Browsing the Terminological Structure of a Specialized Domain: A Method Based on Lexical Functions and their Classification

Browsing the Terminological Structure of a Specialized Domain: A Method Based on Lexical Functions and their Classification Marie-Claude L’Homme, Benoît Robichaud, Nathalie Prévil Observatoire de linguistique Sens-Texte (OLST) Université de Montréal C.P. 6128, succ. Centre-ville Montréal (Québec) H3C 3J7 CANADA {mc.lhomme, benoit.robichaud, nathalie.previl}@umontreal.ca Abstract This paper describes a method for browsing relations between terms and unveiling the terminological structure of a specialized domain. The method consists in expanding a graph that takes as input the relations encoded in a multilingual terminological resource called the DiCoEnviro that contains terms in the field of the environment. In the DiCoEnviro, terminological relations are encoded using lexical functions (Melčuk et al. 1995) and further classified in families defined on the basis of the properties of relations. We seek to provide users with an explicit and intuitive representation of a wide variety of relations. We also make the most of the richness of the encoding, while implementing some graphical choices to make their interpretation as clear as possible for end users. The method is implemented in a tool called NeoVisual that provides access to more than 11,000 relations in English and 15,000 relations in French. Portuguese is also included and coverage in all languages will increase as new entries are added to the DiCoEnviro. Keywords: terminological relation; paradigmatic relation; syntagmatic relation; graphical representation; lexical function 1. Introduction In terminology, it is assumed that concepts in specialized domains and terms used to express them are part of a structure. Concepts or terms are defined according to the place they have in this structure and the relations they hold with others. This principle is taken for granted (at least partially) when describing terminological data and some resources represent this structure explicitly (thesauri, terminological knowledge bases, ontologies).1 In this paper we describe a method for browsing relations between terms and gradually unveiling the terminological structure of a specialized domain. The method is based on the contents of a terminological resource called DiCoEnviro that contains terms linked to the domain of the environment. This specific work builds on three sources (that are further presented in Section 2):  Terminological relations manually encoded in the DiCoEnviro with lexical functions (Mel’čuk et al. 1995);  A method devised by Robichaud (2012) developed on the basis of relations encoded in another term base very similar in structure to the DiCoEnviro, i.e. the DiCoInfo (on computing and the Internet);  A classification of terminological relations and natural language explanations superimposed on LFs to facilitate their consultation in a user-friendly version of our resources (L’Homme et al. 2012). The method consists in expanding a graph that takes the relations from the DiCoEnviro as input and allow users to explore different parts of the terminological structure in which they are interested. Our method is targeted at end users of terminological resources. So we seek to provide 1 However, it is worth pointing out that the most widely used terminological resources, i.e. term banks, do not represent terminological structure explicitly. them with an explicit and intuitive …

Content truncated to 500 words