Machine Learning

, Volume 14, Issue 2, pp 193–217

A Polynomial Approach to the Constructive Induction of Structural Knowledge

  • Jörg-Uwe Kietz
  • Katharina Morik
Article

Abstract

The representation formalism as well as the representation language is of great importance for the success of machine learning. The representation formalism should be expressive, efficient, useful, and applicable. First-order logic needs to be restricted in order to be efficient for inductive and deductive reasoning. In the field of knowledge representation, term subsumption formalisms have been developed which are efficient and expressive. In this article, a learning algorithm, KLUSTER, is described that represents concept definitions in this formalism. KLUSTER enhances the representation language if this is necessary for the discrimination of concepts. Hence, KLUSTER is a constructive induction program. KLUSTER builds the most specific generalization and a most general discrimination in polynomial time. It embeds these concept learning problems into the overall task of learning a hierarchy of concepts.

Constructive induction restrictions of first-order logic for learning learning most specific generalizations 

References

  1. Bisson, G. (1990). KBG, a knowledge-based generalizer. In 7th ICML-90 (pp. 9–15). Morgan Kaufmann.Google Scholar
  2. Borgida, A., Brachman, R.J., McGuinness, D.L., & Resnick, L.A. (1989). Classic: a structural data model for objects. Proceedings of ACM SIGMOD-89 (pp. 58–67). Portland, OR.Google Scholar
  3. Brachman, R.J., and Schmolze, J.G. (1985). An overview of the KL-ONE knowledge representation system. Cognitive Science, 9, 171–216.Google Scholar
  4. Brachman, R.J. (1977). What's in a concept: structural foundations for semantic networks. International Journal of Man-Machine Studies, 9, 127–152.Google Scholar
  5. Brachman, R.J., Gilbert, V.P., & Levesque, H.J. (1985). An essential hybrid reasoning system. In IJCAI-85 (pp. 532–538), Morgan Kaufmann.Google Scholar
  6. Buntine, W. (1988). Generalized subsumption and its applications to induction and redundancy. Artificial Intelligence, 36, 149–176.Google Scholar
  7. Ceri, S., Gottlob, G., & Tanca, L. (1990). Logic programming and databases. New York: Springer.Google Scholar
  8. Cohen, W.W., Borgida, A., & Hirsh, H. (in press). Computing least common subsumers in description logic. Proceedings of AAAI-92.Google Scholar
  9. Cohen, W.W., & Hirsh, H. (1992). Learnability of description logics. Proceedings of the Fourth COLT, ACM Press, pp. 116–127.Google Scholar
  10. Donini, F.M., Lenzerini, M., Nardi, C., & Nutt, W. (1991). Tractble concept languages. Proceedings IJCAI-91 (pp. 458–463).Google Scholar
  11. Emde, W., Habel, C. & Rollinger, C.R. (1983). The discovery of the equator or concept-driven learning. Proceedings IJCAI-83 (pp. 455–458), Morgan Kaufmann.Google Scholar
  12. Fisher, D.H. (1987). Knowledge acquisition via incremental conceptual clustering, Machine Learning, 2, 139–172.Google Scholar
  13. Garey, M.R., & Johnson, D.S. (1979). Computers and intractability—A guide to the theory of NP-completeness. New York: Freeman.Google Scholar
  14. Haussler, D. (1989). Learning conjunctive concepts in structural domains. Machine Learning, 4, 7–40.Google Scholar
  15. Kearns, M.J. (1990). The computational complexity of machine learning. Cambridge, MA, London: MIT Press.Google Scholar
  16. Kietz, J.-U. (1992). A comparative study of structural most specific generalizations used in machine learning. Proceedings of ECAI'92 Workshop W18.Google Scholar
  17. Kietz, J.-U. (1993). Some lower bounds for the computational complexity of inductive logic programming. Proceedings of Machine Learning ECML-93. Berlin: Springer, pp. 115–123.Google Scholar
  18. Kietz, J.-U., & Wrobel, S. (1991). Controlling the complexity of learning in logic through syntactic and task-oriented models. Proceedings of Inductive Logic Programming Workshop, Porto. Also in S. Muggleton (Ed.) (1992). Inductive logic programming. New York: Academic Press, pp. 311–333.Google Scholar
  19. Kietz, J.-U. (1988). Incremental and reversible acquisition of taxonomies. Proceedings of the European Knowledge Acquisition Workshop. Birlinghoven: GMD-Studien No. 143.Google Scholar
  20. Kodratoff, Y., and Tecuci, G. (1989). The central role of explanations in DISCIPLE. In K. Morik (Ed.), Knowledge representation and organization in machine learning. New York: Springer, pp. 135–147.Google Scholar
  21. Lebowitz, M. (1987). Experiments with incremental concept formation: UNIMEM. Machine Learning, 2, 103–138.Google Scholar
  22. Luck, K.V., Nebel, B., Peltason, C., & Schmiedel, A. (1987). The anatomy of the BACK system (KIT-Report No. 41). Berlin: Technical University Berlin.Google Scholar
  23. Michalski, R.S. (1983). A theory and methodology of inductive learning. In Machine learning—An artificial intelligence approach (Vol. I). Los Altos, CA: Morgan Kaufmann, pp. 83–134.Google Scholar
  24. Michalski, R.S. (1990). Learning flexible concepts: fundamental ideas and a method based on two-tiered representation. In Y. Kodratoff & R.S. Michalski (Eds.), Machine Learning—An artificial intelligence approach (Vol. III). San Mateo, CA: Morgan Kaufmann, pp. 63–111.Google Scholar
  25. Morik, K., & Kietz, J.-U. (1989). A bootstrapping approach to conceptual clustering. Proceedings of the Sixth International Workshop on Machine Learning, Morgan Kaufmann.Google Scholar
  26. Morik, K., Wrobel, S., Kietz, J.-U., & Emde, W. (1993). Knowledge Acquisition and Machine Learning—Theory, Methods, and Applications. London: Academic Press.Google Scholar
  27. Moser, M.G. (1983). An overview of NIKL, the new implementation of KL-one. In Research in knowledge representatin and natural language understanding. Cambridge, MA: B. Beranek and Newman Inc.Google Scholar
  28. Muggleton, S., and Buntine, W. (1988). Machine invention of first-order predicates by inverting resolution. Proceedings of IWML-88. Ann Arbor, MI: Morgan Kaufmann.Google Scholar
  29. Muggleton, S. (1990). Inductive logic programming. Proceedings of the First Conference on Algorithmic Learning Theory. Tokyo: Ohmsha.Google Scholar
  30. Muggleton, S. & Feng, C. (1990). Efficient induction of logic programs. Proceedings of the First Conference on Algorithmic Learning Theory. Tokyo: Ohmsha.Google Scholar
  31. Nebel, B. (1990). Reasoning and revision in hybrid representation systems. New York: Springer.Google Scholar
  32. Peltason, C., Luck, K., & Kindermann, C.K. (1991). Terminological logic users workshop (KIT-Report 95). Berlin: Technical University Berlin.Google Scholar
  33. Peltason, C., Schmiedel, A., Kindermann, C., & Quantz, J. (1989). The BACK System revisited (KIT-Report 75). Berlin: Technical University Berlin.Google Scholar
  34. Plotkin, G.D. (1970). A note on inductive generalization. Machine Intelligence, 5, 153–163.Google Scholar
  35. Quinlan, R. (1990). Learning logical defnitions from relations. Machine Learning, 5, 239–266.Google Scholar
  36. Shapiro, E. (1983). Algorithmic program debugging. Cambridge, MA: MIT Press.Google Scholar
  37. Stepp, R.E. & Michalski, R.S. (1986). Conceptual clustering: Inventing goal-oriented classifications of structured objects. In R. Michalski, J. Carbonell, & T. Michell (Eds.), Machine learning—An AI approach (Vol. II). San Mateo: Morgan Kaufmann, pp. 471–498.Google Scholar
  38. Thompson, K., & Langley, P. (1989). Incremental concept formation with composite objects. Proceedings of the 6th International Workshop on Machine Learning, Morgan Kaufmann, pp. 373–374.Google Scholar
  39. Vilain, M. (1985). The restricted language architecture of a hybrid reasoning system. IJCAI-85 (pp. 547–551).Google Scholar
  40. Wrobel, S. (1987). Higher-order concepts in a tractable knowledge representation. In K. Morik (Ed.), Proceedings of the German Workshop on Artificial Intelligence. Berlin: Springer, pp. 129–138.Google Scholar

Copyright information

© Kluwer Academic Publishers 1994

Authors and Affiliations

  • Jörg-Uwe Kietz
    • 1
  • Katharina Morik
    • 2
  1. 1.German National Research Centre for Computer Science (GMD)Institute for Applied Information Technology Schloβ BirlinghovenSt. AugustinGermany
  2. 2.Department of Computer ScienceUniversity Dortmund, Artificial Intelligence (LS VIII)Dortmund 50Germany

Personalised recommendations