Uni-Parser: Unified Semantic Parser for Question Answering on Knowledge Base and Database

Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 8858 - 8869 December 7-11, 2022 ©2022 Association for Computational Linguistics Uni-Parser: Unified Semantic Parser for Question Answering on Knowledge Base and Database Ye Liu1, Semih Yavuz1, Rui Meng1, Dragomir Radev2, Caiming Xiong1, Yingbo Zhou1 1 Salesforce Research, 2 Yale University {yeliu, syavuz, ruimeng, yingbo.zhou, cxiong}@salesforce.com, dragomir.radev@yale.edu Abstract Parsing natural language questions into exe- cutable logical forms is a useful and inter- pretable way to perform question answering on structured data such as knowledge bases (KB) or databases (DB). However, existing ap- proaches on semantic parsing cannot adapt to both modalities, as they suffer from the expo- nential growth of the logical form candidates and can hardly generalize to unseen data. In this work, we propose Uni-Parser, a unified semantic parser for question answering (QA) on both KB and DB. We introduce the primi- tive (relation and entity in KB, and table name, column name and cell value in DB) as an essen- tial element in our framework. The number of primitives grows linearly with the number of retrieved relations in KB and DB, preventing us from dealing with exponential logic form candi- dates. We leverage the generator to predict final logical forms by altering and composing top- ranked primitives with different operations (e.g. select, where, count). With sufficiently pruned search space by a contrastive primitive ranker, the generator is empowered to capture the com- position of primitives enhancing its generaliza- tion ability. We achieve competitive results on multiple KB and DB QA benchmarks more ef- ficiently, especially in the compositional and zero-shot settings. 1 Introduction With the recent advances in deep neural networks, question answering (QA) systems enable users to interact with massive data using queries in natu- ral language. However, it remains challenging to assess structured data, such as knowledge bases and databases. Semantic parsing is a core step of question answering for structured data. The goal is to convert a natural language question to an exe- cutable logical form (Berant et al., 2013; Yih et al., 2015), e.g., SQL for databases and S-expression for knowledge bases. To improve the accuracy and faithfulness in exe- cution of semantic parsing, recent KBQA studies propose to generate logical form candidates by enu- merating and selecting the best logical form by ranking (Berant and Liang, 2014; Yih et al., 2015; Sun et al., 2020; Ye et al., 2021). However, the number of logical form candidates may grow expo- nentially with the increase of reasoning depth for complex questions. Thus this approach can suffer from poor runtime performance due to the time- consuming logical form enumeration (Gu et al., 2021) and inefficient candidate ranking (Ye et al., 2021). For example, given an entity in a KB, we collect its logical form candidates by enumerating paths up to two hops. If the first hop with respect to the given entity contains N relations and the second hop contains M relations, the enumeration would result in N × M logical forms. …

Content truncated to 500 words