Graph Representation of Synonymy and Translation Resources for Crosslinguistic Modelisation of Meaning

Graph representation of synonymy and translation resources for cross-linguistic modelisation of meaning⋆ Benoˆıt Gaillard a, Yannick Chudy a, Pierre Magistry b,c, Shu-Kai Hsieh c, Emmanuel Navarro d aCLLE-ERSS, University of Toulouse II - Le Mirail, Toulouse, France benoit.gaillard@univ-tlse2.fr, yannick.chudy@gmail.com bAlpage, INRIA Paris-Rocquencourt & Universit´e Paris 7, Paris, France. magistry@gate.sinica.edu.tw cGraduate Institute of Linguistics, National Taiwan University, Taipei, Taiwan shukai@ntu.edu.tw dIRIT, University of Toulouse III, Toulouse, France navarro@irit.fr Abstract In this paper we describe the data that will be used to compare the semantic struc- tures that emerge from synonymy in French and in Mandarin. We aim at studying these semantic structures at both a global, lexicographic level, using lexicons, synonymy and trans- lation dictionaries and at a more localised, experimental level, using data collected in parallel psycholinguistic experiments in French and Mandarin. After presenting our research project, the data we need to carry it out and the available resources, we analyse several linguistic issues arising from the structural differences between the French and Mandarin lexicons. We then explain the construction of the synonymy and translation networks from the available resources and detail specific choices that will enable us to produce meaningful experimental results based on this prepared data. Two kinds of networks are built: lexicographic networks and smaller movie-based networks extracted from experimental recordings. We conclude by describing how we intend to use this data. Keywords: Synonymy, graphs, lexicon, translation, concept similarity 1 Introduction This work describes the resources built to be used in a project that aims to compare, or contrast, synonymy structures across languages. We suppose that we can trace concepts of a community of speakers by detecting salient patterns in the networks constituted by all the words of the language, linked by synonymy. There is a long standing debate on the question whether there exist universal cognitive concepts (Pinker, 1994) whether they necessarily translate into language, or on the con- trary whether the absence of universal concepts in the variety of languages in the world calls for an entirely different account of human cognition (Evans and Levinson, 2009). Evans and Levinson also argue that understanding the diversity of languages is fundamental to a realistic account of human cognition by psycholinguistics. We attempt, in this project, to move one step further than the debate over linguistic universals, and to create a measure of similarity between lexico-semantic concepts across languages. In this approach, the notion of a common concept, and a fortiori of universal concepts is broadened, because, more than asserting whether concepts are common to several languages, we aim to measure their similarity which can cover the whole range of possible values. ⋆ The research presented in this paper is supported by the ANR-NSC (France-Taiwan) bilateral project M3 (Modeling and Measurement of Meaning). Copyright 2010 by Benoˆıt Gaillard, Yannick Chudy, Pierre Magistry, Shu-Kai Hsieh, Emmanuel Navarro PACLIC 24 Proceedings 819 What is a concept emerging from the paradigmatic structure of the lexicon of a language re- mains to be defined. In order to model this , paradigmatic structure …

Content truncated to 500 words