Improving Word Sense Disambiguation with Translations

Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing, pages 4055–4065, November 16–20, 2020. c⃝2020 Association for Computational Linguistics 4055 Improving Word Sense Disambiguation with Translations Yixing Luan, Bradley Hauer, Lili Mou, Grzegorz Kondrak Alberta Machine Intelligence Institute, Department of Computing Science University of Alberta, Edmonton, Canada {yixing1,bmhauer,lmou,gkondrak}@ualberta.ca Abstract It has been conjectured that multilingual infor- mation can help monolingual word sense dis- ambiguation (WSD). However, existing WSD systems rarely consider multilingual informa- tion, and no effective method has been pro- posed for improving WSD by generating trans- lations. In this paper, we present a novel approach that improves the performance of a base WSD system using machine transla- tion. Since our approach is language indepen- dent, we perform WSD experiments on several languages. The results demonstrate that our methods can consistently improve the perfor- mance of WSD systems, and obtain state-of- the-art results in both English and multilingual WSD. To facilitate the use of lexical transla- tion information, we also propose BABALIGN, an precise bitext alignment algorithm which is guided by multilingual lexical correspon- dences from BabelNet. 1 Introduction Word sense disambiguation (WSD) is one of the core tasks in natural language processing. Given a predefined sense inventory, a WSD system aims to identify the correct sense of a content word in context. Although WSD is a monolingual task, it has been conjectured that multilingual information could help (Resnik and Yarowsky, 1999; Carpuat, 2009). Attempts have been made to leverage par- allel corpora for sense tagging (Diab and Resnik, 2002), but no effective method for improving WSD with translations has been proposed to date. Much of the history of WSD has been deter- mined by the availability of manually created lexi- cal resources in English, including SemCor (Miller et al., 1994) and WordNet (Miller, 1995). The situation changed with the introduction of Babel- Net (Navigli and Ponzetto, 2012a), a massive mul- tilingual semantic network, created by automati- cally integrating WordNet, Wikipedia, and other Figure 1: An overview of our approach to leverage translations to improve a base WSD system. resources. In particular, BabelNet synsets contain translations in multiple languages for each indi- vidual word sense. Methods have been proposed to use multilingual information in BabelNet for WSD (Navigli and Ponzetto, 2012b; Apidianaki and Gong, 2015), but they do not directly exploit the mapping between senses and translations in multiple languages. While there have been many attempts to apply WSD to machine translation (MT) (Liu et al., 2018; Pu et al., 2018), our goal instead is to harness advances in MT to improve WSD. Rather than develop a new WSD system, we propose a gen- eral method that can make existing and future sys- tems more accurate by leveraging translations. We evaluate our methods with several supervised and knowledge-based WSD systems. Our approach is based on the assumption of abso- lute synonymy between the senses of mutual trans- lations in context (Hauer and Kondrak, 2020). The principal method SOFTCONSTRAINT refines sense 4056 predictions of a given base …

Content truncated to 500 words