Using parallel and comparable corpora for multilingual linguistic resources extraction

Bilingual lexicons are central components of machine translation and cross-lingual information retrieval systems. Their manual construction requires extensive expertise in both languages involved and it is a costly process. Several automatic methods were proposed as an alternative but they often rely of resources available in a limited number of languages and their performances are still far behind the quality of manual translations.Our work concerns bilingual lexicon extraction from multilingual parallel and comparable corpora, in other words, the process of finding translation pairs among the common multilingual vocabulary available in such corpora.