(Designing and Evaluating Extraction Tools for Collocations in Dictionaries and Corpora)

Partners: Université de Liège (Coordinator), Rank Xerox Research Centre (Grenoble, France).

Framework: Funded partly by the Commission of the European Community, Luxemburg, DG XIII E4, under the Multilingual Action Plan (MLAP).

Description of the project: The project aims at designing, implementing and using a computational toolbox for the acquisition of collocations from text corpora and dictionaries. It builds on the elements of a definition of collocational phenomena from lexicography and computational and corpus linguistics, makes a comparative state of the art in tools for lexical acquisition from text and from machine readable dictionaries, then specifies, implements and uses its own toolbox for that purpose. The tools include linguistically informed dictionary and corpus query, as well as linguistic and statistical processing of monolingual and parallel corpora. The project produces parallel fragments of collocational dictionaries for English, French and German, for a given lexical semantic field, on the basis of an integration of information from many different sources.

Duration: 2 years (01/94 -- 12/95). IMS has 30p/m. Total: 120p/m

IMS Contact Point: Bruno Maximilian Schulze (

IMS Stuttgart / WWW@IMS.Uni-Stuttgart.DE / Mon Feb 23 14:52:28 1998 (hofmanaa)