Transfer Mixed-hop Prediction on FoodOn (test)
70Precision (Random Negatives)all-MiniLM-L12-v2 + HIT
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| all-MiniLM-L12-v2 + HITTraining=Fine-tuned, Component=Hierarchy Transformer2024.01 | 70 | 50 | 58.3 | 59.4 | 44.2 | 50.6 | |
| all-MiniLM-L12-v2Mode=HIT2024.01 | 69 | 46.3 | 55.4 | 74.1 | 38.5 | 50.7 | |
| all-MiniLM-L12-v2Mode=fine-tune2024.01 | 43.6 | 38.2 | 40.7 | 69 | 17.7 | 28.2 | |
| all-MiniLM-L12-v2 + fine-tuneTraining=Fine-tuned2024.01 | 37.8 | 54 | 44.5 | 63.8 | 37.1 | 46.9 | |
| PoincaréGloVe2024.01 | 19.2 | 22.4 | 20.7 | 18.9 | 20 | 19.5 | |
| all-MiniLM-L12-v2Training=Pre-trained2024.01 | 13.5 | 65.6 | 22.4 | 9.9 | 83.3 | 17.6 | |
| all-MiniLM-L12-v2Mode=Pre-trained2024.01 | 13.5 | 65.6 | 22.4 | 9.9 | 83.3 | 17.6 | |
| NaivePrior2024.01 | 9.1 | 9.1 | 9.1 | 9.1 | 9.1 | 9.1 |