Knowledge Graph Completion on FB60K-NYT10 20% training ratio (test)
45.59Hits@5TagReal
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| TagRealModel Category=PLM-based2023.05 | 45.59 | 51.34 | 35.41 | |
| PKGCModel Category=PLM-based2023.05 | 35.77 | 43.82 | 28.62 | |
| TuckERModel Category=KGE-based2023.05 | 29.5 | 32.48 | 24.44 | |
| ConvEModel Category=KGE-based2023.05 | 29.49 | 33.3 | 24.31 | |
| TransEModel Category=KGE-based2023.05 | 29.13 | 32.67 | 15.8 | |
| JointNREModel Category=Text&KGE-based2023.05 | 16.93 | 20.74 | 11.39 | |
| RotatEModel Category=KGE-based2023.05 | 15.91 | 18.32 | 12.65 | |
| CPLModel Category=RL-based2023.05 | 15.19 | 18 | 10.87 | |
| RC-NetModel Category=Text&KGE-based2023.05 | 13.48 | 15.37 | 13.26 | |
| TransE+LineModel Category=Text&KGE-based2023.05 | 12.17 | 15.16 | 4.88 | |
| MINERVAModel Category=RL-based2023.05 | 11.64 | 14.16 | 8.93 | |
| ComplExModel Category=KGE-based2023.05 | 4.32 | 5.48 | 3.16 | |
| DisMultModel Category=KGE-based2023.05 | 3.44 | 4.31 | 2.64 |