Automatic combination of multiple ranked retrieval systems

作者: Richard K. Belew , Brian T. Bartell , Garrison W. Cottrell

DOI: 10.5555/188490.188554

关键词:

摘要: Retrieval performance can often be improved significantly by using a number of different retrieval algorithms and combining the results, in contrast to just single algorithm. This is because algorithms, or experts, emphasize document query features when determining relevance therefore retrieve sets documents. However, it unclear how experts are combined, general, yield superior overall estimate. We propose method which estimates made automatically combined result performance. apply two expert combination tasks. The applications demonstrate that identify high combinations also novel means for effectiveness experts.

参考文章(18)
Donna Harman, Overview of the First Text REtrieval Conference. international acm sigir conference on research and development in information retrieval. pp. 36- 47 ,(1993)
D K Harman, The first text REtrieval conference (TREC-1) Special Publication (NIST SP) - 500-207. ,(1993) , 10.6028/NIST.SP.500-207
Paul Thompson, Description of the PRC CEO algorithm for TREC-2 text retrieval conference. pp. 271- 274 ,(1993)
Edward A. Fox, Durgesh Rao, Russell Modlin, Joseph A. Shaw, M. Prabhakar Koushik, Combining evidence from multiple searches text retrieval conference. pp. 319- 328 ,(1992)
Brian Theodore Bartell, Optimizing ranking functions: a connectionist approach to adaptive information retrieval University of California at San Diego. ,(1994)
Donna Harman, Overview of the first TREC conference international acm sigir conference on research and development in information retrieval. pp. 36- 47 ,(1993) , 10.1145/160688.160692
Nicholas J. Belkin, C. Cool, W. Bruce Croft, James P. Callan, The effect multiple query representations on information retrieval system performance international acm sigir conference on research and development in information retrieval. pp. 339- 346 ,(1993) , 10.1145/160688.160760
Jeffrey Katzer, Judith Tessier, William Frakes, Padmini Das-Gupta, A Study of the Overlap Among Document Representations international acm sigir conference on research and development in information retrieval. ,vol. 51, pp. 106- 114 ,(1983) , 10.1145/3130348.3130352
Norbert Fuhr, Chris Buckley, A probabilistic learning approach for document indexing international acm sigir conference on research and development in information retrieval. ,vol. 9, pp. 223- 248 ,(1991) , 10.1145/125187.125189