Answering while Summarizing: Multi-task Learning for Multi-hop QA with Evidence Extraction
About
Question answering (QA) using textual sources for purposes such as reading comprehension (RC) has attracted much attention. This study focuses on the task of explainable multi-hop QA, which requires the system to return the answer with evidence sentences by reasoning and gathering disjoint pieces of the reference texts. It proposes the Query Focused Extractor (QFE) model for evidence extraction and uses multi-task learning with the QA model. QFE is inspired by extractive summarization models; compared with the existing method, which extracts each evidence sentence independently, it sequentially extracts evidence sentences by using an RNN with an attention mechanism on the question sentence. It enables QFE to consider the dependency among the evidence sentences and cover important information in the question sentence. Experimental results show that QFE with a simple RC baseline model achieves a state-of-the-art evidence extraction score on HotpotQA. Although designed for RC, it also achieves a state-of-the-art evidence extraction score on FEVER, which is a recognizing textual entailment task on a large textual database.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Multi-hop Question Answering | HotpotQA fullwiki setting (test) | Answer F138.1 | 64 | |
| Question Answering | HotpotQA distractor (dev) | Answer F168.7 | 45 | |
| Question Answering | HotpotQA distractor setting (test) | Answer F168.1 | 34 | |
| Supporting Fact Prediction | HotpotQA distractor (dev) | F1 Score84.7 | 13 | |
| Question Answering | HotpotQA Full Wiki hidden (test) | F138.1 | 12 | |
| Supporting Facts Prediction | HotpotQA Full Wiki hidden (test) | F1 Score44.4 | 11 | |
| Fact Extraction and Verification | FEVER leaderboard March 2019 (test) | Evidence F177.7 | 8 |