	NUM
cana-5777	179	4	]	]	X
cana-5777	179	5	wang	wang	PROPN
cana-5777	179	6	,	,	PUNCT
cana-5777	179	7	s.	s.	PROPN
cana-5777	179	8	,	,	PUNCT
cana-5777	179	9	&	&	CCONJ
cana-5777	179	10	wang	wang	PROPN
cana-5777	179	11	,	,	PUNCT
cana-5777	179	12	s.	s.	PROPN
cana-5777	179	13	(	(	PUNCT
cana-5777	179	14	2023	2023	NUM
cana-5777	179	15	)	)	PUNCT
cana-5777	179	16	.	.	PUNCT
cana-5777	180	1	disease	disease	NOUN
cana-5777	180	2	prediction	prediction	NOUN
cana-5777	180	3	using	use	VERB
cana-5777	180	4	graph	graph	NOUN
cana-5777	180	5	machine	machine	NOUN
cana-5777	180	6	learning	learning	NOUN
cana-5777	180	7	based	base	VERB
cana-5777	180	8	on	on	ADP
cana-5777	180	9	electronic	electronic	ADJ
cana-5777	180	10	health	health	NOUN
cana-5777	180	11	data.healthcare	data.healthcare	NOUN
cana-5777	180	12	,	,	PUNCT
cana-5777	180	13	11(7	11(7	NUM
cana-5777	180	14	)	)	PUNCT
cana-5777	180	15	,	,	PUNCT
cana-5777	180	16	1031	1031	NUM
cana-5777	180	17	.	.	PUNCT
cana-5777	181	1	https://www.mdpi.com/22279032/11/7/1031	https://www.mdpi.com/22279032/11/7/1031	PRON
cana-5777	182	1	[	[	X
cana-5777	182	2	8	8	NUM
cana-5777	182	3	]	]	X
cana-5777	182	4	mahmud	mahmud	PROPN
cana-5777	182	5	,	,	PUNCT
cana-5777	182	6	f.	f.	PROPN
cana-5777	182	7	b.	b.	PROPN
cana-5777	182	8	,	,	PUNCT
cana-5777	182	9	rayhan	rayhan	PROPN
cana-5777	182	10	,	,	PUNCT
cana-5777	182	11	m.	m.	PROPN
cana-5777	182	12	md	md	PROPN
cana-5777	182	13	.	.	PUNCT
cana-5777	183	1	s.	s.	PROPN
cana-5777	183	2	,	,	PUNCT
cana-5777	183	3	shuvo	shuvo	PROPN
cana-5777	183	4	,	,	PUNCT
cana-5777	183	5	m.	m.	PROPN
cana-5777	183	6	h.	h.	PROPN
cana-5777	183	7	,	,	PUNCT
cana-5777	183	8	sadia	sadia	PROPN
cana-5777	183	9	,	,	PUNCT
cana-5777	183	10	i.	i.	PROPN
cana-5777	183	11	,	,	PUNCT
cana-5777	183	12	&	&	CCONJ
cana-5777	183	13	morol	morol	PROPN
cana-5777	183	14	,	,	PUNCT
cana-5777	183	15	m.	m.	PROPN
cana-5777	183	16	k.	k.	PROPN
cana-5777	183	17	(	(	PUNCT
cana-5777	183	18	2022	2022	NUM
cana-5777	183	19	)	)	PUNCT
cana-5777	183	20	.	.	PUNCT
cana-5777	184	1	a	a	DET
cana-5777	184	2	comparative	comparative	ADJ
cana-5777	184	3	analysis	analysis	NOUN
cana-5777	184	4	of	of	ADP
cana-5777	184	5	graph	graph	NOUN
cana-5777	184	6	neural	neural	ADJ
cana-5777	184	7	networks	network	NOUN
cana-5777	184	8	and	and	CCONJ
cana-5777	184	9	commonly	commonly	ADV
cana-5777	184	10	used	use	VERB
cana-5777	184	11	machine	machine	NOUN
cana-5777	184	12	learning	learn	VERB
cana-5777	184	13	algorithms	algorithm	NOUN
cana-5777	184	14	on	on	ADP
cana-5777	184	15	fake	fake	ADJ
cana-5777	184	16	news	news	NOUN
cana-5777	184	17	detection.ieee	detection.ieee	PROPN
cana-5777	184	18	access.https://arxiv.org/pdf/2203.14132	access.https://arxiv.org/pdf/2203.14132	PROPN
cana-5777	185	1	[	[	X
cana-5777	185	2	9	9	NUM
cana-5777	185	3	]	]	X
cana-5777	185	4	wang	wang	PROPN
cana-5777	185	5	,	,	PUNCT
cana-5777	185	6	y.	y.	PROPN
cana-5777	185	7	,	,	PUNCT
cana-5777	185	8	ding	ding	NOUN
cana-5777	185	9	,	,	PUNCT
cana-5777	185	10	a.	a.	PROPN
cana-5777	185	11	,	,	PUNCT
cana-5777	185	12	guan	guan	PROPN
cana-5777	185	13	,	,	PUNCT
cana-5777	185	14	k.	k.	PROPN
cana-5777	185	15	,	,	PUNCT
cana-5777	185	16	we	we	PRON
cana-5777	185	17	,	,	PUNCT
cana-5777	185	18	s.	s.	PROPN
cana-5777	185	19	,	,	PUNCT
cana-5777	185	20	&	&	CCONJ
cana-5777	185	21	du	du	PROPN
cana-5777	185	22	,	,	PUNCT
cana-5777	185	23	y.	y.	PROPN
cana-5777	185	24	(	(	PUNCT
cana-5777	185	25	2022	2022	NUM
cana-5777	185	26	)	)	PUNCT
cana-5777	185	27	.	.	PUNCT
cana-5777	186	1	graph	graph	NOUN
cana-5777	186	2	-	-	PUNCT
cana-5777	186	3	based	base	VERB
cana-5777	186	4	ensemble	ensemble	ADJ
cana-5777	186	5	machine	machine	NOUN
cana-5777	186	6	learning	learn	VERB
cana-5777	186	7	for	for	ADP
cana-5777	186	8	student	student	NOUN
cana-5777	186	9	performance	performance	NOUN
cana-5777	186	10	prediction.arxiv	prediction.arxiv	NUM
cana-5777	186	11	preprint.https://arxiv.org/pdf/2112.07893	preprint.https://arxiv.org/pdf/2112.07893	NOUN
cana-5777	187	1	[	[	X
cana-5777	187	2	10]shaila	10]shaila	NUM
cana-5777	187	3	,	,	PUNCT
cana-5777	187	4	s.	s.	PROPN
cana-5777	187	5	j.	j.	PROPN
cana-5777	187	6	,	,	PUNCT
cana-5777	187	7	&	&	CCONJ
cana-5777	187	8	varsha	varsha	PROPN
cana-5777	187	9	,	,	PUNCT
cana-5777	187	10	h.	h.	PROPN
cana-5777	187	11	s.	s.	PROPN
cana-5777	187	12	(	(	PUNCT
cana-5777	187	13	2024	2024	NUM
cana-5777	187	14	)	)	PUNCT
cana-5777	187	15	.	.	PUNCT
cana-5777	188	1	application	application	NOUN
cana-5777	188	2	of	of	ADP
cana-5777	188	3	graph	graph	NOUN
cana-5777	188	4	theory	theory	NOUN
cana-5777	188	5	in	in	ADP
cana-5777	188	6	machine	machine	NOUN
cana-5777	188	7	learning	learning	NOUN
cana-5777	188	8	.	.	PUNCT
cana-5777	189	1	researchgate	researchgate	NOUN
cana-5777	189	2	.	.	PUNCT
cana-5777	190	1	https://www.researchgate.net/publication/379825214_applications_of_graph_theor	https://www.researchgate.net/publication/379825214_applications_of_graph_theor	NOUN
cana-5777	190	2	y_in_machine_learning	y_in_machine_learne	VERB
cana-5777	190	3	[	[	X
cana-5777	190	4	11	11	NUM
cana-5777	190	5	]	]	X
cana-5777	190	6	erdem	erdem	PROPN
cana-5777	190	7	,	,	PUNCT
cana-5777	190	8	t.	t.	PROPN
cana-5777	190	9	(	(	PUNCT
cana-5777	190	10	2023	2023	NUM
cana-5777	190	11	)	)	PUNCT
cana-5777	190	12	.	.	PUNCT
cana-5777	191	1	cancer	cancer	NOUN
cana-5777	191	2	data	datum	NOUN
cana-5777	191	3	[	[	X
cana-5777	191	4	data	datum	NOUN
cana-5777	191	5	set	set	NOUN
cana-5777	191	6	]	]	PUNCT
cana-5777	191	7	.	.	PUNCT
cana-5777	192	1	kaggle	kaggle	PROPN
cana-5777	192	2	.	.	PUNCT
cana-5777	193	1	https://www.kaggle.com/datasets/erdemtaha/cancer-data	https://www.kaggle.com/datasets/erdemtaha/cancer-data	NOUN
cana-5777	193	2	communications	communication	NOUN
cana-5777	193	3	on	on	ADP
cana-5777	193	4	applied	apply	VERB
cana-5777	193	5	nonlinear	nonlinear	ADJ
cana-5777	193	6	analysis	analysis	NOUN
cana-5777	193	7	issn	issn	NOUN
cana-5777	193	8	:	:	PUNCT
cana-5777	193	9	1074	1074	NUM
cana-5777	193	10	-	-	PUNCT
cana-5777	193	11	133x	133x	NUM
cana-5777	193	12	vol	vol	NOUN
cana-5777	193	13	31	31	NUM
cana-5777	193	14	no	no	NOUN
cana-5777	193	15	.	.	PUNCT
cana-5777	194	1	8s	8s	PROPN
cana-5777	194	2	(	(	PUNCT
cana-5777	194	3	2024	2024	NUM
cana-5777	194	4	)	)	PUNCT
cana-5777	194	5	https://internationalpubls.com	https://internationalpubls.com	X
cana-5777	194	6	1072	1072	NUM
cana-5777	194	7	https://ieeexplore.ieee.org/document/4700287	https://ieeexplore.ieee.org/document/4700287	PROPN
cana-5777	194	8	https://arxiv.org/pdf/1609.02907	https://arxiv.org/pdf/1609.02907	PROPN
cana-5777	194	9	https://arxiv.org/abs/1810.00826	https://arxiv.org/abs/1810.00826	ADV
cana-5777	194	10	https://arxiv.org/pdf/1710.10903	https://arxiv.org/pdf/1710.10903	PROPN
cana-5777	194	11	https://arxiv.org/pdf/1706.02216	https://arxiv.org/pdf/1706.02216	PROPN
cana-5777	194	12	https://dl.acm.org/doi/pdf/10.1145/3292500.3330961	https://dl.acm.org/doi/pdf/10.1145/3292500.3330961	PROPN
cana-5777	194	13	https://www.mdpi.com/2227-9032/11/7/1031	https://www.mdpi.com/2227-9032/11/7/1031	PROPN
cana-5777	194	14	https://arxiv.org/pdf/2203.14132	https://arxiv.org/pdf/2203.14132	NOUN
cana-5777	194	15	https://arxiv.org/pdf/2112.07893	https://arxiv.org/pdf/2112.07893	VERB
cana-5777	194	16	https://www.researchgate.net/publication/379825214_applications_of_graph_theory_in_machine_learning	https://www.researchgate.net/publication/379825214_applications_of_graph_theory_in_machine_learne	VERB
cana-5777	194	17	https://www.researchgate.net/publication/379825214_applications_of_graph_theory_in_machine_learning	https://www.researchgate.net/publication/379825214_applications_of_graph_theory_in_machine_learne	VERB
cana-5777	194	18	https://www.kaggle.com/datasets/erdemtaha/cancer-data	https://www.kaggle.com/datasets/erdemtaha/cancer-data	NOUN
cana-5777	194	19	article	article	NOUN
cana-5777	194	20	history	history	NOUN
cana-5777	194	21	:	:	PUNCT
cana-5777	194	22	received	receive	VERB
cana-5777	194	23	20.09.2024	20.09.2024	NUM
cana-5777	194	24	revised	revise	VERB
cana-5777	194	25	:	:	PUNCT
cana-5777	194	26	24.10.2024	24.10.2024	NUM
cana-5777	194	27	accepted	accept	VERB
cana-5777	194	28	:	:	PUNCT
cana-5777	194	29	30.11.2024	30.11.2024	NUM
cana-5777	194	30	abstract	abstract	ADJ
cana-5777	194	31	2	2	NUM
cana-5777	194	32	.	.	PUNCT
cana-5777	194	33	literature	literature	NOUN
cana-5777	194	34	review	review	NOUN
cana-5777	194	35	references	reference	NOUN