Identifying Bad Semantic Neighbors for Improving Distributional Thesauri

Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics, pages 561–571, Sofia, Bulgaria, August 4-9 2013. c⃝2013 Association for Computational Linguistics Identifying Bad Semantic Neighbors for Improving Distributional Thesauri Olivier Ferret CEA, LIST, Vision and Content Engineering Laboratory, Gif-sur-Yvette, F-91191 France. olivier.ferret@cea.fr Abstract Distributional thesauri are now widely used in a large number of Natural Lan- guage Processing tasks. However, they are far from containing only interesting semantic relations. As a consequence, improving such thesaurus is an impor- tant issue that is mainly tackled indirectly through the improvement of semantic sim- ilarity measures. In this article, we pro- pose a more direct approach focusing on the identification of the neighbors of a thesaurus entry that are not semantically linked to this entry. This identification re- lies on a discriminative classifier trained from unsupervised selected examples for building a distributional model of the entry in texts. Its bad neighbors are found by ap- plying this classifier to a representative set of occurrences of each of these neighbors. We evaluate the interest of this method for a large set of English nouns with various frequencies. 1 Introduction The work we present in this article focuses on the automatic building of a thesaurus from a corpus. As illustrated by Table 1, such thesaurus gives for each of its entries a list of words, called seman- tic neighbors, that are supposed to be semanti- cally linked to the entry. Generally, each neigh- bor is associated with a weight that characterizes the strength of its link with the entry and all the neighbors of an entry are sorted according to the decreasing order of their weight. The term semantic neighbor is very generic and can have two main interpretations according to the kind of semantic relations it is based on: one re- lies only on paradigmatic relations, such as hy- pernymy or synonymy, while the other consid- ers syntagmatic relations, called collocation rela- tions by (Halliday and Hasan, 1976) in the context of lexical cohesion or “non-classical relations” by (Morris and Hirst, 2004). The distinction between these two interpretations refers to the distinction between the notions of semantic similarity and se- mantic relatedness as it was done in (Budanitsky and Hirst, 2006) or in (Zesch and Gurevych, 2010) for instance. However, the limit between these two notions is sometimes hard to find in existing work as terms semantic similarity and semantic relat- edness are often used interchangeably. Moreover, semantic similarity is frequently considered as in- cluded into semantic relatedness and the two prob- lems are often tackled by using the same methods. In the remainder of this article, we will use the term semantic similarity with its generic sense and the term semantic relatedness for referring more specifically to similarity based on syntagmatic re- lations. Following work such as (Grefenstette, 1994), a widespread way to build a thesaurus from a cor- pus is to use a semantic similarity measure for ex- tracting the semantic neighbors of the entries of the thesaurus. Three …

Content truncated to 500 words