Skip to main content

Table 20 HR@K results on Diginetica, \(K=[1, 3, 5, 10, 15, 20]\)

From: Improving session-based recommendation with contrastive learning

Method Diginetica
HR@1 HR@3 HR@5 HR@10 HR@15 HR@20
STAMP 0.0856 0.1962 0.2675 0.3821 0.4579 0.5126
STAMP-CL 0.0875 0.1975 0.2688 0.3857 0.4600 0.5152
RepeatNet 0.0771 0.1820 0.2532 0.3707 0.4461 0.5020
RepeatNet-CL 0.0795 0.1871 0.2854 0.3800 0.4580 0.5165
SR-GNN 0.0878 0.1918 0.2617 0.3757 0.4507 0.5037
SR-GNN-CL 0.0890 0.1970 0.2684 0.3834 0.4567 0.5103
SR-IEM 0.0829 0.1878 0.2596 0.3768 0.4540 0.5100
SR-IEM-CL 0.0844 0.1924 0.2618 0.3808 0.4570 0.5135
PIE 0.0861 0.1928 0.2657 0.3821 0.4589 0.5154
PIE-CL 0.0882 0.1976 0.2707 0.3887 0.4652 0.5220