Papers with answer selection models
Intent-calibrated Self-training for Answer Selection in Open-domain Dialogues (2023.tacl-1)
Copied to clipboard
| Challenge: | Existing answer selection models require large amounts of labeled data to produce accurate answers. |
| Approach: | They propose intent-calibrated self-training to calibrate answer labels using labeled data . they propose intentcalibration to improve quality of pseudo answer labels . |
| Outcome: | The proposed intent-calibrated answer selection paradigm outperforms baselines with 1%, 5%, and 10% labeled data on two benchmark datasets. |
Knowledge as A Bridge: Improving Cross-domain Answer Selection with External Knowledge (C18-1)
Copied to clipboard
| Challenge: | Existing approaches to answer selection are limited in domains with limited labeled data. |
| Approach: | They propose a Knowledge-aware Attentive Network framework for cross-domain answer selection that uses the knowledge base as a bridge to enable knowledge transfer from the source domain to the target domain. |
| Outcome: | The proposed model outperforms strong competitors by a noticeable margin in cross-domain answer selection. |