This paper describes the official runs of the Twenty-One group for the first CLEF workshop. The Twenty-One group participated in the monolingual, bilingual and multilingual tasks. The following new techniques are introduced in this paper. In the bilingual task we experimented with different methods to estimate translation probabilities. In the multilingual task we experimented with refinements on raw-score merging techniques and with a new relevance feedback algorithm that re-estimates both the model's translation probabilities and the relevance weights. Finally, we performed preliminary experiments to exploit the web to generate translation probabilities and bilingual dictionaries, notably for English-Italian and English-Dutch.
|Number of pages||13|
|Journal||Lecture notes in artificial intelligence|
|Publication status||Published - 2001|
|Event||1st Workshop of the Cross-Language Evaluation Forum, CLEF 2000 - Lisbon, Portugal|
Duration: 21 Sep 2000 → 22 Sep 2000
Conference number: 1