Ranking Translation Candidates Acquired from Comparable Corpora

Domain-specific bilingual lexicons extracted from domain-specific comparable corpora provide for one term a list of ranked translation candidates. This study proposes to re-rank these translation candidates. We suggest that a term and its translation appear in comparable sentences that can be extracted from domainspecific comparable corpora. For a source term and a list of translation candidates, we propose a method to identify and align the best source and target sentences that contain the term and its translation candidates. We report results with two language pairs (French-English and FrenchGerman) using domain-specific comparable corpora. Our method significantly improves the top 1, top 5 and top 10 precisions of a domain-specific bilingual lexicon, and thus, provides a better user-