Efficient θ-subsumption based on graph algorithms

  • Tobias Scheffer
  • Ralf Herbrich
  • Fritz Wysotzki
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1314)


The θ-subsumption problem is crucial to the efficiency of ILP learning systems. We discuss two θ-subsumption algorithms based on strategies for preselecting suitable matching literais. The class of clauses, for which subsumption becomes polynomial, is a superset of the deterministic clauses. We further map the general problem of θ-subsumption to a certain problem of finding a clique of fixed size in a graph, and in return show that a specialization of the pruning strategy of the Carraghan and Pardalos clique algorithm provides a dramatic reduction of the subsumption search space. We also present empirical results for the mesh design data set.


Pruning Strategy Argument Position Maximum Clique Problem Connection Graph Graph Context 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [BM92]
    B. Dolsak and S. Muggleton. The application of inductive logic programming to finite-element mesh design. In Inductive Logic Programming, London, 1992. Academic Press.Google Scholar
  2. [CP90]
    R. Carraghan and P. Pardalos. An exact algorithm for the maximum clique problem. Operations Research Letters, 9:375–382, 1990.Google Scholar
  3. [DB93]
    L. DeRaedt and M. Bruynooghe. A theory of clausal discovery. In Proc. Workshop on ILP, 1993.Google Scholar
  4. [DMR92]
    S. Dzeroski, S. Muggleton, and S. Russel. Pac-learnability of determinate logic programs. In Proc. 5th ACM Workshop on Computational Learning Theory, pages 128–135, 1992.Google Scholar
  5. [Eis81]
    N. Eisinger. Subsumption and connection graphs. In Proc. IJCAI, 1981.Google Scholar
  6. [FGL+91]
    U. Feige, S. Goldwasser, L. Lovasz, S. Safra, and M. Szegedy. Approximating the maxclique is almost NP-complete. In Proc. 32nd IEEE Symp. on Foundations of Comp. Sci., 1991.Google Scholar
  7. [GHP96]
    L. Gibbons, D. Hearn, and P. Pardalos. A continuous based heuristic for the maximum clique problem. In Clique, Graph Coloring and Satisfiability: Second DIMACS Implementation Challenge, 1996.Google Scholar
  8. [GL85]
    G. Gottlob and A. Leitsch. On the efficiency of subsumption algorithms. J. ACM, 32(2):280–295, 1985.Google Scholar
  9. [Got87]
    G. Gottlob. Subsumption and implication. Information Processing Letters, 24:109–111, 1987.Google Scholar
  10. [GW96a]
    P. Geibel and F. Wysotzki.Learning relational concepts with decision trees. In Proc. ICML, 1996.Google Scholar
  11. [GW96b]
    P. Geibel and F. Wysotzki. Relational learning with decision trees. In Proc. ECAI, 1996.Google Scholar
  12. [JT96]
    D. S. Johnson and M. A. Trick, editors. Clique, Graph Coloring and Satisfiability: Second DIMACS Implementation Challenge, DIMACS series, 1996.Google Scholar
  13. [KC92]
    B. M. Kim and J. W. Cho. A new subsumption method in the connection graph proof procedure. Theoretical Computer Science, 103:283–309, 1992.Google Scholar
  14. [KL94]
    J.-U. Kietz and M. Liibbe. An efficient subsumption algorithm for inductive logic programming. In Proc. International Conference on Machine Learning, 1994.Google Scholar
  15. [KN86]
    D. Kapur and P. Narendran. NP-completeness of the set unification and matching problems. In Proc. 8th International Conference on Automated Deduction, 1986.Google Scholar
  16. [Kow75]
    R. Kowalski. A proof procedure using connection graphs. J. ACM, 22(4):572–595, 1975.Google Scholar
  17. [Lov78]
    D. W. Loveland.Automated theorem proving: A logical basis. Elsevier, North Holland, 1978.Google Scholar
  18. [MF90]
    S. Muggleton and C. Feng. Efficient induction of logic programs. In Proc. 1st Conf. on Algorithmic Learning Theory, pages 368–381, 1990.Google Scholar
  19. [Mug93]
    S. Muggleton. Inverting implication. Artificial Intelligence Journal, 1993.Google Scholar
  20. [P1o70]
    G. D. Plotkin.A note on inductive generalization.In B. Meltzer and D. Michie, editors, Machine Intelligence, volume 5, pages 153–163, 1970.Google Scholar
  21. [Rob65]
    J. A. Robinson. A machine-oriented logic based on the resolution principle. J. ACM, 12(1):23–41, 1965.Google Scholar
  22. [Sic76]
    Sharon Sickel. A search technique for clause interconnectivity graphs. IEEE Transactions on Computers, C-25(8):823–835, 1976.Google Scholar
  23. [Soc88]
    R. Socher. A subsumption algorithm based on characteristic matrices. In Proc. 9th Int. Conf. on Automated Deduction, 1988.Google Scholar
  24. [Tin76]
    G. Tinhofer. Zum algorithmischen Nachweis der Isomorphie von endlichen Graphen.In H. Noltemeier, editor, Graphen, Algorithmen, Datenstrukturen. 2. Fachtagung über Graph entheoretische Konzepte der Informatik. Carl Hanser Verlag, 1976.Google Scholar
  25. [UW81]
    S. Unger and F. Wysotzki. Lernfähige Klassifizierungssysteme. Akademie Verlag Berlin, 1981.Google Scholar
  26. [vdLNC93]
    P. van der Laag and S. Nienhuys-Cheng. Subsumption and refinement in model inference. In Machine Learning: ECML, 1993.Google Scholar
  27. [Wei76]
    B. Weisfeiler. On Construction and Identification of Graphs. Number 558 in Lecture Notes in Mathematics. Springer Verlag, Berlin, 1976.Google Scholar
  28. [WSK81]
    F. Wysotzki, J. Selbig, and W. Kolbe. Concept learning by structured examples-an algebraic approach. In Proceedings of the 7th International Joint Conference on Artificial Intelligence, 1981.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1997

Authors and Affiliations

  • Tobias Scheffer
    • 1
  • Ralf Herbrich
    • 1
  • Fritz Wysotzki
    • 1
  1. 1.Artificial Intelligence Research Group, Sekr. FR 5-8Technische Universität BerlinBerlin

Personalised recommendations