Automatic Acquisition of Meaning Elements for the Creation of Semantic Lexicons

This paper presents, in a unified way, two new trends in natural language processing, that is a new kind of lexicons that are cornerstones of a lot of current natural language applications which tackle the problem of meaning, and different corpus-based lexical knowledge acquisition studies that have emerged with the big amounts of electronic texts available on the nets. More precisely, this paper is an overview of corpus-based methods that could assist a lexicographer in order to build semantic lexicons. The presentation is directed by the lexicographer’s tasks that these studies can facilitate, i.e. term acquisition in order to circumscribe the vocabulary of the lexicon of a domain, structuring of a lexicon with linked semantic classes, adjustment to a language, and determination of the predicates of a domain. We end up by a discussion about the integration of these methods into a fully automatic construction of semantic lexicons.