An Implementation of Optimal Rules Discovery System: An Integrated Approach Based on Concept Hierarchies, Information Gain, and Rough Sets

최적 규칙 발견 시스템의 구현: 개념 계층과 정보 이득 및 라프셋에 의한 통합 접근

  • 김진상 (계명대학교 컴퓨터전자공학부)
  • Published : 2000.06.01

Abstract

This study suggests an integrated method based on concept hierarchies, information gain, and rough set theory for efficient discovery rules from a large amount of data, and implements an optimal rules discovery system. Our approach applies attribute-oriented concept ascension technique to extract generalized knowledge from a database, knowledge reduction technique to remove superfluous attributes and attribute values, and significance of attributes to induce optimal rules. The system first reduces the size of database by removing the duplicate tuples through the condition attributes which have no influences on the decision attributes, and finally induces simplified optimal rules by removing the superfluous attribute values by analyzing the dependency relationships among the attributes. And we induce some decision rules from actual data by using the system and test rules to new data, and evaluate that the rules are well suited to them.

본 연구는 대량의 데이터에서 효율적으로 최적 규칙을 발견하기 위해 개념 계층과 정보 이득 및 라프셋 이론에 딕반한 통합 방법을 제시하고,이를 최적 규칙 발견 시스템으로 구현한다. 본 방법은 데이터베이스에 있는 데이터에서 일반화된 지식을 추출하기 위한 속성중심의 개념 상승 기법과 불필요한 속성 및 속성값을 제거하기 위한 지식 감축 기법을 적용하며, 최적 규칙의 도출을 위해 속성의 중요도를 사용한다. 본 시스템은 먼저, 속성값 개념의 일반화에 의해 종복 튜플을 제거함으로써 데이터 베이스의 크기를 줄이고, 결정속성에 뎡향을 주지않는 조건속성을 제거하여 간략화된 최적 규칙을 유도한다.그리고 실제 데이터에 적용하여 결정 규칙을 유도하고 그 규칙을 새로운 데이터에 테스트햐 봄으로써 새로운 데이터에도 잘 적용됨을 보인다.

Keywords

References

  1. Rough Sets and Data Mining Data Mining Using Attribute-Oriented Generalization and Information Reduction N. Cercone;H. Hamilton;X. Hu;N. Shan;T. Lin;N. Cercone(eds.)
  2. Knowledge Discovery in Databases : A Rule-Based Attribute-Oriented Approach D. Cheung;A. Fu;J. Han
  3. Intelligent Integration of Information v.6 no.2/3 CoBase : A Scalable and Extensible Cooperative Information System W. Chu;H. Yang;K. Chiang;M. Minock;G. Chow;C. Larson;G. Wiederhold(eds.)
  4. Advances in Knowledge Discovery and Data Mining U. Fayyad;G. Piatetsky Shapiro;P. Smyth;R. Uthurusamy(eds.)
  5. Machine Learning v.2 no.2 Knowledge Acquisition via Incremental Conceptual Clustering D. Fisher
  6. Proceeding of the 18th Conference on Very Large Data Bases Knowledge Discovery in Database : An Attribute-Oriented Approach J. Han;Y. Cai;N. Cercone
  7. Proc. RSKD'93 An Attribute-Oriented Rough Set Approach for Knowledge Discovery in Databases X. Hu;N. Cercone;J. Han
  8. Generalization and Decision Tree Induction : Efficient Classification in Data Mining M. Kamber;L. Winstone;W. Gong;S. Cheng;J. Han
  9. Proc. RSKD'93 Finding Reducts in Composed Information Systems M. Kryszkiewicz;Henryk Rybinski
  10. Rough Sets-Theoretical Aspects of Reasoning about Data Z. Paw"lak
  11. Machine Learning v.1 no.1 Induction of Decision Trees J. Quinlan
  12. Siftware : Tools for Data Mining and Knowledge Discovery
  13. Intelligent Decision Support Handbook of Advances and Applications of the Rough Set Theory The Discernibility Matrices and Functions in Information Systems A. Skowron;C. Rauszer;Slowinski(eds.)
  14. 정보과학회지 v.13 no.5 ID3계열의 귀납적 기계학습 박영택;이강로
  15. 러프집합과 응용 이성주;정환묵;최완규
  16. FUZZ-IEEE'99, The 8th Int'l Conf. on Fuzzy Systems Generation of Approximation Rules Using Information Gain 정홍;최경옥;정환묵
  17. Rough Sets and Data Mining Generation of Multiple Knowledge from Databases Based on Rough Set Theory X. Hu;N. Cercone;W. Ziarko;T. Lin(ed.);N. Cercone(ed.)