Two graph-based algorithms for state-of-the-art WSD

Proceedings of the 2006 Conference on Empirical Methods in Natural Language Processing (EMNLP 2006), pages 585–593, Sydney, July 2006. c⃝2006 Association for Computational Linguistics Two graph-based algorithms for state-of-the-art WSD Eneko Agirre, David Mart´ınez, Oier L´opez de Lacalle and Aitor Soroa IXA NLP Group University of the Basque Country Donostia, Basque Contry a.soroa@si.ehu.es Abstract This paper explores the use of two graph algorithms for unsupervised induction and tagging of nominal word senses based on corpora. Our main contribution is the op- timization of the free parameters of those algorithms and its evaluation against pub- licly available gold standards. We present a thorough evaluation comprising super- vised and unsupervised modes, and both lexical-sample and all-words tasks. The results show that, in spite of the infor- mation loss inherent to mapping the in- duced senses to the gold-standard, the optimization of parameters based on a small sample of nouns carries over to all nouns, performing close to supervised sys- tems in the lexical sample task and yield- ing the second-best WSD systems for the Senseval-3 all-words task. 1 Introduction Word sense disambiguation (WSD) is a key enabling-technology. Supervised WSD tech- niques are the best performing in public evalu- ations, but need large amounts of hand-tagged data. Existing hand-annotated corpora like Sem- Cor (Miller et al., 1993), which is annotated with WordNet senses (Fellbaum, 1998) allow for a small improvement over the simple most frequent sense heuristic, as attested in the all-words track of the last Senseval competition (Snyder and Palmer, 2004). In theory, larger amounts of training data (SemCor has approx. 700K words) would improve the performance of supervised WSD, but no cur- rent project exists to provide such an expensive re- source. Supervised WSD is based on the “fixed-list of senses” paradigm, where the senses for a tar- get word are a closed list coming from a dic- tionary or lexicon. Lexicographers and seman- ticists have long warned about the problems of such an approach, where senses are listed sepa- rately as discrete entities, and have argued in fa- vor of more complex representations, where, for instance, senses are dense regions in a contin- uum (Cruse, 2000). Unsupervised WSD has followed this line of thinking, and tries to induce word senses directly from the corpus. Typical unsupervised WSD sys- tems involve clustering techniques, which group together similar examples. Given a set of induced clusters (which represent word uses or senses1), each new occurrence of the target word will be compared to the clusters and the most similar clus- ter will be selected as its sense. Most of the unsupervised WSD work has been based on the vector space model, where each example is represented by a vector of features (e.g. the words occurring in the context), and the induced senses are either clusters of ex- amples (Sch¨utze, 1998; Purandare and Peder- sen, 2004) or clusters of words (Pantel and Lin, 2002). Recently, V´eronis (V´eronis, 2004) has pro- posed HyperLex, an application of graph models to WSD based on the small-world properties …

Content truncated to 500 words