Paper: Semantic Parsing for Single-Relation Question Answering

ACL ID P14-2105
Title Semantic Parsing for Single-Relation Question Answering
Venue Annual Meeting of the Association of Computational Linguistics
Session Main Conference
Year 2014
Authors

We develop a semantic parsing framework based on semantic similarity for open do- main question answering (QA). We focus on single-relation questions and decom- pose each question into an entity men- tion and a relation pattern. Using convo- lutional neural network models, we mea- sure the similarity of entity mentions with entities in the knowledge base (KB) and the similarity of relation patterns and re- lations in the KB. We score relational triples in the KB using these measures and select the top scoring relational triple to answer the question. When evaluated on an open-domain QA task, our method achieves higher precision across different recall points compared to the previous ap- proach, and can improve F 1 by 7 points.