FBK-irst: Lexical Substitution Task Exploiting Domain and Syntagmatic Coherence

Proceedings of the 4th International Workshop on Semantic Evaluations (SemEval-2007), pages 145–148, Prague, June 2007. c⃝2007 Association for Computational Linguistics FBK-irst: Lexical Substitution Task Exploiting Domain and Syntagmatic Coherence Claudio Giuliano and Alfio Gliozzo and Carlo Strapparava FBK-irst, I-38050, Povo, Trento, ITALY {giuliano, gliozzo, strappa}@itc.it Abstract This paper summarizes FBK-irst participa- tion at the lexical substitution task of the SEMEVAL competition. We submitted two different systems, both exploiting synonym lists extracted from dictionaries. For each word to be substituted, the systems rank the associated synonym list according to a simi- larity metric based on Latent Semantic Anal- ysis and to the occurrences in the Web 1T 5-gram corpus, respectively. In particular, the latter system achieves the state-of-the-art performance, largely surpassing the baseline proposed by the organizers. 1 Introduction The lexical substitution (Glickman et al., 2006a) can be regarded as a subtask of the lexical entailment, in which for a given word in context the system is asked to select an alternative word that can be re- placed in that context preserving the meaning. Lex- ical Entailment, and in particular lexical reference (Glickman et al., 2006b)1, is in turn a subtask of tex- tual entailment, which is formally defined as a rela- tionship between a coherent text T and a language expression, the hypothesis H. T is said to entail H, denoted by T → H, if the meaning of H can be in- ferred from the meaning of T (Dagan et al., 2005; Dagan and Glickman., 2004). Even though this no- tion has been only recently proposed in the computa- tional linguistics literature, it attracts more and more attention due to the high generality of its settings and to the usefulness of its (potential) applications. 1In the literature, slight variations of this problem have been also referred to as sense matching (Dagan et al., 2006). With respect to lexical entailment, the lexical sub- stitution task has a more restrictive criterion. In fact, two words can be substituted when meaning is preserved, while the criterion for lexical entailment is that the meaning of the thesis is implied by the meaning of the hypothesis. The latter condition is in general ensured by substituting either hyperonyms or synonyms, while the former is more rigid because only synonyms are in principle accepted. Formally, in a lexical entailment task a system is asked to decide whether the substitution of a par- ticular term w with the term e in a coherent text Hw = HlwHr generates a sentence He = HleHr such that Hw → He, where Hl and Hr denote the left and the right context of w, respectively. For example, given the source word ‘weapon’ a system may substitute it with the target synonym ‘arm’, in order to identify relevant texts that denote the sought concept using the latter term. A particular case of lexical entailment is recog- nizing synonymy, where both Hw → He and He → Hw hold. The lexical substitution task at SEMEVAL addresses exactly this problem. The task is …

Content truncated to 500 words