Literal and Metaphorical Senses in Compositional Distributional Semantic Models

Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics, pages 183–193, Berlin, Germany, August 7-12, 2016. c⃝2016 Association for Computational Linguistics Literal and Metaphorical Senses in Compositional Distributional Semantic Models E. Dar´ıo Guti´errez1 Ekaterina Shutova2 Tyler Marghetis3 Benjamin K. Bergen4 1 University of California San Diego 2 University of Cambridge 3 Indiana University Bloomington edg@icsi.berkeley.edu tmarghet@cogsci.ucsd.edu es407@cam.ac.uk bkbergen@ucsd.edu Abstract Metaphorical expressions are pervasive in natural language and pose a substan- tial challenge for computational seman- tics. The inherent compositionality of metaphor makes it an important test case for compositional distributional semantic models (CDSMs). This paper is the first to investigate whether metaphorical compo- sition warrants a distinct treatment in the CDSM framework. We propose a method to learn metaphors as linear transforma- tions in a vector space and find that, across a variety of semantic domains, explicitly modeling metaphor improves the result- ing semantic representations. We then use these representations in a metaphor iden- tification task, achieving a high perfor- mance of 0.82 in terms of F-score. 1 Introduction An extensive body of behavioral and corpus- linguistic studies suggests that metaphors are per- vasive in everyday language (Cameron, 2003; Steen et al., 2010) and play an important role in how humans define and understand the world. Ac- cording to Conceptual Metaphor Theory (CMT) (Lakoff and Johnson, 1981), individual metaphor- ical expressions, or linguistic metaphors (LMs), are instantiations of broader generalizations re- ferred to as conceptual metaphors (CMs). For example, the phrases half-baked idea, food for thought, and spoon-fed information are LMs that instantiate the CM IDEAS ARE FOOD. These phrases reflect a mapping from the source domain of FOOD to the target domain of IDEAS (Lakoff, 1989). Two central claims of the CMT are that this mapping is systematic, in the sense that it con- sists of a fixed set of ontological correspondences, such as thinking is preparing, communication is feeding, understanding is digestion; and that this mapping can be productively extended to produce novel LMs that obey these correspondences. Recent years have seen the rise of statistical techniques for metaphor detection. Several of these techniques leverage distributional statistics and vector-space models of meaning to classify ut- terances as literal or metaphorical (Utsumi, 2006; Shutova et al., 2010; Hovy et al., 2013; Tsvetkov et al., 2014). An important insight of these studies is that metaphorical meaning is not merely a prop- erty of individual words, but rather arises through cross-domain composition. The meaning of sweet, for instance, is not intrinsically metaphorical. Yet this word may exhibit a range of metaphori- cal meanings—e.g., sweet dreams, sweet person, sweet victory–that are created through the inter- play of source and target domains. If metaphor is compositional, how do we represent it, and how can we use it in a compositional framework for meaning? Compositional distributional semantic models (CDSMs) provide a compact model of composi- tionality that produces vector representations of phrases while avoiding the sparsity and storage issues associated with storing vectors for each phrase in a language explicitly. One …

Content truncated to 500 words