Report on CLIR Task for the NTCIR-4 Evaluation Campaign

This paper describes our first participation in an evaluation campaign involving three Asian languages (NTCIR-4). Our project has three objectives: 1) to compare the retrieval performances of eleven IR models used to carry out monolingual retrievals with these languages; 2) to analyze the relative merit of various freely available translation tools used to translate English-language topics into Chinese, Japanese or Korean; and 3) to evaluate the relative performance of the various merging strategies used to combine separate result lists extracted from a corpus written in English, Chinese, Japanese or Korean.