NTCIR-5 Chinese, English, Korean Cross Language Retrieval Experiments using PIRCS

In NTCIR-5 our focus is to see if web-assisted query expansion is useful, and to test an EnglishKorean bilingual dictionary. We participated in Chinese, Japanese, Korean and English monolingual retrieval using also web expansion for Chinese and English. We also performed Chinese-English, English-Chinese, English-Korean bilingual, and Chinese-Korean pivot bilingual CLIR. The query translation approach was employed. MT translation was combined with our web-based entity-oriented translation which translates named entities extracted from the original query as well as web-based expansion terms. For English-Korean, a dictionary translation method was also used. In general, monolingual retrieval results are about median, while cross-lingual runs (except Chinese-English) appear to be among the top results. For English-Korean, our bilingual dictionary translation by itself is not competitive, but when concatenated with web translation came to within 5% of MT with web translation. Using the web to expand a query before retrieval was not successful except for monolingual English and pivot bilingual ChineseKorean retrieval.