On Arabic-English cross-language information retrieval: a machine translation approach

Abstract

Machine translation (MT) is an automatic process that translates from one human language to another by using context information. We empirically evaluate the use of an MT-based approach for query translation in an Arabic-English cross-language information retrieval (CLIR) system, called ALKAFI, using the TREC-7 and TREC-9 topics and collections. The effect of the query length on the MT performance is also investigated in order to explore how much context is actually required for successful MT processing.