METRICC: Harnessing Comparable Corpora for Multilingual Lexicon Development
Résumé
Research on comparable corpora has grown in recent years bringing about the possibility of developing multilingual lexicons through the exploitation of comparable corpora to create corpus-driven multilingual dictionaries. To date, this issue has not been widely addressed. This paper focuses on the use of the mechanism of collocational networks proposed by Williams (1998) for exploiting comparable corpora. The paper first provides a description of the METRICC project, which is aimed at the automatically creation of comparable corpora and describes one of the crawlers developed for comparable corpora building, and then discusses the power of collocational networks for multilingual corpus-driven dictionary development.
Fichier principal
EURALEX_pp389-403_Alonso_Blancafort_De_Groc_Millon_and_Williams.pdf (408.96 Ko)
Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...