Entity Classification on AM (test)
89.29AccuracyR-GCN
Evaluation Results
| Method | Links | |
|---|---|---|
| R-GCNlayers=2, hidden units=10, decomposition=basis function, optimizer=Adam, epochs=50, learning rate=0.012017.03 | 89.29 | |
| RDF2Vecdescription=Skipgram-based entity embeddings2017.03 | 88.33 | |
| WLdescription=Weisfeiler-Lehman kernels2017.03 | 87.37 | |
| Featdescription=hand-designed feature extractor (in- and out-degree per relation)2017.03 | 66.66 |