Abstract
The reliability of the topics within the Cross Language Evaluation Forum (CLEF) needs to be validated constantly to justify the efforts for experiments within CLEF and to demonstrate the reliability of the results as far as possible. The analysis presented in this paper is concerned with several aspects. Continuing and expanding a study from 2002, we investigate the difficulty of topics and the correlation between the retrieval quality for topics and the occurrence of proper names. © Springer-Verlag 2004.
Cite
CITATION STYLE
Mandl, T., & Womser-Hacker, C. (2004). Analysis of the reliability of the multilingual topic set for the cross language evaluation forum. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 3237, 21–28. https://doi.org/10.1007/978-3-540-30222-3_3
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
