Research Article
Few-Shot Multihop Question Answering over Knowledge Base
Table 4
Comparative results between our best model with other models.
| Method | Negative sample | Avg |
| Wang (baseline) | 3 | 56.70 | Wang (bridging) | 3 | 58.60 | Wang (bridging+literal match) | 3 | 61.50 | Wang (bridging+literal match) | 1 | 61.10 | Wang (bridging+literal match) | 5 | 59.40 | Our model (with 10% data) | | 58.54 | Our model (with 100% data) | | 62.55 |
|
|