Studying the Recursive Behaviour of Adjectival Modification with Compositional Distributional Semantics

Proceedings of the 2013 Conference on Empirical Methods in Natural Language Processing, pages 141–151, Seattle, Washington, USA, 18-21 October 2013. c⃝2013 Association for Computational Linguistics Studying the recursive behaviour of adjectival modification with compositional distributional semantics Eva Maria Vecchi and Roberto Zamparelli and Marco Baroni Center for Mind/Brain Sciences (University of Trento, Italy) (evamaria.vecchi|roberto.zamparelli|marco.baroni)@unitn.it Abstract In this study, we use compositional distribu- tional semantic methods to investigate restric- tions in adjective ordering. Specifically, we focus on properties distinguishing Adjective- Adjective-Noun phrases in which there is flex- ibility in the adjective ordering from those bound to a rigid order. We explore a number of measures extracted from the distributional representation of AAN phrases which may in- dicate a word order restriction. We find that we are able to distinguish the relevant classes and the correct order based primarily on the degree of modification of the adjectives. Our results offer fresh insight into the semantic properties that determine adjective ordering, building a bridge between syntax and distri- butional semantics. 1 Introduction A prominent approach for representing the meaning of a word in Natural Language Processing (NLP) is to treat it as a numerical vector that codes the pat- tern of co-occurrence of that word with other ex- pressions in a large corpus of language (Sahlgren, 2006; Turney and Pantel, 2010). This approach to semantics (sometimes called distributional seman- tics) scales well to large lexicons and does not re- quire words to be manually disambiguated (Sch¨utze, 1997). Until recently, however, this method had been almost exclusively limited to the level of sin- gle content words (nouns, adjectives, verbs), and had not directly addressed the problem of composition- ality (Frege, 1892; Montague, 1970; Partee, 2004), the crucial property of natural language which al- lows speakers to derive the meaning of a complex linguistic constituent from the meaning of its imme- diate syntactic subconstituents. Several recent proposals have strived to ex- tend distributional semantics with a component that also generates vectors for complex linguistic con- stituents, using compositional operations in the vec- tor space (Baroni and Zamparelli, 2010; Guevara, 2010; Mitchell and Lapata, 2010; Grefenstette and Sadrzadeh, 2011; Socher et al., 2012). All of these approaches construct distributional represen- tations for novel phrases starting from the corpus- derived vectors for their lexical constituents and exploiting the geometric quality of the representa- tion. Such methods are able to capture complex se- mantic information of adjective-noun (AN) phrases, such as characterizing modification (Boleda et al., 2012; Boleda et al., 2013), and can detect seman- tic deviance in novel phrases (Vecchi et al., 2011). Furthermore, these methods are naturally recursive: they can derive a representation not only for, e.g., red car, but also for new red car, fast new red car, etc. This aspect is appealing since trying to extract meaningful representations for all recursive phrases directly from a corpus will result in a problem of sparsity, since most large phrases will never occur in any finite sample. Once we start seriously looking into recursive modification, however, the …

Content truncated to 500 words