Translingual Information Retrieval: A Comparative Evaluation

Translingual information retrieval TIR con sists of providing a query in one language and searching document collections in one or more di erent languages This paper introduces new TIR methods and reports on comparative TIR experiments with these new methods and with previously reported ones in a realistic setting Methods fall into two categories query trans lation based and statistical IR approaches es tablishing translingual associations The re sults show that using bilingual corpora for au tomated extraction of term equivalences in con text outperforms other methods Translin gual versions of the Generalized Vector Space Model GVSM and Latent Semantic Indexing LSI perform relatively well as does translin gual pseudo relevance feedback PRF All showed relatively small performance loss be tween monolingual and translingual versions Query translation based on a general machine readable bilingual dictionary heretofore the most popular method did not match the per formance of other more sophisticated methods Also the previous very high LSI results in the literature were discon rmed by more realistic relevance based evaluations