Lei Yu, Karl Moritz Hermann, Phil Blunsom, Stephen Pulman
Answer sentence selection is the task of identifying sentences that contain the answer to a given question. This is an important problem in its own right as well as in the larger context of open domain question answering. We propose a novel approach to solving this task via means of distributed representations, and learn to match questions with answers by considering their semantic encoding. This contrasts prior work on this task, which typically relies on classifiers with large numbers of hand-crafted syntactic and semantic features and various external resources. Our approach does not require any feature engineering nor does it involve specialist linguistic data, making this model easily applicable to a wide range of domains and languages. Experimental results on a standard benchmark dataset from TREC demonstrate that---despite its simplicity---our model matches state of the art performance on the answer sentence selection task.
| Task | Dataset | Metric | Value | Model |
|---|---|---|---|---|
| Question Answering | QASent | MAP | 0.7113 | Bigram-CNN (lexical overlap + dist output) |
| Question Answering | QASent | MRR | 0.7846 | Bigram-CNN (lexical overlap + dist output) |
| Question Answering | QASent | MAP | 0.5693 | Bigram-CNN |
| Question Answering | QASent | MRR | 0.6613 | Bigram-CNN |
| Question Answering | TrecQA | MAP | 0.711 | CNN |
| Question Answering | TrecQA | MRR | 0.785 | CNN |
| Question Answering | WikiQA | MAP | 0.652 | Bigram-CNN (lexical overlap + dist output) |
| Question Answering | WikiQA | MRR | 0.6652 | Bigram-CNN (lexical overlap + dist output) |
| Question Answering | WikiQA | MAP | 0.619 | Bigram-CNN |
| Question Answering | WikiQA | MRR | 0.6281 | Bigram-CNN |