UTD-HLT-CG: Semantic Architecture for Metonymy Resolution and Classification of Nominal Relations

Proceedings of the 4th International Workshop on Semantic Evaluations (SemEval-2007), pages 454–459, Prague, June 2007. c⃝2007 Association for Computational Linguistics UTD-HLT-CG: Semantic Architecture for Metonymy Resolution and Classification of Nominal Relations Cristina Nicolae, Gabriel Nicolae and Sanda Harabagiu Human Language Technology Research Institute The University of Texas at Dallas Richardson, Texas {cristina, gabriel, sanda}@hlt.utdallas.edu Abstract In this paper we present a semantic archi- tecture that was employed for processing two different SemEval 2007 tasks: Task 4 (Classification of Semantic Relations be- tween Nominals) and Task 8 (Metonymy Resolution). The architecture uses multi- ple forms of syntactic, lexical, and semantic information to inform a classification-based approach that generates a different model for each machine learning algorithm that imple- ments the classification. We used decision trees, decision rules, logistic regression and lazy classifiers. A voting module selects the best performing module for each task evalu- ated in SemEval 2007. The paper details the results obtained when using the semantic ar- chitecture. 1 Introduction Automatic semantic interpretations of natural lan- guage text rely on (1) semantic theories that cap- ture the subtleties employed by human communi- cations; (2) lexico-semantic resources that encode various forms of semantic knowledge; and (3) com- putational methods that model the selection of the optimal interpretation derived from the textual data. Two of the SemEval 2007 tasks, namely Task 4 (Classification of Semantic Relations between Nom- inals)and Task 8 (Metonymy Resolution) employed distinct theories for the interpretation of their cor- responding semantic phenomena, but, nevertheless, they also shared several lexico-semantic resources, and, furthermore, both these tasks could have been cast as classification problems, in vein with most of the recent work in computational semantic process- ing. Based on this observation, we have designed and implemented a semantic architecture that was used in both tasks. In Section 2 of this paper we give a brief description of the semantic theories cor- responding to each of the two tasks, while in Section 3 we detail the semantic architecture. Section 4 de- scribes the experimental results and evaluation. We have used three lexico-semantic resources: (i) the WordNet lexico-semantic database; (ii) VerbNet; and (iii) the Lexical Conceptual Structure (LCS) database. Used only by Task 4, WordNet is a lexico- semantic database created at Princeton University1 (Fellbaum, 1998), which encodes a vast majority of the English nouns, verbs, adjectives and adverbs, and groups synonym words into synsets. VerbNet2 is a broad-coverage, comprehensive verb lexicon created at University of Pennsylvania, compatible with WordNet, but with explicitly stated syntactic and semantic information, using Levin verb classes (Levin, 1993) to systematically construct lexical en- tities. Classes are hierarchically organized and each class in the hierarchy has its corresponding syntac- tic frames, semantic predicates and a list of typical verb arguments. The Lexical Conceptual Structure (Traum and Habash, 2000) is a compositional ab- straction with language-independent properties. An LCS is a directed graph with a root. Each node is as- sociated with certain information, including a type, a primitive and a field. An LCS captures the semantics 1http://wordnet.princeton.edu …

Content truncated to 500 words