A Split-and-Recombine Approach for Follow-up Query Analysis

Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing, pages 5316–5326, Hong Kong, China, November 3–7, 2019. c⃝2019 Association for Computational Linguistics 5316 A Split-and-Recombine Approach for Follow-up Query Analysis Qian Liu†∗, Bei Chen§, Haoyan Liu♦∗, Jian-Guang Lou§, Lei Fang§, Bin Zhou†, Dongmei Zhang§ †State Key Laboratory of Virtual Reality Technology and Systems, School of Computer Science and Engineering, Beihang University, China ♦State Key Lab of Software Development Environment, Beihang University, China §Microsoft Research, Beijing, China †♦{qian.liu, haoyan.liu, zhoubin}@buaa.edu.cn; §{beichen, jlou, leifa, dongmeiz}@microsoft.com Abstract Context-dependent semantic parsing has proven to be an important yet challenging task. To leverage the advances in context- independent semantic parsing, we propose to perform follow-up query analysis, aiming to restate context-dependent natural language queries with contextual information. To accomplish the task, we propose STAR, a novel approach with a well-designed two-phase process. It is parser-independent and able to handle multifarious follow-up scenarios in different domains. Experiments on the FollowUp dataset show that STAR outperforms the state-of-the-art baseline by a large margin of nearly 8%. The superiority on parsing results verifies the feasibility of follow-up query analysis. We also explore the extensibility of STAR on the SQA dataset, which is very promising. 1 Introduction Recently, Natural Language Interfaces to Data- bases (NLIDB) has received considerable atten- tion, as they allow users to query databases by directly using natural language. Current stud- ies mainly focus on context-independent seman- tic parsing, which translates a single natural lan- guage sentence into its corresponding executable form (e.g. Structured Query Language) and re- trieves the answer from databases regardless of its context. However, context does matter in real world applications. Users tend to issue queries in a coherent way when communicating with NLIDB. For example, after the query “How much money has Smith earned?” (Precedent Query), users may pose another query by simply asking “How about Bill Collins?” (Follow-up Query) in- stead of the complete “How much money has Bill ∗Work done during an internship at Microsoft Research. Collins earned?” (Restated Query). Therefore, contextual information is essential for more accu- rate and robust semantic parsing, namely context- dependent semantic parsing. Compared with context-independent semantic parsing, context-dependent semantic parsing has received less attention. Several attempts include a statistical model with parser trees (Miller et al., 1996), a linear model with context-dependent log- ical forms (Zettlemoyer and Collins, 2009) and a sequence-to-sequence model (Suhr et al., 2018). However, all these methods cannot apply to differ- ent domains, since the ATIS dataset (Dahl et al., 1994) they rely on is domain-specific. A search- based neural method DynSP∗ arises along with the SequentialQA (SQA) dataset (Iyyer et al., 2017), which takes the first step towards cross-domain context-dependent semantic parsing. Neverthe- less, DynSP∗ focuses on dealing with relatively simple scenarios. All the aforementioned methods design context-dependent semantic parser from scratch. Instead, inspired by Liu et al. (2019), we propose to directly leverage the technical ad- vances in context-independent semantic parsing. We define follow-up …

Content truncated to 500 words