Node Classification on Cora (Accuracy, G↑, L↑)
0.929AccuracySAGE
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| SAGENode Features=h_TAPE2023.05 | 0.929 | 5.28 | 18.13 | |
| RevGATNode Features=h_TAPE2023.05 | 0.928 | 4.14 | 18.04 | |
| GCNNode Features=h_TAPE2023.05 | 0.9119 | 2.33 | 16.59 | |
| GCNNode Features=h_shallow2023.05 | 0.8911 | — | — | |
| RevGATNode Features=h_shallow2023.05 | 0.8911 | — | — | |
| SAGENode Features=h_shallow2023.05 | 0.8824 | — | — | |
| MLPNode Features=h_TAPE2023.05 | 0.8778 | 37.41 | 13.35 | |
| SAGENode Features=h_GIANT2023.05 | 0.8455 | — | — | |
| GCNNode Features=h_GIANT2023.05 | 0.8423 | — | — | |
| RevGATNode Features=h_GIANT2023.05 | 0.8353 | — | — | |
| LM_finetuneModel=DeBERTa, Protocol=fine-tuned2023.05 | 0.7606 | — | — | |
| MLPNode Features=h_GIANT2023.05 | 0.7196 | — | — | |
| LLMModel=ChatGPT gpt-3.5-turbo, Protocol=zero-shot2023.05 | 0.6769 | — | — | |
| MLPNode Features=h_shallow2023.05 | 0.6388 | — | — |