Word Sense Disambiguation Incorporating Lexical and Structural Semantic Information

Proceedings of the 2007 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning, pp. 477–485, Prague, June 2007. c⃝2007 Association for Computational Linguistics Word Sense Disambiguation Incorporating Lexical and Structural Semantic Information Takaaki Tanaka† Francis Bond♦ Timothy Baldwin♠ Sanae Fujita† Chikara Hashimoto♣ † {takaaki, sanae}@cslab.kecl.ntt.co.jp ♦ bond@nict.go.jp ♠ tim@csse.unimelb.edu.au ♣ ch@yz.yamagata-u.ac.jp † NTT Communication Science Laboratories, Nippon Telegraph and Telephone Corporation ♦ National Institute of Information and Communications Technology ♠ The University of Melbourne ♣ Yamagata University Abstract We present results that show that incorporat- ing lexical and structural semantic informa- tion is effective for word sense disambigua- tion. We evaluated the method by using pre- cise information from a large treebank and an ontology automatically created from dic- tionary sentences. Exploiting rich semantic and structural information improves preci- sion 2–3%. The most gains are seen with verbs, with an improvement of 5.7% over a model using only bag of words and n-gram features. 1 Introduction Recently, significant improvements have been made in combining symbolic and statistical approaches to various natural language processing tasks. In parsing, for example, symbolic grammars are be- ing combined with stochastic models (Riezler et al., 2002; Oepen et al., 2002; Malouf and van Noord, 2004). Statistical techniques have also been shown to be useful for word sense disambiguation (Steven- son, 2003). However, to date, there have been few combinations of sense information together with symbolic grammars and statistical models. Klein and Manning (2003) show that much of the gain in statistical parsing using lexicalized models comes from the use of a small set of function words. Features based on general relations provide little improvement, presumably because the data is too sparse: in the Penn treebank normally used to train and test statistical parsers stocks and skyrocket never appear together. They note that this should motivate the use of similarity and/or class based approaches: the superordinate concepts capital (⊃ stocks) and move upward (⊃ sky rocket) frequently appear to- gether. However, there has been little success in this area to date. For example, Xiong et al. (2005) use se- mantic knowledge to parse Chinese, but gain only a marginal improvement. Focusing on WSD, Steven- son (2003) and others have shown that the use of syntactic information (predicate-argument relations) improve the quality of word sense disambiguation (WSD). McCarthy and Carroll (2003) have shown the effectiveness of the selectional preference infor- mation for WSD. However, there is still little work on combining WSD and parse selection. We hypothesize that one of the reasons for the lack of success is that there has been no resource annotated with both syntactic (or structural seman- tic information) and lexical semantic information. For English, there is the SemCor corpus (Fellbaum, 1998) which is annotated with parse trees and Word- Net senses, but it is fairly small, and does not ex- plicitly include any structural semantic information. Therefore, we decided to construct and use a tree- bank with both syntactic information (e.g. HPSG parses) and lexical semantic information …

Content truncated to 500 words