Attention-Based Learning for Predicting Drug-Drug Interactions in Knowledge Graph Embedding Based on Multisource Fusion Information

Table 2

Temperature and wildlife count in the three areas covered by the study.

Methods	KEGG				OGB-biokg
	AUC	Acc	F1	AUPR	AUC	Acc	F1	AUPR
LAP	0.8677	0.7981	0.8047	0.8354	0.9428	0.9053	0.8992	0.9313
DeepWalk	0.9121	0.8396	0.8451	0.8980	0.9900	0.9572	0.9581	0.9887
LINE	0.9040	0.8273	0.8338	0.8896	0.9910	0.9599	0.9607	0.9898
TransE	0.9319	0.8594	0.8651	0.9208	0.9930	0.9703	0.9708	0.9908
Complex	0.9454	0.8753	0.8812	0.9364	0.9942	0.9710	0.9715	0.9924
KGNN	0.9414	0.8845	0.8899	0.9187	0.9949	0.9715	0.972	0.9943
DDKG	0.9534	0.9012	0.9041	0.9359	0.9951	0.973	0.9733	0.9945
LaGAT	0.9643	0.9154	0.9175	0.9518	0.9952	0.9737	0.9739	0.9946
Ours	0.9705	0.9240	0.9257	0.9611	0.9959	0.9777	0.9779	0.9954