ABSTRACT
The intuition that different text classifiers behave in qualitatively different ways has long motivated attempts to build a better metaclassifier via some combination of classifiers. We introduce a probabilistic method for combining classifiers that considers the context-sensitive reliabilities of contributing classifiers. The method harnesses reliability indicators---variables that provide a valuable signal about the performance of classifiers in different situations. We provide background, present procedures for building metaclassifiers that take into consideration both reliability indicators and classifier outputs, and review a set of comparative studies undertaken to evaluate the methodology.
- K. Al-Kofahi, A. Tyrrell, A. Vacher, T. Travers, and P. Jackson. Combining multiple classifiers for text categorization. In CIKM '01, pages 97--104, 2001. Google ScholarDigital Library
- B. T. Bartell, G. W. Cottrell, and R. K. Belew. Automatic combination of multiple ranked retrieval systems. In SIGIR '94, pages 173--181, 1994. Google ScholarDigital Library
- N. Belkin, C. Cool, W. Croft, and J. Callan. The effect of multiple query representations on information retrieval system performance. In SIGIR '93, pages 339--346, 1993. Google ScholarDigital Library
- D. Chickering, D. Heckerman, and C. Meek. A Bayesian approach to learning Bayesian networks with local structure. In UAI '97, pages 80--89, 1997. Google ScholarDigital Library
- M. Corporation. WinMine Toolkit v1.0. http://research.microsoft.com/\~ dmax /WinMine/ContactInfo.html, 2001.Google Scholar
- S. T. Dumais and H. Chen. Hierarchical classification of web content. In SIGIR '00, pages 256--263, 2000. Google ScholarDigital Library
- S. T. Dumais, J. Platt, D. Heckerman, and M. Sahami. Inductive learning algorithms and representations for text categorization. In CIKM '98, pages 148--155, 1998. Google ScholarDigital Library
- D. Heckerman, D. Chickering, C. Meek, R. Rounthwaite, and C. Kadie. Dependency networks for inference, collaborative filtering, and data visualization. Journal of Machine Learning Research, 1:49--75, 2000. Google ScholarDigital Library
- D. Hull, J. Pedersen, and H. Schuetze. Method combination for document filtering. In SIGIR '96, pages 279--287, 1996. Google ScholarDigital Library
- T. Joachims. Text categorization with support vector machines: Learning with many relevant features. In ECML '98, pages 137--142, 1998. Google ScholarDigital Library
- J. Katzer, M. McGill, J. Tessier, W. Frakes, and P. DasGupta. A study of the overlap among document representations. Information Technology: Research and Development, 1:261--274, 1982.Google Scholar
- W. Lam and K.-Y. Lai. A meta-learning approach for text categorization. In SIGIR '01, pages 303--309, 2001. Google ScholarDigital Library
- L. S. Larkey and W. B. Croft. Combining classifiers in text categorization. In SIGIR '96, pages 289--297, 1996. Google ScholarDigital Library
- D. D. Lewis. A sequential algorithm for training text classifiers: Corrigendum and additional data. SIGIR Forum, 29(2):13--19, Fall 1995. Google ScholarDigital Library
- D. D. Lewis. Reuters-21578, distribution 1.0. http://www.daviddlewis.com/resources /testcollections/reuters21578, January 1997.Google Scholar
- D. D. Lewis and W. A. Gale. A sequential algorithm for training text classifiers. In SIGIR '94, pages 3--12, 1994. Google ScholarDigital Library
- D. D. Lewis, R. E. Schapire, J. P. Callan, and R. Papka. Training algorithms for linear text classifiers. In SIGIR '96, pages 298--306, 1996. Google ScholarDigital Library
- Y. Li and A. Jain. Classification of text documents. The Computer Journal, 41(8):537--546, 1998.Google ScholarCross Ref
- A. McCallum and K. Nigam. A comparison of event models for naive bayes text classification. In Working Notes of AAAI 1998, Workshop on Learning for Text Categorization, pages 41--48, 1998.Google Scholar
- J. C. Platt. Probabilistic outputs for support vector machines and comparisons to regularized likelihood methods. In A. J. Smola, P. Bartlett, B. Scholkopf, and D. Schuurmans, editors, Advances in Large Margin Classifiers. MIT Press, 1999.Google Scholar
- F. Provost and T. Fawcett. Robust classification for imprecise environments. Machine Learning, 42:203--231, 2001. Google ScholarDigital Library
- T. Rajashekar and W. Croft. Combining automatic and manual index representations in probabilistic retrieval. Journal of the American Society for Information Science, 6(4):272--283, 1995. Google ScholarDigital Library
- R. E. Schapire and Y. Singer. BoosTexter: A boosting-based system for text categorization. Machine Learning, 39:135--168, 2000. Google ScholarDigital Library
- J. Shaw and E. Fox. Combination of multiple searches. In D. K. Harman, editor, TREC-3 Conference, number 500-225 in NIST Special Publication, pages 105--108, 1995.Google Scholar
- K. Ting and I. Witten. Issues in stacked generalization. Journal of Artificial Intelligence Research, 10:271--289, 1999. Google ScholarCross Ref
- K. Toyama and E. Horvitz. Bayesian modality fusion: Probabilistic integration of multiple vision algorithms for head tracking. In ACCV 2000, Fourth Asian Conference on Computer Vision, 2000.Google Scholar
- C. van Rijsbergen. Information Retrieval. Butterworths, London, 1979. Google ScholarDigital Library
- S. Weiss, C. Apte, F. Damerau, D. Johnson, F. Oles, T. Goets, and T. Hampp. Maximizing text-mining performance. IEEE Intelligent Systems, 14(4), 1999. Google ScholarDigital Library
- D. H. Wolpert. Stacked generalization. Neural Networks, 5:241--259, 1992. Google ScholarDigital Library
- Y. Yang, T. Ault, and T. Pierce. Combining multiple learning strategies for effective cross validation. In ICML '00, pages 1167--1182, 2000. Google ScholarDigital Library
- Y. Yang and X. Liu. A re-examination of text categorization methods. In SIGIR '99, pages 42--49, 1999. Google ScholarDigital Library
Index Terms
- Probabilistic combination of text classifiers using reliability indicators: models and results
Recommendations
The Combination of Text Classifiers Using Reliability Indicators
AbstractThe intuition that different text classifiers behave in qualitatively different ways has long motivated attempts to build a better metaclassifier via some combination of classifiers. We introduce a probabilistic method for combining classifiers ...
Ensemble of feature sets and classification algorithms for sentiment classification
In this paper, we make a comparative study of the effectiveness of ensemble technique for sentiment classification. The ensemble framework is applied to sentiment classification tasks, with the aim of efficiently integrating different feature sets and ...
Training more discriminative multi-class classifiers for hand detection
In this paper, an effective algorithm is developed to learn more discriminative multi-class classifiers for achieving more accurate hand detection. At each round of boosting, a set of shared stump classifiers with relatively low discrimination power are ...
Comments