Öz

Combination of multiple evidence of document relevance for effective retrieval has been subject of many past and recent researches.We propose a formal approach to normalizing scores for metasearch by taking the distributions of the scores into account. It is shown that by equalizing the exponential distribution of scores of the top nonrelevant documents yields the best metasearch performance reported in the literature.

Skor Dağılımlı Üst Arama Modeli

Öz

Bir belgenin kullanıcının bilgi ihtiyacı ile ilgililiği hakkında birden çok kanıtın birleştirilmesi, geçmişte ve günümüzde bir çok araştırmaya konu olmuştur. Bu çalışmada farklı arama sistemlerine ait ilgisiz belgelerin skor dağılımlarının eşitlenmesi ile elde edilen ortalama üst arama performansının, literatürde şimdiye kadar rapor edilmiş olan diğer yöntemlerden daha iyi sonuçlar verdiği gösterilmiştir.

Kaynakça

1. Belkin, N., Cool, C., Croft, W., Callan, J., The effect of multiple query representation on information retrieval system performance. In: Proceedings of the 16th Annual International ACM-SIGIR Conference on Research and Development in Information Retrieval. Pittsburg, PA, USA, pp. 339-346, 1993.
2. Belkin, N., Kantor, P., Fox, E., Shaw, J., Combining the evidence of multiple query representations for information retrieval. Information Processing & Management 31 (3), 431-448, 1995.
3. C. Buckley., Implementation of the SMART information retrieval system. Cornell University, Technical report 35-686, 1985.
4. Croft, W., Combining approaches to in formation retrieval. In: Croft, W. (Ed.), Advances in Information Retrieval. Kluwer Academic Publishers, pp. 1-36, 2000.
5. Croft, W., Combining approaches to information retrieval. In: Croft, W. (Ed.), Advances in Information Retrieval. Kluwer Academic Publishers, pp. 1-36, 2000.
6. Deogun, J. S., Sever, H., Raghavan, V. V., Structural abstractions of hypertext documents for web-based retrieval. In: Wagner, R. R. (Ed.). Proceedings of Ninth international Workshop on Database and Expert Systems Applications, (in conjunction with DEXA'98). Vienna, Austria, pp. 385 390 , Aug. 1998.
7. Gauch, S., Wang, G., Gomez, M., Profusion: Intelligent fusion from multiple, distributed search engines. Journal of Universal Computer Science 2(9), 637-649, Sep. 1996. http://www.jucs.org/jucs_2_9/profusion_intelligen t_fusion .from.
8. Katzer, J.,McGill, M.J., Frakes, W., DasGupta, P. A study of overlap among document representations. Information Technology: Research and Development 1 (4), 261-274, Oct 1982.

9. Lee, J. H., Combining multiple evidence from different properties of weighting schemes, in: Fox. E. A. (Ed.). Proceedings of the 18th Annual Inter¬national ACM-SIGIR Conference on Research and Development in Information Retrieval. Seattle, WA, pp. 180-188, July 1995.
10. Montague, M., Aslam, J., Relevance score normalization for metasearch. In: Proceedings of the ACM 10th Annual International Conference on Information and Knowledge Management (CIKM). Atlanta, Georgia pp. 427-433, November 2001.
11. Manmatha, R., Sever, H., A Formal Approach to Score Normalization for Metasearch, In:Human Language Technology Conference (HLT'02) San Diego, March 2002.
12. Pinto, D., McCallum, A., Wei, X. and Croft, W.B. Table Extraction Using Conditional Random Fields. In: Proceedings of SIGIR ’03 Conference, Toronto, Canada , 2003.
13. Rodrigo, A., Botafoga, R., Ehud, R. Shneiderman, B., Structural Analysis of Hypertexts: Identifying Hierarchies and Useful Metrics. ACMTIS. pp. 142-180, 1992.
14. Saracevic, T., Kantor P., A study of information seeking and retrieving. III. searchers, searchs, and overlap. Journel of American Society for Information Science 39(3), 197-216, 1998.
15. Tonta, Y., Bitirim, Y. and Sever, H., Türkçe Arama Motorlarında Performans Değerlendirme, Damla Yayınevi Ltd., pp. 1-152, 2002.
16. Tumer, K., Ghosh, J., Linear and order statistics combiners for pattern classification, In: Sharkey. A. (Ed.), Combining Artificial Neural Networks. Springer-Verlag, pp. 127 162, 1999.
17. Vogt, C., Cottrel, G., Fusion via a linear combination of scores. Information Retrieval 1(2-3), 151-173, 1999.
18. H. Sever and M. Tolun., Comparison of Normalization Techniques for Metasearch. ADVIS'02, Lecture Notes in Computer Science, Springer Verlag, Vol. 2457, pp. 133-143, 2002.
19. J. H. Lee., Analyses of multiple evidence combination. In the Proc. of the 20th Intl. Conf. on Research and Development in Information Retrieval (SIGIR’97), pp. 267–276, 1997.
20. A. Arampatzis and A. van Hameren, Maximum likelihood estimation for filtering thresholds. In In the Proc. of the 24th ACM SIGIR conf. on Research and Developement in Information Retrieval, pp. 285–293, Sept 2001.
21. Y. Zhang and J. Callan., Maximum likelihood estimation for filtering thresholds. In the Proc. of the 24th ACM SIGIR conf. on Research and Developement in Information Retrieval, pages 294–302, Sept 2001.
22. R. Manmatha, T. Rath, and F. Feng, Sept 2001. Modeling score distributions for combining the outputs of search engines. In the Proc. of the 24th ACM SIGIR conf. on Research and Developement in Information Retrieval, pages 267–275.
23. Sweets, J., 1963. Information Retrieval Systems. Science 141,245-250.