Experiments with transitive dictionary translation and pseudo-relevance feedback using graded relevance assessments

  • Authors:

  • Raija Lehtokangas;Heikki Keskustalo;Kalervo Järvelin

  • Affiliations:

  • Department of Information Studies, FIN-33014 University of Tampere, Finland;Department of Information Studies, FIN-33014 University of Tampere, Finland;Department of Information Studies, FIN-33014 University of Tampere, Finland

  • Venue:
  • Journal of the American Society for Information Science and Technology
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this article, the authors present evaluation results fortransitive dictionary-based cross-language information retrieval(CLIR) using graded relevance assessments in a best match retrievalenvironment. A text database containing newspaper articles and arelated set of 35 search topics were used in the tests. Sourcelanguage topics (in English, German, and Swedish) wereautomatically translated into the target language (Finnish) via anintermediate (or pivot) language. Effectiveness of the transitivelytranslated queries was compared to that of the directly translatedand monolingual Finnish queries. Pseudo-relevance feedback (PRF)was also used to expand the original transitive target queries.Cross-language information retrieval performance was evaluated onthree relevance thresholds: stringent, regular, and liberal. Thetransitive translations performed well achieving, on the average,8593% of the direct translation performance, and 6672% ofmonolingual performance. Moreover, PRF was successful in raisingthe performance of transitive translation routes in absolute termsas well as in relation to monolingual and direct translationperformance applying PRF. © 2008 Wiley Periodicals, Inc.