Classement des combinaisons lexicales spécialisées à base nominale dans un dictionnaire d’informatique

An increasing number of lexicographical and terminological resources present information about lexical combinations, and more specifically collocations. However, the access to this information is not always optimal. The question addressed here is how to organize collocations to help users locate them efficiently: for example, during the production of a text, a translator or a technical writer might already know the meaning of a collocation, but looks up a dictionary to find the form that expresses this meaning. This paper proposes a model for organizing specialized lexical combinations (L’Homme 1998)—“collocations” used in texts for special purposes—in an online specialized dictionary of computing and the Internet. Firstly, the notion of “collocation” in lexicology and in terminology is circumscribed. Secondly, the processing of collocations in selected lexicographical and terminological resources is reviewed, the focus being placed on the methods for organizing collocations and problems raised by them. Thirdly, a model of classification of collocations is devised based on an analysis of a preliminary corpus in order to observe the regularities between the key words and collocates. In this model, collocations are organized in “generic” and “intermediate” semantic classes that are defined according to the observed regularities. Finally, the proposed model is validated by means of an additional corpus.