Detecting Compositionality of Verb-Object Combinations using Selectional Preferences

Proceedings of the 2007 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning, pp. 369–379, Prague, June 2007. c⃝2007 Association for Computational Linguistics Detecting Compositionality of Verb-Object Combinations using Selectional Preferences Diana McCarthy University of Sussex Falmer, East Sussex BN1 9QH, UK dianam@sussex.ac.uk Sriram Venkatapathy International Institute of Information Technology Hyderabad, India sriram@research.iiit.ac.in Aravind K. Joshi University of Pennsylvania, Philadelphia PA, USA. joshi@linc.cis.upenn.edu Abstract In this paper we explore the use of se- lectional preferences for detecting non- compositional verb-object combinations. To characterise the arguments in a given gram- matical relationship we experiment with three models of selectional preference. Two use WordNet and one uses the entries from a distributional thesaurus as classes for rep- resentation. In previous work on selectional preference acquisition, the classes used for representation are selected according to the coverage of argument tokens rather than be- ing selected according to the coverage of argument types. In our distributional the- saurus models and one of the methods us- ing WordNet we select classes for represent- ing the preferences by virtue of the number of argument types that they cover, and then only tokens under these classes which are representative of the argument head data are used to estimate the probability distribution for the selectional preference model. We demonstrate a highly significant correlation between measures which use these ‘type- based’ selectional preferences and composi- tionality judgements from a data set used in previous research. The type-based models perform better than the models which use to- kens for selecting the classes. Furthermore, the models which use the automatically ac- quired thesaurus entries produced the best results. The correlation for the thesaurus models is stronger than any of the individ- ual features used in previous research on the same dataset. 1 Introduction Characterising the semantic behaviour of phrases in terms of compositionality has particularly attracted attention in recent years (Lin, 1999; Schone and Ju- rafsky, 2001; Bannard, 2002; Bannard et al., 2003; Baldwin et al., 2003; McCarthy et al., 2003; Ban- nard, 2005; Venkatapathy and Joshi, 2005). Typi- cally the phrases are putative multiwords and non- compositionality is viewed as an important feature of many such “words with spaces” (Sag et al., 2002). For applications such as paraphrasing, information extraction and translation, it is essential to take the words of non-compositional phrases together as a unit because the meaning of a phrase cannot be ob- tained straightforwardly from the constituent words. In this work we are investigate methods of deter- mining semantic compositionality of verb-object 1 combinations on a continuum following previous research in this direction (McCarthy et al., 2003; Venkatapathy and Joshi, 2005). Much previous research has used a combination of statistics and distributional approaches whereby distributional similarity is used to compare the con- stituents of the multiword with the multiword itself. In this paper, we will investigate the use of selec- tional preferences of verbs. We will use the pref- erences to find atypical verb-object combinations as we anticipate that such combinations are …

Content truncated to 500 words