Conversational Retrieval on TopiOCQA
73.07Recall@5Dragon + Rewrite
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Dragon + RewriteBackbone=Dragon, Query processing=Rewrite, Training protocol=Unsupervised2024.01 | 73.07 | 88.19 | |
| Adlakha et al. (2022)2024.01 | 70.4 | — | |
| Dragon + Fine-tuneBackbone=Dragon, Query processing=History concatenation, Training protocol=Fine-tune, Training dataset=HumanAnnotatedConvQA2024.01 | 67.86 | 86.28 | |
| Dragon + Fine-tune (SyntheticConvQA)Backbone=Dragon, Query processing=History concatenation, Training protocol=Fine-tune, Training dataset=SyntheticConvQA2024.01 | 64.48 | 85.24 | |
| DragonBackbone=Dragon, Training protocol=Unsupervised2024.01 | 57.68 | 78.8 | |
| E5-unsupervised + Fine-tuneBackbone=E5, Query processing=History concatenation, Training protocol=Fine-tune, Training dataset=HumanAnnotatedConvQA2024.01 | 41.01 | 51.07 | |
| E5-unsupervised + RewriteBackbone=E5, Query processing=Rewrite, Training protocol=Unsupervised2024.01 | 27.58 | 39.15 | |
| E5-unsupervisedBackbone=E5, Training protocol=Unsupervised2024.01 | 26.25 | 37.67 |