Node Classification on ACM-5
0.8792Macro F1 ScoreHGConv
Evaluation results
| Method | Links | ||
|---|---|---|---|
| HGConvTraining ratio=100%2020.12 | 0.8792 | 0.8889 | |
| HGConvRatio=100%, learning rate=0.008, dropout=0.82020.12 | 0.8792 | 0.8889 | |
| HGConvTraining ratio=80%2020.12 | 0.8766 | 0.8855 | |
| HGConvRatio=80%, learning rate=0.003, dropout=0.82020.12 | 0.8766 | 0.8855 | |
| RGCNTraining ratio=100%2020.12 | 0.8721 | 0.8841 | |
| RGCNRatio=100%, learning rate=0.001, dropout=0.52020.12 | 0.8721 | 0.8841 | |
| HGTTraining ratio=100%2020.12 | 0.8715 | 0.8825 | |
| HGTRatio=100%, learning rate=0.01, dropout=0.62020.12 | 0.8715 | 0.8825 | |
| HGConvTraining ratio=60%2020.12 | 0.8701 | 0.8794 | |
| HGConvRatio=60%, learning rate=0.003, dropout=0.82020.12 | 0.8701 | 0.8794 | |
| RGCNTraining ratio=80%2020.12 | 0.8699 | 0.8809 | |
| RGCNRatio=80%, learning rate=0.003, dropout=0.62020.12 | 0.8699 | 0.8809 | |
| HGTTraining ratio=80%2020.12 | 0.8692 | 0.878 | |
| HGTRatio=80%, learning rate=0.01, dropout=0.72020.12 | 0.8692 | 0.878 | |
| RGCNTraining ratio=60%2020.12 | 0.863 | 0.8722 | |
| RGCNRatio=60%, learning rate=0.003, dropout=0.62020.12 | 0.863 | 0.8722 | |
| HANTraining ratio=100%2020.12 | 0.8617 | 0.872 | |
| HANRatio=100%, learning rate=0.01, dropout=0.92020.12 | 0.8617 | 0.872 | |
| HANTraining ratio=80%2020.12 | 0.861 | 0.8715 | |
| HANRatio=80%, learning rate=0.05, dropout=0.92020.12 | 0.861 | 0.8715 | |
| HGTTraining ratio=60%2020.12 | 0.8573 | 0.8668 | |
| HGTRatio=60%, learning rate=0.01, dropout=0.62020.12 | 0.8573 | 0.8668 | |
| HANTraining ratio=60%2020.12 | 0.8526 | 0.8626 | |
| HANRatio=60%, learning rate=0.08, dropout=0.82020.12 | 0.8526 | 0.8626 | |
| GCNTraining ratio=100%2020.12 | 0.8492 | 0.8597 | |
| GCNRatio=100%, learning rate=0.005, dropout=02020.12 | 0.8492 | 0.8597 | |
| HGConvTraining ratio=40%2020.12 | 0.8478 | 0.8616 | |
| HGConvRatio=40%, learning rate=0.005, dropout=0.72020.12 | 0.8478 | 0.8616 | |
| GATTraining ratio=100%2020.12 | 0.8466 | 0.8572 | |
| GATRatio=100%, learning rate=0.03, dropout=0.62020.12 | 0.8466 | 0.8572 | |
| GATTraining ratio=80%2020.12 | 0.8459 | 0.8562 | |
| GATRatio=80%, learning rate=0.005, dropout=0.52020.12 | 0.8459 | 0.8562 | |
| GCNTraining ratio=80%2020.12 | 0.8448 | 0.8554 | |
| GCNRatio=80%, learning rate=0.01, dropout=0.42020.12 | 0.8448 | 0.8554 | |
| GATTraining ratio=60%2020.12 | 0.8441 | 0.8544 | |
| GATRatio=60%, learning rate=0.03, dropout=0.72020.12 | 0.8441 | 0.8544 | |
| GCNTraining ratio=60%2020.12 | 0.844 | 0.8545 | |
| GCNRatio=60%, learning rate=0.03, dropout=0.22020.12 | 0.844 | 0.8545 | |
| HGTTraining ratio=40%2020.12 | 0.8428 | 0.8573 | |
| HGTRatio=40%, learning rate=0.01, dropout=0.92020.12 | 0.8428 | 0.8573 | |
| HANTraining ratio=40%2020.12 | 0.8404 | 0.8525 | |
| HANRatio=40%, learning rate=0.05, dropout=0.52020.12 | 0.8404 | 0.8525 | |
| RGCNTraining ratio=40%2020.12 | 0.8368 | 0.8501 | |
| RGCNRatio=40%, learning rate=0.005, dropout=0.52020.12 | 0.8368 | 0.8501 | |
| GATTraining ratio=40%2020.12 | 0.8367 | 0.8475 | |
| GATRatio=40%, learning rate=0.05, dropout=0.52020.12 | 0.8367 | 0.8475 | |
| CHoEPre-trained dataset=ACM, Shots=5-shot, Evaluation protocol=fine-tuned2026.05 | 0.8327 | 0.8379 | |
| HEROPre-trained dataset=ACM, Shots=5-shot, Evaluation protocol=fine-tuned2026.05 | 0.8324 | 0.8383 | |
| GCNTraining ratio=40%2020.12 | 0.8317 | 0.8433 | |
| GCNRatio=40%, learning rate=0.01, dropout=0.52020.12 | 0.8317 | 0.8433 | |
| HGConvTraining ratio=20%2020.12 | 0.827 | 0.8428 | |
| HGConvRatio=20%, learning rate=0.005, dropout=0.52020.12 | 0.827 | 0.8428 | |
| GRACEPre-trained dataset=ACM, Shots=5-shot, Evaluation protocol=fine-tuned2026.05 | 0.8255 | 0.8201 | |
| GATTraining ratio=20%2020.12 | 0.8253 | 0.8388 | |
| GATRatio=20%, learning rate=0.01, dropout=0.62020.12 | 0.8253 | 0.8388 | |
| GCNTraining ratio=20%2020.12 | 0.8221 | 0.8364 | |
| GCNRatio=20%, learning rate=0.005, dropout=0.52020.12 | 0.8221 | 0.8364 | |
| GPF-plusPre-trained dataset=ACM, Shots=5-shot, Evaluation protocol=fine-tuned2026.05 | 0.822 | 0.8267 | |
| HANTraining ratio=20%2020.12 | 0.8191 | 0.8334 | |
| HANRatio=20%, learning rate=0.01, dropout=0.52020.12 | 0.8191 | 0.8334 | |
| DGIPre-trained dataset=ACM, Shots=5-shot, Evaluation protocol=fine-tuned2026.05 | 0.8185 | 0.8173 | |
| RGCNTraining ratio=20%2020.12 | 0.8148 | 0.8333 | |
| RGCNRatio=20%, learning rate=0.03, dropout=0.52020.12 | 0.8148 | 0.8333 | |
| HGMAEPre-trained dataset=ACM, Shots=5-shot, Evaluation protocol=fine-tuned2026.05 | 0.8132 | 0.8131 | |
| HGTTraining ratio=20%2020.12 | 0.81 | 0.8286 | |
| HGTRatio=20%, learning rate=0.008, dropout=0.82020.12 | 0.81 | 0.8286 | |
| MLPTraining ratio=100%2020.12 | 0.7594 | 0.7745 | |
| MLPRatio=100%, learning rate=0.008, dropout=0.92020.12 | 0.7594 | 0.7745 | |
| HetGNNTraining ratio=100%2020.12 | 0.7565 | 0.7721 | |
| HetGNNRatio=100%, learning rate=0.01, dropout=0.82020.12 | 0.7565 | 0.7721 | |
| MLPTraining ratio=80%2020.12 | 0.7503 | 0.7642 | |
| MLPRatio=80%, learning rate=0.01, dropout=0.82020.12 | 0.7503 | 0.7642 | |
| HetGNNTraining ratio=80%2020.12 | 0.7445 | 0.7592 | |
| HetGNNRatio=80%, learning rate=0.01, dropout=0.82020.12 | 0.7445 | 0.7592 | |
| HeCoPre-trained dataset=ACM, Shots=5-shot, Evaluation protocol=fine-tuned2026.05 | 0.744 | 0.7637 | |
| MLPTraining ratio=60%2020.12 | 0.7252 | 0.7354 | |
| MLPRatio=60%, learning rate=0.008, dropout=0.92020.12 | 0.7252 | 0.7354 | |
| HGPromptPre-trained dataset=ACM, Shots=5-shot, Evaluation protocol=fine-tuned2026.05 | 0.7136 | 0.7108 | |
| HetGNNTraining ratio=60%2020.12 | 0.7133 | 0.7248 | |
| HetGNNRatio=60%, learning rate=0.01, dropout=0.82020.12 | 0.7133 | 0.7248 | |
| GraphPromptPre-trained dataset=ACM, Shots=5-shot, Evaluation protocol=fine-tuned2026.05 | 0.7064 | 0.7019 | |
| GraphLoRAPre-trained dataset=ACM, Shots=5-shot, Evaluation protocol=fine-tuned2026.05 | 0.6786 | 0.6947 | |
| MLPTraining ratio=40%2020.12 | 0.6585 | 0.6887 | |
| MLPRatio=40%, learning rate=0.03, dropout=0.82020.12 | 0.6585 | 0.6887 | |
| HetGPTPre-trained dataset=ACM, Shots=5-shot, Evaluation protocol=fine-tuned2026.05 | 0.6556 | 0.7274 | |
| HetGNNTraining ratio=40%2020.12 | 0.6476 | 0.6872 | |
| HetGNNRatio=40%, learning rate=0.01, dropout=0.82020.12 | 0.6476 | 0.6872 | |
| MLPTraining ratio=20%2020.12 | 0.6156 | 0.6469 | |
| MLPRatio=20%, learning rate=0.01, dropout=0.82020.12 | 0.6156 | 0.6469 | |
| HetGNNTraining ratio=20%2020.12 | 0.6022 | 0.642 | |
| HetGNNRatio=20%, learning rate=0.01, dropout=0.82020.12 | 0.6022 | 0.642 | |
| EdgePromptPre-trained dataset=ACM, Shots=5-shot, Evaluation protocol=fine-tuned2026.05 | 0.3372 | 0.468 |