Cross-Language Information Retrieval via Dictionary-based and Statistics-based Methods

SADAT FATIHA  (9851207)

As Internet resources become accessible to more and more countries, there is a need to develop methods for Cross Language Information Retrieval for different languages. Most Cross-Language Information Retrieval researches have focused on dictionary-based method with a combination to statistics approach to avoid the problem of ambiguity. Query expansion is used to improve the effectiveness of information retrieval and to dramatically reduce the errors such an approach normally makes. In this research, the approach of using a bilingual dictionary, by including an interactive feedback loop in the task of query expansion before translation, is aimed. A second query expansion is used after the disambiguation of translation candidates. We apply this method to an English-French module, in the Cross-language Information Retrieval.
This method achieved 91.95% of the monolingual counterpart, in term of average precision.
Finally, a combination of French-English and English-Japanese modules via a transitive translation is discussed.



サダト ファテイア (9851207)