The ISLE in the ocean. Transatlantic standards for multilingual lexicons (with an eye to machine translation)

The ISLE in the Ocean Transatlantic Standards for Multilingual Lexicons (with an Eye to Machine Translation) Nicoletta Calzolari, Alessandro Lenci, Antonio Zampolli,1 Nuria Bel, Marta Villegas,2 Gregor Thurmair3 1Istituto di Linguistica Computazionale, CNR, Pisa Consorzio Pisa Ricerche Università di Pisa, Dipartimento di Linguistica Italy {glottolo,lenci,eagles}@ilc.pi.cnr.it 2GILCUB (Grup Investigació Lingüística Computacional Universitat Barcelona Spain {nuria,tona}@gilcub.es 3Sail Labs, Munich Germany Gregor.Thurmair@sail-labs.de Abstract The ISLE project is a continuation of the long standing EAGLES initiative, carried out under the Human Language Technology (HLT) programme in collaboration between American and European groups in the framework of the EU-US International Research Co-operation, supported by NSF and EC. In this paper we concentrate on the current position of the ISLE Computational Lexicon Working Group (CLWG), whose activities aim at defining a general schema for a multilingual lexical entry (MILE), as the basis for a standard framework for multilingual computational lexicons. The needs and features of existing Machine Translation systems provide the main reference points for the process of consensual definition of the MILE. The overall structure of the MILE will be illustrated with particular attention to some of the issues raised for multilingual lexicons by the need of expressing complex transfer conditions among translation equivalents Keywords standards, multilingual resources, computational lexicons for MT 1. Introduction the One of the crucial aspects for HLT is how to optimise production, maintenance and extension of computational lexical resources, as well as the process leading to their integration in applications. An essential precondition to achieve these results is to establish a common and standardized framework for computational lexicon construction, which may ensure the encoding of linguistic information in such a way to grant its reusability by different applications and in different tasks. This is even more true when multilingual lexicons and machine translation (MT) are taken into consideration. Here two specific problems arise, which respectively concern architectural and representational issues: (i.) how to build new bilingual (multilingual) lexicons from available monolingual resources, and how to establish the proper relation among these two types of architectures; (ii.) how to state in the most proper way the translation correspondences among entries in the multilingual lexicon. With respect to the latter problem, the passage from source language (SL) to target language (TL) makes it necessary to express very complex and articulated transfer conditions, which have to take into account as difficult and pervasive phenomena as argument switching, multi-word expressions, collocational patterns, etc. In turn, the representational issues are crucially connected to the architectural ones, mainly depending on how linguistic information is organized in the monolingual parts, and how it can be accessed at the multilingual layer. The ongoing work of the Computational Lexicon Working Group (CLWG) in the ISLE project, which we illustrate in the following sections, pursues the main goal of establishing a general and consensual standardized environment for the development and integration of multilingual resources, so as to provide a satisfactory answers to the issues above. The general vision of the project adheres to the idea of enhancing the sharing and …

Content truncated to 500 words