Prerequisite Chain Learning on NLP to CV domain
0.8172F1 ScoreCLS + Node2vec
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| CLS + Node2vecBase Embedding Model=Node2vec, Learning Protocol=Supervised / Upper Bound2021.05 | 0.8172 | 0.8197 | 0.8223 | 0.814 | |
| CD-VGAE + P2VBase Embedding Model=P2V, Learning Protocol=Zero-shot / Proposed2021.05 | 0.6754 | 0.5759 | 0.5468 | 0.8837 | |
| CD-VGAE + BERTBase Embedding Model=BERT, Learning Protocol=Zero-shot / Proposed2021.05 | 0.6391 | 0.5593 | 0.5441 | 0.7884 | |
| VGAE + P2VBase Embedding Model=P2V, Learning Protocol=Zero-shot / Baseline2021.05 | 0.6202 | 0.55 | 0.5368 | 0.7349 | |
| VGAE + BERTBase Embedding Model=BERT, Learning Protocol=Zero-shot / Baseline2021.05 | 0.5885 | 0.5477 | 0.5398 | 0.6488 | |
| CLS + P2VBase Embedding Model=P2V, Learning Protocol=Zero-shot / Baseline2021.05 | 0.4881 | 0.5757 | 0.6106 | 0.407 | |
| CLS + BERTBase Embedding Model=BERT, Learning Protocol=Zero-shot / Baseline2021.05 | 0.4277 | 0.548 | 0.5743 | 0.3419 |