Table 3.

Performance comparison with the state-of-the-art methods on three tasks of Dataset1

	ACC	AUPR	AUC	F1	Precision	Recall
Task1
MDDI-SCL	0.9378	0.9782	0.9983	0.8755	0.8804	0.8767
MDF-SA-DDI	0.9301	0.9737	0.9989	0.8878	0.9085	0.8760
DDIMDL	0.8852	0.9208	0.9976	0.7585	0.8471	0.7182
Lee et al.'s methods	0.9094	0.9562	0.9961	0.8391	0.8509	0.8339
DeepDDI	0.8371	0.8899	0.9961	0.6848	0.7275	0.6611
DNN	0.8797	0.9134	0.9963	0.7223	0.8047	0.7027
RF	0.7775	0.8349	0.9956	0.5936	0.7893	0.5161
KNN	0.7214	0.7716	0.9813	0.4831	0.7174	0.4081
LR	0.7920	0.8400	0.9960	0.5948	0.7437	0.5236
Task2
MDDI-SCL	0.6767	0.6947	0.9634	0.5304	0.6254	0.4814
MDF-SA-DDI	0.6633	0.6776	0.9497	0.5584	0.6547	0.5078
DDIMDL	0.6415	0.6558	0.9799	0.4460	0.5607	0.4319
Lee et al.'s methods	0.6405	0.6244	0.9247	0.5039	0.5388	0.4891
DeepDDI	0.5774	0.5594	0.9575	0.3416	0.3630	0.3890
DNN	0.6239	0.6361	0.9796	0.2997	0.4237	0.2840
Task3
MDDI-SCL	0.4589	0.3938	0.9053	0.1919	0.2585	0.1678
MDF-SA-DDI	0.4338	0.3873	0.8630	0.2329	0.2715	0.2226
DDIMDL	0.4075	0.3635	0.9512	0.1590	0.2408	0.1452
Lee et al.'s methods	0.4097	0.3184	0.8302	0.2022	0.2216	0.2027
DeepDDI	0.3602	0.2781	0.9059	0.1373	0.1586	0.1450
DNN	0.4087	0.3776	0.9550	0.1152	0.1836	0.1093