Lexical Inference over Multi-Word Predicates: A Distributional Approach

Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics, pages 644–654, Baltimore, Maryland, USA, June 23-25 2014. c⃝2014 Association for Computational Linguistics Lexical Inference over Multi-Word Predicates: A Distributional Approach Omri Abend Shay B. Cohen Mark Steedman School of Informatics, University of Edinburgh, Edinburgh EH8 9AB, United Kingdom {oabend,scohen,steedman}@inf.ed.ac.uk Abstract Representing predicates in terms of their argument distribution is common practice in NLP. Multi-word predicates (MWPs) in this context are often either disregarded or considered as fixed expressions. The lat- ter treatment is unsatisfactory in two ways: (1) identifying MWPs is notoriously diffi- cult, (2) MWPs show varying degrees of compositionality and could benefit from taking into account the identity of their component parts. We propose a novel approach that integrates the distributional representation of multiple sub-sets of the MWP’s words. We assume a latent distri- bution over sub-sets of the MWP, and esti- mate it relative to a downstream prediction task. Focusing on the supervised identi- fication of lexical inference relations, we compare against state-of-the-art baselines that consider a single sub-set of an MWP, obtaining substantial improvements. To our knowledge, this is the first work to address lexical relations between MWPs of varying degrees of compositionality within distributional semantics. 1 Introduction Multi-word expressions (MWEs) constitute a large part of the lexicon and account for much of its growth (Jackendoff, 2002; Seaton and Macaulay, 2002). However, despite their impor- tance, MWEs remain difficult to define and model, and consequently pose serious difficulties for NLP applications (Sag et al., 2001). Multi-word Predi- cates (MWPs; sometimes termed Complex Predi- cates) form an important and much addressed sub- class of MWEs and are the focus of this paper. MWPs are informally defined as multiple words that constitute a single predicate (Alsina et al., 1997). MWPs encompass a wide range of phe- nomena, including causatives, light verbs, phrasal verbs, serial verb constructions and many others, and pose considerable challenges to both linguistic theory and NLP applications (see Section 2). Part of the difficulty in treating them stems from their position on the borderline between syntax and the lexicon. It is therefore often unclear whether they should be treated as fixed expressions, as compo- sitional phrases that reflect the properties of their component parts or as both. This work addresses the modelling of MWPs within the context of distributional semantics (Tur- ney and Pantel, 2010), in which predicates are represented through the distribution of arguments they may take. In order to collect meaningful statistics, the predicate’s lexical unit should be suf- ficiently frequent and semantically unambiguous. MWPs pose a challenge to such models, as na¨ıvely collecting statistics over all instances of highly ambiguous verbs is likely to result in noisy representations. For instance, the verb “take” may appear in MWPs as varied as “take time”, “take effect” and “take to the hills”. This heterogene- ity of “take” is likely to have a negative effect on downstream systems that use its distributional rep- resentation. For instance, while “take” and “ac- cept” are often …

Content truncated to 500 words