GPLSI-IXA: Using Semantic Classes to Acquire Monosemous Training Examples from Domain Texts

Proceedings of the 5th International Workshop on Semantic Evaluation, ACL 2010, pages 402–406, Uppsala, Sweden, 15-16 July 2010. c⃝2010 Association for Computational Linguistics GPLSI-IXA: Using Semantic Classes to Acquire Monosemous Training Examples from Domain Texts Rub´en Izquierdo & Armando Su´arez GPLSI Group University of Alicante. Spain {ruben,armando}@dlsi.ua.es German Rigau IXA NLP Group. EHU. Donostia, Spain german.rigau@ehu.es Abstract This paper summarizes our participation in task #17 of SemEval–2 (All–words WSD on a specific domain) using a su- pervised class-based Word Sense Disam- biguation system. Basically, we use Sup- port Vector Machines (SVM) as learning algorithm and a set of simple features to build three different models. Each model considers a different training corpus: Sem- Cor (SC), examples from monosemous words extracted automatically from back- ground data (BG), and both SC and BG (SCBG). Our system explodes the monosemous words appearing as mem- bers of a particular WordNet semantic class to automatically acquire class-based annotated examples from the domain text. We use the class-based examples gathered from the domain corpus to adapt our tra- ditional system trained on SemCor. The evaluation reveal that the best results are achieved training with SemCor and the background examples from monosemous words, obtaining results above the first sense baseline and the fifth best position in the competition rank. 1 Introduction As empirically demonstrated by the last SensEval and SemEval exercises, assigning the appropriate meaning to words in context has resisted all at- tempts to be successfully addressed. In fact, super- vised word-based WSD systems are very depen- dent of the corpora used for training and testing the system (Escudero et al., 2000). One possible reason could be the use of inappropriate level of abstraction. Most supervised systems simply model each polysemous word as a classification problem where each class corresponds to a particular synset of the word. But, WordNet (WN) has been widely criticized for being a sense repository that often provides too fine–grained sense distinctions for higher level applications like Machine Translation or Question & Answering. In fact, WSD at this level of granularity has resisted all attempts of in- ferring robust broad-coverage models. It seems that many word–sense distinctions are too subtle to be captured by automatic systems with the cur- rent small volumes of word–sense annotated ex- amples. Thus, some research has been focused on deriv- ing different word-sense groupings to overcome the fine–grained distinctions of WN (Hearst and Sch¨utze, 1993), (Peters et al., 1998), (Mihalcea and Moldovan, 2001), (Agirre and LopezDeLa- Calle, 2003), (Navigli, 2006) and (Snow et al., 2007). That is, they provide methods for grouping senses of the same word, thus producing coarser word sense groupings for better disambiguation. In contrast, some research have been focused on using predefined sets of sense-groupings for learn- ing class-based classifiers for WSD (Segond et al., 1997), (Ciaramita and Johnson, 2003), (Villarejo et al., 2005), (Curran, 2005), (Kohomban and Lee, 2005) and (Ciaramita and Altun, 2006). That is, grouping senses of different words into the same explicit and comprehensive semantic class. Most of the …

Content truncated to 500 words