DOI QR코드

DOI QR Code

A Study on the Performance Improvement of Rocchio Classifier with Term Weighting Methods

용어 가중치부여 기법을 이용한 로치오 분류기의 성능 향상에 관한 연구

  • 김판준 (연세대학교 문헌정보학과)
  • Published : 2008.03.30

Abstract

This study examines various weighting methods for improving the performance of automatic classification based on Rocchio algorithm on two collections(LISA, Reuters-21578). First, three factors for weighting are identified as document factor, document factor, category factor for each weighting schemes, the performance of each was investigated. Second, the performance of combined weighting methods between the single schemes were examined. As a result, for the single schemes based on each factor, category-factor-based schemes showed the best performance, document set-factor-based schemes the second, and document-factor-based schemes the worst. For the combined weighting schemes, the schemes(idf*cat) which combine document set factor with category factor show better performance than the combined schemes(tf*cat or ltf*cat) which combine document factor with category factor as well as the common schemes (tfidf or ltfidf) that combining document factor with document set factor. However, according to the results of comparing the single weighting schemes with combined weighting schemes in the view of the collections, while category-factor-based schemes(cat only) perform best on LISA, the combined schemes(idf*cat) which combine document set factor with category factor showed best performance on the Reuters-21578. Therefore for the practical application of the weighting methods, it needs careful consideration of the categories in a collection for automatic classification.

로치오 알고리즘에 기반한 자동분류의 성능 향상을 위하여 두 개의 실험집단(LISA, Reuters-21578)을 대상으로 여러 가중치부여 기법들을 검토하였다. 먼저, 가중치 산출에 사용되는 요소를 크게 문헌요소(document factor), 문헌집합 요소(document set factor), 범주 요소(category factor)의 세 가지로 구분하여 각 요소별 단일 가중치부석 기법의 분류 성능을 살펴보았고, 다음으로 이들 가중치 요소들 간의 조합 가중치부여 기법에 따른 성능을 알아보았다. 그 결과, 각 요소별로는 범주 요소가 가장 좋은 성능을 보였고, 그 다음이 문헌집합 요소, 그리고 문헌 요소가 가장 낮은 성능을 나타냈다. 가중치 요소 간의 조합에서는 일반적으로 사용되는 문헌 요소와 문헌집합 요소의 조합 가중치(tfidf or ltfidf)와 함께 문헌 요소를 포함하는 조합(tf*cat or ltf*cat) 보다는, 오히려 문헌 요소를 배제하고 문헌 집합 요소를 범주 요소와 결합한 조합 가중치 기법(idf*cat)이 가장 좋은 성능을 보였다. 그러나 실험집단 측면에서 단일 가중치와 조합 가중치를 서로 비교한 결과에 따르면, LISA에서 범주 요소만을 사용한 단일 가중치(cat only)가 가장 좋은 성능을 보인 반면, Reuters-21578에서는 문헌집합 요소와 범주 요소간의 조합 가중치(idf*cat)의 성능이 가장 우수한 것으로 나타났다. 따라서 가중치부여 기법에 대한 실제 적용에서는, 분류 대상이 되는 문헌집단 내 범주들의 특성을 신중하게 고려할 필요가 있다.

References

  1. 김판준. 2007, 로치오 알고리즘을 이용한 자동분류에서 용어 가중치 기법. $\ulcorner$문헌정보학논집$\lrcorner$, 명지대학교, 문헌정보학회, 제9호: 157-185
  2. 김판준. 200sa. 기계학습을 통한 디스크립터 자동부여에 관한 연구. $\ulcorner$정보관리학회지$\lrcorner$, 23(1) : 279-299 https://doi.org/10.3743/KOSIM.2006.23.1.279
  3. 김판준. 2006b. 로치오 알고리즘을 이용한 학술지 논문의 디스크립터 자동부여에 관한연구 $\ulcorner$정보관리학회지$\lrcorner$, 23(3): 69-89 https://doi.org/10.3743/KOSIM.2006.23.3.069
  4. 이재윤. 2005. 자질 선정 기준과 가중치 할당 방식간의 관계를 고려한 문서 자동분류의개선에 대한 연구 $\ulcorner$한국문헌정보학회지$\lrcorner$, 39(2) 123-146
  5. 이재윤, 최보영, 정영미. 2000. 문헌 자동분류에서 용어가중치 기법에 대한 연구. 제7회 한국정보관리학회 학술대회 논문집, 2000년 8월 17일 이화여자대학교, pp.41-44
  6. 정영미. 1993. $\ulcorner$정보검색론$\lrcorner$. 서울: 구미무역 (주) 출판부
  7. Brank, J., M. Grobelnik, N. Milic-Frayling & D. Mladenic. 2002. "Interaction of feature selection methods and linear classification models." In: Proceedings of the ICML-02 Workshop on Text Learning, Sydney. [cited 2007. 5. 3].
  8. Castillo M. D. and Serrano J. I. 2004. "A multistrategy approach for digital text categorization from imbalanced documents." ACM SIGKDD Explorations Newsletter: Special Issue on Leaning from Imbalanced Datasets, 6(1): 70-79 https://doi.org/10.1145/1007730.1007740
  9. Debole, Franca and F. Sebastiani. 2003. "Supervised term weighting for automated text categorization." In: Proceedings of SAC-03, 18th ACM Symposium on Applied Computing, New York: ACM. 784-788 https://doi.org/10.1145/952532.952688
  10. Deng, Zhi-Hong et al. 2004. A Comparative study on feature weight in text categorization. In Proceedings of The Sixth Asia Pacific Web Conference(APWEB 2004), Hangzhou, China, April 14-17, LNCS 3007, 588-597
  11. Forman G. 2003. "An extensive empirical study of feature selection metrics for text classification." The Journal of Machine Learning Research. Special Issue on Variable and Feature Selection. 3: 1289-1305 https://doi.org/10.1162/153244303322753670
  12. Geng, L. and Howard J. Hamilton. 2006. "Choosing the right lens: finding what is interesting in data mining." eds. by Guillet, Fabrice, Howard J. Hamilton. Quality Measures in Data Mining. Springer, pp. 3-24
  13. How, Bong Chih and Narayanan K. 2004. An empirical study of feature selection for text categorization based on term weightage. In Proceedings of the 2004 IEEE/WIC/ACM International Conference on Web Intelligence. pp.592-602 https://doi.org/10.1109/WI.2004.10060
  14. Joachims. Thorsten. 1996. "A probabilistic analysis of the rocchio algorithm with TFIDF for text categorization." Proceedings of ICML-97. 14th International Conference on Machine Learning, Nashville. TN: 143-151
  15. Joachims, Thorsten. 1998. "Text categorization with support vector machines: learning with many relevant features." In: Proceedings of the 10th European Conference on Machine Learning: 137-142
  16. Lan, Man. Chew-Lim Tan, and Hwee-Boon LOW. 2006. "Proposing a new term weighting scheme for text categorization." In 21st National Conference on Artificial Intelligence, AAAI-2006, 16-20 July. 2006. Boston. Massachusetts. USA. [cited 2007. 3. 7.].
  17. Liu, Ying. Han Tong Loh. Kamal Yousef Tourni. and Shu Beng Tor. 2007. "Han dling of imbalanced data in text classification: category- based term weights." In: Kao, Anne and Stephen R. Poteet eds. Natural Language Processing and Text Mining. Springer.. pp.171-192
  18. Papineni, K. 2001. "Why inverse document frequency?" Proceedings of the North American Association for Computational Linguistics, NAACI New York. pp. 25- 32
  19. Prabowo. Ruby and Mike Thelwall. 2006. "A comparison of feature selection methods for an evloving RSS feed corpus." Information Processing and Management. 42: 1491-1512 https://doi.org/10.1016/j.ipm.2006.03.018
  20. Robertson S. 2004. Understanding inverse document frequency: on theoretical arguments for IDF. Journal of Documentation. 60(5): 503-520 https://doi.org/10.1108/00220410410560582
  21. Robertson. S. E. and K. Sparck Jones. 1976. "Relevance weighting of search terms." JASIS. 27(3): 129-146 https://doi.org/10.1002/asi.4630270302
  22. Rogati, M and Y. Yang. 2002. High-Performing Feature Selection for Text Classification." In: Proceedings of the eleventh international conference on Information and knowledge management. CIKM '02. [cited 2007. 8. 23.]. < citeseer .ist.psu.edu/rogati02highperforming.html>
  23. Salton. G.. H. Wu, and C. T. Yu. 1981. "The Measurement of term importance in automatic indexing." JASIS, 32(3): 175-186 https://doi.org/10.1002/asi.4630320304
  24. Salton. G. and M. J. McGill. 1983. Introduction to Modern Information Retrieval. N. Y.: McGraw-Hill
  25. Sebastiani. Fabrizio. 2002. "Machine learning in automated text categorization." ACM Computing Surveys, 34(1): 1-47 https://doi.org/10.1145/505282.505283
  26. Soucy P. and Guy W. Mineau. 2005. "Beyond TFIDF weighting for text categorization in the vector space model." IJCAI-05 proceedings, 1130-1135. [cited 2007. 2. 20].
  27. Yang, Y and Liu X. 1999. "A re-examination of text categorization methods." In: Proceedings of the 22nd Annual International ACM SIGIR Conference on Research and Development in Information Retrieval https://doi.org/10.1145/312624.312647
  28. Yang, Y. 1999. "Evaluation of statistical approaches to text categorization." Information Retrieval. 1: 69-90 https://doi.org/10.1023/A:1009982220290
  29. Yang. Y. and Pedersen J. O. 1997. "A comparative Study on Feature Selection in Text Categorization." In: Proceedings of ICML-97, 14th International Conference on Machine Learning: 412420 https://doi.org/10.1023/A:1009982220290
  30. Yu. C. T. and G. Salton 1976. "Precision weighting-an effective automatic indexing method." Journal of Association for Computing Machinery, 23(1): 7688
  31. Yu. C. T.. K. Lam. and G. Salton. 1982. "Term weighting in information retrieval using the term precision model." Journal of Association for Computing Machinery, 29(1): 152-170 https://doi.org/10.1145/322290.322300
  32. Zheng Z., X. Wu and R. Srihari. 2004. "Feature selection for text categorization on imbalanced data." ACM SIGKDD Explorations Newsletter: Special Issue on Leaning from Imbalanced Datasets, 6(1): 80-89 https://doi.org/10.1145/1007730.1007741