Improving Semantic Parsing with Enriched Synchronous Context-Free Grammar

Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing, pages 1455–1465, Lisbon, Portugal, 17-21 September 2015. c⃝2015 Association for Computational Linguistics. Improving Semantic Parsing with Enriched Synchronous Context-Free Grammar Junhui Li1 Muhua Zhu2 Wei Lu3 Guodong Zhou1 1Natural Language Processing Lab, Soochow University, China {lijunhui, gdzhou}@suda.edu.cn 2 Alibaba Inc., Hangzhou, China muhua.zmh@alibaba-inc.com 3 Information Systems Technology and Design, Singapore University of Technology and Design luwei@sutd.edu.sg Abstract Semantic parsing maps a sentence in natu- ral language into a structured meaning rep- resentation. Previous studies show that se- mantic parsing with synchronous context- free grammars (SCFGs) achieves favor- able performance over most other alter- natives. Motivated by the observation that the performance of semantic pars- ing with SCFGs is closely tied to the translation rules, this paper explores ex- tending translation rules with high qual- ity and increased coverage in three ways. First, we introduce structure informed non-terminals, better guiding the parsing in favor of well formed structure, instead of using a uninformed non-terminal in SCFGs. Second, we examine the differ- ence between word alignments for seman- tic parsing and statistical machine transla- tion (SMT) to better adapt word alignment in SMT to semantic parsing. Finally, we address the unknown word translation is- sue via synthetic translation rules. Eval- uation on the standard GeoQuery bench- mark dataset shows that our approach achieves the state-of-the-art across various languages, including English, German and Greek. 1 Introduction Semantic parsing, the task of mapping natural language (NL) sentences into a formal meaning representation language (MRL), has recently re- ceived a significant amount of attention with vari- ous models proposed over the past few years. Con- sider the NL sentence paired with its correspond- ing MRL in Figure 1(a). Semantic parsing can be NL: What is the area of Sea0le MRL: answer(area_1(cityid(‘sea0le’, _))) NL’: what be the area of sea0le MRL’: answer@1 area_1@1 cityid@2 sea0le@s _@0 (a) before pre-­‐processing (b) aGer pre-­‐processing Figure 1: Example of a sentence pair in NL and MRL. naturally viewed as a statistical machine transla- tion (SMT) task, which translates a sentence in NL (i.e., the source language in SMT) into its mean- ing representation in MRL (i.e., the target lan- guage in SMT). Indeed, many attempts have been made to directly apply statistical machine transla- tion (SMT) systems (or methodologies) to seman- tic parsing (Papineni et al., 1997; Macherey et al., 2001; Wong and Mooney, 2006; Andreas et al., 2013). However, although recent studies (Wong and Mooney, 2006; Andreas et al., 2013) show that semantic parsing with SCFGs, which form the ba- sis of most existing statistical syntax-based trans- lation models (Yamada and Knight, 2001; Chiang, 2007), achieves favorable results, this approach is still behind the most recent state-of-the-art. For details, please see performance comparison in An- dreas et al. (2013) and Lu (2014). The key issues behind the limited success of ap- plying SMT systems directly to semantic parsing lie in the difference between semantic parsing and SMT: MRL is not a real natural language with different …

Content truncated to 500 words