DOI QR코드

DOI QR Code

An Analytical Study on Performance Factors of Automatic Classification based on Machine Learning

기계학습에 기초한 자동분류의 성능 요소에 관한 연구

  • Received : 2016.05.13
  • Accepted : 2016.06.02
  • Published : 2016.06.30

Abstract

This study examined the factors affecting the performance of automatic classification for the domestic conference papers based on machine learning techniques. In particular, In view of the classification performance that assigning automatically the class labels to the papers in Proceedings of the Conference of Korean Society for Information Management using Rocchio algorithm, I investigated the characteristics of the key factors (classifier formation methods, training set size, weighting schemes, label assigning methods) through the diversified experiments. Consequently, It is more effective that apply proper parameters (${\beta}$, ${\lambda}$) and training set size (more than 5 years) according to the classification environments and properties of the document set. and If the performance is equivalent, I discovered that the use of the more simple methods (single weighting schemes) is very efficient. Also, because the classification of domestic papers is corresponding with multi-label classification which assigning more than one label to an article, it is necessary to develop the optimum classification model based on the characteristics of the key factors in consideration of this environment.

국내 학술회의 논문으로 구성된 문헌집합을 대상으로 기계학습에 기초한 자동분류의 성능에 영향을 미치는 요소들을 검토하였다. 특히 구현이 쉽고 컴퓨터 처리 속도가 빠른 로치오 알고리즘을 사용하여 "한국정보관리학회 학술대회 논문집"의 논문에 주제 범주를 자동 할당하는 분류 성능 측면에서 분류기 생성 방법, 학습집합 규모, 가중치부여 기법, 범주 할당 방법 등 주요 요소들의 특성을 다각적인 실험을 통해 살펴보았다. 결과적으로 분류 환경 및 문헌집합의 특성에 따라 파라미터(${\beta}$, ${\lambda}$)와 학습집합의 크기(5년 이상)를 적절하게 적용하는 것이 효과적이며, 동등한 성능 수준이라면 보다 단순한 단일 가중치부여 기법을 사용하여 분류의 효율성을 높일 수 있음을 발견하였다. 또한 국내 학술회의 논문의 분류는 특정 논문에 하나 이상의 범주가 부여되는 복수-범주 분류(multi-label classification)가 실제 환경에 부합한다고 할 수 있으므로, 이러한 환경을 고려하여 주요 성능 요소들의 특성에 기초한 최적의 분류 모델을 개발할 필요가 있다.

Keywords

References

  1. 강승식 (2002). 한국어 형태소 분석과 정보검색. 서울: 홍릉출판사. (Kang, Seung-Shik (2002). Korean Morphology and Information Retrieval. Hongrung Publishing Company.)
  2. 김성희, 엄재은 (2008). 기계학습을 이용한 문서 자동분류에 관한 연구. 정보관리연구, 39(4), 47-66. http://dx.doi.org/10.1633/jim.2008.39.4.047 (Kim, Seong-Hee, & Eom, Jae-Eun (2008). A study on the documents' automatic classification using machine learning. Journal of Information Management, 39(4), 47-66. http://dx.doi.org/10.1633/JIM.2008.39.4.047)
  3. 김용환, 정영미 (2012). 위키피디아를 이용한 분류자질 선정에 관한 연구. 정보관리학회지, 29(2), 155-171. http://dx.doi.org/10.3743/kosim.2012.29.2.155 (Kim, Yong-Hwan, & Chung, Young-Mee (2012). An experimental study on feature selection using Wikipedia for text categorization. Journal of the Korean Society for Information Management, 29(2), 155-171. http://dx.doi.org/10.3743/kosim.2012.29.2.155)
  4. 김종민, 유창동 (2014). 특징 추출 비용에 민감한 분류를 위한 선형 분류기 최적화 알고리즘. 2014년도 대한전자공학회 하계학술대회 논문집, 37(1), 2021-2024. (Kim, Jong-Min, & Yoo, Chang D. (2014). Linear classifier optimization for feature acquisition cost-sensitive classification. Proceedings of the IEEK Conference, 37(1), 2021-2024.)
  5. 김판준 (2006a). 기계학습을 통한 디스크립터 자동부여에 관한 연구. 정보관리학회지, 23(1), 279-299. http://dx.doi.org/10.3743/kosim.2006.23.1.279 (Kim, Pan Jun (2006a). A study on automatic assignment of descriptors using machine learning. Journal of the Korean Society for Information Management, 23(1), 279-299. http://dx.doi.org/10.3743/kosim.2006.23.1.279)
  6. 김판준 (2006b). 로치오 알고리즘을 이용한 학술지 논문의 디스크립터 자동부여에 관한 연구. 정보관리학회지, 23(3), 69-89. http://dx.doi.org/10.3743/kosim.2006.23.3.069 (Kim, Pan Jun (2006b). A study on the automatic descriptor assignment for scientific journal articles uing rocchio algorithm. Journal of the Korean Society for Information Management, 23(3), 69-89. http://dx.doi.org/10.3743/kosim.2006.23.3.069)
  7. 김판준 (2008). 용어 가중치부여 기법을 이용한 로치오 분류기의 성능 향상에 관한 연구. 정보관리학회지, 25(1), 211-233. http://dx.doi.org/10.3743/kosim.2008.25.1.211 (Kim, Pan Jun (2008). A study on the performance improvement of rocchio classifier with term weighting methods. Journal of the Korean Society for Information Management, 25(1), 211-233. http://dx.doi.org/10.3743/kosim.2008.25.1.211)
  8. 김판준, 이재윤 (2007). 문헌간 유사도를 이용한 자동분류에서 미분류 문헌의 활용에 관한 연구. 정보관리학회지, 24(1), 251-271. http://dx.doi.org/10.3743/kosim.2007.24.1.251 (Kim, Pan Jun, & Lee, Jae Yun (2007). Utilizing unlabeled documents in automatic classification with inter-document similarities. Journal of the Korean Society for Information Management, 24(1), 251-271. http://dx.doi.org/10.3743/kosim.2007.24.1.251)
  9. 김판준, 이재윤 (2012). 디스크립터 자동 할당을 위한 저자키워드의 재분류에 관한 실험적 연구. 정보관리학회지, 29(2), 225-246. http://dx.doi.org/10.3743/kosim.2012.29.2.225 (Kim, Pan Jun, & Lee, Jae Yun (2012). A study on the reclassification of author keywords for automatic assignment of descriptors. Journal of the Korean Society for Information Management, 29(2), 225-246. http://dx.doi.org/10.3743/kosim.2012.29.2.225)
  10. 김판준, 이재윤 (2014). 해외 데이터베이스의 통제키 워드에 기초한 국내 학술지 논문의 자동분류 성능향상에 관한 실험적 연구. 한국문헌정보학회지, 48(3), 491-510. http://dx.doi.org/10.4275/kslis.2014.48.3.491 (Kim, Pan Jun, & Lee, Jae Yun (2014). An experimental study on the performance improvement of automatic classification for the articles of Korean journals based on controlled keywords in international database. Journal of the Korean Society for Library and Information Science, 48(3), 491-510. http://dx.doi.org/10.4275/kslis.2014.48.3.491)
  11. 송성전, 정영미 (2012). 용어의 문맥활용을 통한 문헌 자동 분류의 성능 향상에 관한 연구. 정보관리학회지, 29(2), 205-224. http://dx.doi.org/10.3743/kosim.2012.29.2.205 (Song, Sung-Jeon, & Chung, Young-Mee (2012). A study on improving the performance of document classification using the context of terms. Journal of the Korean Society for Information Management, 29(2), 205-224. http://dx.doi.org/10.3743/kosim.2012.29.2.205)
  12. 심 경 (2006). 문헌범주화에서 학습문헌수 최적화에 관한 연구. 정보관리학회지, 23(4), 277-294. http://dx.doi.org/10.3743/kosim.2006.23.4.277 (Shim, Kyung (2006). Optimization of number of training documents in text categorization. Journal of the Korean Society for Information Management, 23(4), 277-294. http://dx.doi.org/10.3743/kosim.2006.23.4.277)
  13. 심경, 정영미 (2006). 학습문헌집합에 기 부여된 범주의 정확성과 문헌 범주화 성능. 정보관리학회지, 23(2), 265-285. http://dx.doi.org/10.3743/kosim.2006.23.2.265 (Shim, Kyung, & Chung, Young-Mee (2006). The effect of the quality of pre-assigned subject categories on the text categorization performance. Journal of the Korean Society for Information Management, 23(2), 265-285. http://dx.doi.org/10.3743/kosim.2006.23.2.265)
  14. 이용구 (2009). 기계번역을 이용한 교차언어 문서 범주화의 분류 성능 분석. 한국문헌정보학회지, 43(1), 313-332. http://dx.doi.org/10.4275/kslis.2009.43.1.313 (Lee, Yong-Gu (2009). Classification performance analysis of cross-language text categorization using machine translation. Journal of the Korean Society for Library and Information Science, 43(1), 313-332. http://dx.doi.org/10.4275/kslis.2009.43.1.313)
  15. 이용구 (2013). 문헌빈도와 장서빈도를 이용한 kNN 분류기의 자질선정에 관한 연구. 한국도서관.정보학회지, 44(1), 27-47. http://dx.doi.org/10.16981/kliss.44.1.201303.27 (Lee, Yong-Gu (2013). A study on feature selection for kNN classifier using document frequency and collection frequency. Journal of Korean Library and Information Science Society, 44(1), 27-47. http://dx.doi.org/10.16981/kliss.44.1.201303.27)
  16. 이재윤 (2005a) 문서측 자질선정을 이용한 고속 문서분류기의 성능향상에 관한 연구. 정보관리연구, 36(4), 51-69. http://dx.doi.org/10.1633/jim.2005.36.4.051 (Lee, Jae Yun (2005a). Improving the performance of a fast text classifier with document-side feature selection. Journal of Information Management, 36(4), 51-69. http://dx.doi.org/10.1633/jim.2005.36.4.051)
  17. 이재윤 (2005b). 자질 선정 기준과 가중치 할당 방식간의 관계를 고려한 문서 자동분류의 개선에 대한 연구. 한국문헌정보학회지, 39(2), 123-146. http://dx.doi.org/10.4275/kslis.2005.39.2.123 (Lee, Jae Yun (2005b). An empirical study on improving the performance of text categorization considering the relationships between feature selection criteria and weighting methods. Journal of the Korean Society for Library and Information Science, 39(2), 123-146. http://dx.doi.org/10.4275/kslis.2005.39.2.123)
  18. 정은경 (2009). 문서범주화 성능 향상을 위한 의미기반 자질확장에 관한 연구. 정보관리학회지, 26(3), 261-278. http://dx.doi.org/10.3743/kosim.2009.26.3.261 (Chung, Eun-Kyung (2009). A semantic-based feature expansion approach for improving the effectiveness of text categorization by using wordNet. Journal of the Korean Society for Information Management, 26(3), 261-278. http://dx.doi.org/10.3743/kosim.2009.26.3.261)
  19. 한국연구재단 학술연구분야 분류표 (2015). Retrieved from http://www.nrf.re.kr
  20. 한국학술지인용색인 웹사이트 (2016). Retrieved from https://www.kci.go.kr
  21. AI-Salemi, B., Aziz, M., Juzaiddin, A., & Noah, S. (2015). Boosting algorithms with topic modeling for multi-label text categorization: A comparative empirical study. Journal of Information Science, 41(5), 732-746. http://dx.doi.org/10.1177/0165551515590079
  22. Aliferis, C. F., Statnikov, A., Tsamardinos, I., Mani, S., & Koutsoukos, X. D. (2010a). Local causal and markov blanket induction for causal discovery and feature selection for classification. Part I: Algorithms and empirical evaluation. Journal of Machine Learning Research, 11, 171-234.
  23. Aliferis, C. F., Statnikov, A., Tsamardinos, I., Mani, S., & Koutsoukos, X. D. (2010b). Local causal and markov blanket induction for causal discovery and feature selection for classification. Part II: Analysis and extensions. Journal of Machine Learning Research, 11, 235-284.
  24. Aphinyanaphongs, Y., Fu, L., Li, Z., & Peskin, E. R. (2014). A comprehensive empirical comparison of modern supervised classification and feature selection methods for text categorization. Journal of the Association for Information Science and Technology, 65(10), 1964-1987. http://dx.doi.org/10.1002/asi.23110
  25. Chen, E., Lin, Y., Xiong, H., Luo, Q., & Ma, H. (2011). Exploiting probabilistic topic models to improve text categorization under class imbalance. Information Processing and Management, 47(2), 202-214. http://dx.doi.org/10.1016/j.ipm.2010.07.003
  26. Cohen, W. W., & Singer, Y. (1999). Context-sensitive learning methods for text categorization. ACM Transactions on Information Systems, 17(2), 141-173. http://dx.doi.org/10.1145/306686.306688
  27. Debole, F., & Sebastiani, F. (2003). Supervised term weighting for automated text categorization. Proceedings of the 18th ACM Symposium on Applied Computing (SAC) 2003, 784-788. http://dx.doi.org/10.1145/952532.952688
  28. Devi, P. R., Suganya, B. R., & Abirami, S. (2015). Multi-label learning with class-based features using extended centroid-based classification technique (CCBF). Procedia Computer Science, 54, 405-411. http://dx.doi.org/10.1016/j.procs.2015.06.047
  29. Forman, G. (2003). An extensive empirical study of feature selection metrics for text classification. Journal of Machine Learning Research, 3, 1289-1305.
  30. Forman, G., & Kirshenbaum, E. (2008). Extremely fast text feature extraction for classification and indexing. Proceedings of the 17th ACM Conference on Information and Knowledge Mining (CIKM) 2008, 26-30. http://dx.doi.org/10.1145/1458082.1458243
  31. Foulds, J., & Frank, E. (2010). A review of multi-instance learning assumptions. Knowledge Engineering Review, 25(1), 1-25. http://dx.doi.org/10.1017/s026988890999035x
  32. Genkin, A., Lewis, D. D., & Madigan, D. (2007). Large-scale bayesian logistic regression for text categorization. Technometrics, 49(3), 291-304. http://dx.doi.org/10.1198/004017007000000245
  33. Harish B. S., Guru D. S., & Manjunath, S. (2010). Representation and classification of text documents: A brief review. Proceedings of the IJCA Special Issue on Recent Trends in Image Processing and Pattern Recognition, RTIPPR, 110-119.
  34. Hull, D. A. (1994). Improving text retrieval for the routing problem using latent semantic indexing. SIGIR-94, 282-291. http://dx.doi.org/10.1007/978-1-4471-2099-5_29
  35. Ittner, J. D., Lewis, D. D., & Ahn, D. D. (1995). Text categorization of low quality images. Proceedings of the 4th Annual Symposium on Document Analysis and Information Retrieval (SDAIR) 1995, 301-315.
  36. Jain, R., & Nitin, P. (2015). Feature selection for effective text classification using semantic information. International Journal of Computer Applications, 113(10), 18-25. http://dx.doi.org/10.5120/19861-1818
  37. Jiang, S., Pang, G., Wu, M., & Kuang, L. (2012). An improved k-nearest-neighbor algorithm for text categorization. Expert Systems with Applications, 39(1), 1503-1509. http://dx.doi.org/10.1016/j.eswa.2011.08.040
  38. Joachims, T. (1997). A probabilistic analysis of the rocchio algorithm with tdf for text categorization. Proceedings of the International Conference on Machine Learning (ICML) 1997, 143-151.
  39. Khan, A., Baharudin, B., & Lee, L. H. (2010). A review of machine learning algorithms for text-documents classification. Journal of Advances in Information Technology, 1(1), 4-20. http://dx.doi.org/10.4304/jait.1.1.4-20
  40. Kohavi, R., & John, G. H. (1997). Wrappers for feature subset selection. Artificial Intelligence, 97(1-2), 273-324. http://dx.doi.org/10.1016/s0004-3702(97)00043-x
  41. Korde, V., & Mahender, C. N. (2012). Text classification and classifiers: A survey. International Journal of Artificial Intelligence & Applications (IJAIA), 3(2), 85-99. https://doi.org/10.5121/ijaia.2012.3208
  42. Kumar, M. A., & Gopal, M. (2010). A comparison study on multiple binary-class SVM methods for unilabel text categorization. Pattern Recognition Letters, 31(11), 1437-1444. http://dx.doi.org/10.1016/j.patrec.2010.02.015
  43. Li, C. H., & Park, S. C. (2009). An efficient document classification model using an improved back propagation neural network and singular value decomposition. Expert Systems with Applications, 36(2), 3208-3215. http://dx.doi.org/10.1016/j.eswa.2008.01.014
  44. Liu, Y., Loh, H. T., Yousef-Toumi, K., & Tor, S. B. (2007). Handling of imbalanced data in text classification: Category-based term weights. Natural Language Processing and Text Mining, 171-192. http://dx.doi.org/10.1007/978-1-84628-754-1_10
  45. Moschitti, A. (2003). Study on optimal parameter tuning for rocchio text classifier. Lecture Notes in Computer Science, (2633), 420-435. http://dx.doi.org/10.1007/3-540-36618-0_30
  46. Pang, G., & Jiang, S. (2013). A generalized cluster centroid based classifier for text categorization. Information Processing and Management, 49(2), 576-586. http://dx.doi.org/10.1016/j.ipm.2012.10.003
  47. Patra, A., & Singh, D. (2013). A survey report on text classification with different term weighing methods and comparison between classification algorithms. International Journal of Computer Applications, 75(7), 14-18. http://dx.doi.org/10.5120/13122-0472
  48. Read, J., Pfahringer, B., Holmes, G., & Frank, E. (2011). Classifier chains for multi-label classification. Machine Learning, 85(3), 333-359. http://dx.doi.org/10.1007/s10994-011-5256-5
  49. Rogati, M., & Yang, Y. (2002). High-performing feature selection for text classification. Proceedings of the 11th International Conference on Information and knowledge management (CIKM) 2002, 4-9. http://dx.doi.org/10.1145/584792.584911
  50. Schapire, R. E., & Singer, Y. (2000). BoosTexter: A boosting-based system for text categorization. Machine Learning, 39(2-3), 135-168. https://doi.org/10.1023/A:1007649029923
  51. Schapire, R. E., Singer, Y., & Singhal, A. (1998). Boosting and rocchio applied to text filtering. Proceedings of the 21st Annual International ACM SIGIR conference on research and development in information retrieval (SIGIR) 1998, 215-223. http://dx.doi.org/10.1145/290941.290996
  52. Sebastiani, F. (2002). Machine learning in automated text categorization. ACM Computing Surveys, 34(1), 1-47. https://doi.org/10.1145/505282.505283
  53. Singhal, A., Mitra, M., & Buckley, C. (1997). Learning routing queries in a query zone. Proceedings of the 20th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR) 1997, 25-32. http://dx.doi.org/10.1145/258525.258530
  54. Sokolova, M., & Lapalme, G. (2009). A systematic analysis of performance measures for classification tasks. Information Processing and Management, 45(4), 427-437. http://dx.doi.org/10.1016/j.ipm.2009.03.002
  55. Tan, S. (2008). An improved centroid classifier for text categorization. Expert Systems with Applications, 35(1-2), 279-285. http://dx.doi.org/10.1016/j.eswa.2007.06.028
  56. Tarrago, D. S., Cornelis, C., Bello, R., & Herrera, F. (2014). A multi-instance learning wrapper based on the Rocchio classifier for web index recommendation. Knowledge-Based Systems, 59, 173-181. http://dx.doi.org/10.1016/j.knosys.2014.01.008
  57. Torii, M., Yin, L., Nguyen, T., Mazumdar, C. T., Liu, H., Hartley, D. M., & Nelson, N. P. (2011). An exploratory study of a text classification framework for Internet-based surveillance of emerging epidemics. International Journal of Medical Informatics, 80(1), 56-66. http://dx.doi.org/10.1016/j.ijmedinf.2010.10.015
  58. Tsoumakas, G., & Katakis, I. (2007). Multi-label classification: An overview. International Journal of Data Warehousing and Mining (IJDWM), 3(3), 1-13. http://dx.doi.org/10.4018/jdwm.2007070101
  59. Uysal, A. K., & Gunal, S. (2014). The impact of preprocessing on text classification. Information Processing and Management, 50(1), 104-112. http://dx.doi.org/10.1016/j.ipm.2013.08.006
  60. Villena-Roman, J., Collada-Perez, S., Lana-Serrano, S., & Gonzalez-Cristobal, J. C. (2011). Hybrid approach combining machine learning and a rule-based expert system for text categorization. Proceedings of the 24th International Florida Artificial Intelligence Research Society Conference, 323-328.
  61. Wu, C. (2009). Behavior-based spam detection using a hybrid method of rule-based techniques and neural networks. Expert Systems with Applications, 36(3), 4321-4330. http://dx.doi.org/10.1016/j.eswa.2008.03.002
  62. Yang, Y. (1999). An evaluation of statistical approaches to text categorization. Information Retrieval, 1(1-2), 69-90. https://doi.org/10.1023/A:1009982220290
  63. Yang, Y., & Liu, X. (1999). A re-examination for text categorization methods. Proceedings of the 22nd Annual International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR) 1999, 42-49. http://dx.doi.org/10.1145/312624.312647
  64. Yang, Y., & Pedersen, J. O. (1997). A comparative study on feature selection in text categorization. Proceedings of the 14th International Conference on Machine Learning (ICML) 1997, 412-420.
  65. Yu, B., Xu, Z., & Li, C. (2008). Latent semantic analysis for text categorization using neural network. Knowledge-Based Systems, 21(8), 900-904. http://dx.doi.org/10.1016/j.knosys.2008.03.045
  66. Zeng, A., & Huang, Y. (2011). A text classification algorithm based on rocchio and hierarchical clustering. Lecture Notes in Computer Science, 432-439. http://dx.doi.org/10.1007/978-3-642-24728-6_59
  67. Zhang, W., Yoshida, T., & Tang, X. (2011). A comparative study of TF*IDF, LSI and multiwords for text classification. Expert Systems with Applications, 38(3), 2758-2765. http://dx.doi.org/10.1016/j.eswa.2010.08.066