Recognition of Hotspot Words for Disease Symptoms Incorporating Contextual Weight and Co-Occurrence Degree

Table 3

Performance comparison of different disease symptom feature word extraction models.

Data set	Metrics	TextRank	YAKE	KeyBERT	MDERank	CW-MDERank
DS1	Recall	0.488	0.543	0.575	0.599	0.628
	Precision	0.455	0.52	0.535	0.552	0.589
	F1 score	0.471	0.531	0.554	0.575	0.608
DS2	Recall	0.491	0.559	0.599	0.608	0.639
	Precision	0.452	0.5	0.534	0.571	0.614
	F1 score	0.471	0.528	0.565	0.589	0.626
DS3	Recall	0.495	0.577	0.616	0.619	0.631
	Precision	0.435	0.477	0.488	0.554	0.581
	F1 score	0.463	0.522	0.545	0.585	0.605