Advertisement

Optimized Rule Mining Through a Unified Framework for Interestingness Measures

  • Céline Hébert
  • Bruno Crémilleux
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4081)

Abstract

The large amount of association rules resulting from a KDD process makes the exploitation of the patterns embedded in the database difficult even impossible. In order to address this problem, various interestingness measures were proposed for selecting the most relevant rules. Nevertheless, the choice of an appropriate measure remains a hard task and the use of several measures may lead to conflicting information. In this paper, we propose a unified framework for a set of interestingness measures \(\mathcal{M}\) and prove that most of the usual objective measures behave in a similar way. In the context of classification rules, we show that each measure of \(\mathcal{M}\) admits a lower bound on condition that a minimal frequency threshold and a maximal number of exceptions are considered. Furthermore, our framework enables to characterize the whole collection of the rules simultaneously optimizing all the measures of \(\mathcal{M}\). We finally provide a method to mine a rule cover of this collection.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Agrawal, R., Imielinski, T., Swami, A.N.: Mining association rules between sets of items in large databases. In: Buneman, P., Jajodia, S. (eds.) SIGMOD 1993 Conference, pp. 207–216. ACM Press, New York (1993)CrossRefGoogle Scholar
  2. 2.
    Bayardo, J.R.J., Agrawal, R.: Mining the most interesting rules. In: KDD 1999, pp. 145–154 (1999)Google Scholar
  3. 3.
    Brin, S., Motwani, R., Silverstein, C.: Beyond market baskets: Generalizing association rules to correlations. In: Peckham, J. (ed.) SIGMOD 1997, Proceedings ACM SIGMOD International Conference on Management of Data, Tucson, Arizona, USA, May 13-15, 1997, pp. 265–276. ACM Press, New York (1997)CrossRefGoogle Scholar
  4. 4.
    Crémilleux, B., Boulicaut, J.-F.: Simplest rules characterizing classes generated by delta-free sets. In: 22nd Int. Conf. on Knowledge Based Systems and Applied Artificial Intelligence (ES 2002), Cambridge, UK, December 2002, pp. 33–46 (2002)Google Scholar
  5. 5.
    Dong, G., Li, J.: Efficient mining of emerging patterns: discovering trends and differences. In: Proceedings of the Fifth International Conference on Knowledge Discovery and Data Mining (ACM SIGKDD 1999), San Diego, CA, pp. 43–52. ACM Press, New York (1999)CrossRefGoogle Scholar
  6. 6.
    Francisci, D., Collard, M.: Multi-criteria evaluation of interesting dependencies according to a data mining approach. In: Congress on Evolutionary Computation, Canberra, Australia, December 2003, pp. 1568–1574. IEEE Press, Los Alamitos (2003)CrossRefGoogle Scholar
  7. 7.
    Fürnkranz, J., Flach, P.A.: Roc ’n’ rule learning-towards a better understanding of covering algorithms. Machine Learning 58(1), 39–77 (2005)zbMATHCrossRefGoogle Scholar
  8. 8.
    Hébert, C., Crémilleux, B.: Obtention de règles optimisant un ensemble de mesures. In: Conférence francophone sur l’apprentissage automatique (CAp 2006), Trégastel, France (2006)Google Scholar
  9. 9.
    Hilderman, R.J., Hamilton, H.J.: Measuring the interestingness of discovered knowledge: A principled approach. Intell. Data Anal. 7(4), 347–382 (2003)zbMATHGoogle Scholar
  10. 10.
    Liu, B., Hsu, W., Ma, Y.: Integrating classification and association rules mining. In: Proceedings of Fourth International Conference on Knowledge Discovery & Data Mining (KDD 1998), New York, August 1998, pp. 80–86. AAAI Press, Menlo Park (1998)Google Scholar
  11. 11.
    Pasquier, N., Bastide, Y., Taouil, R., Lakhal, L.: Discovering frequent closed itemsets for association rules. In: Embley, D.W. (ed.) ER 1997. LNCS, vol. 1331, pp. 299–312. Springer, Heidelberg (1997)Google Scholar
  12. 12.
    Piatetsky-Shapiro, G.: Discovery, analysis, and presentation of strong rules. In: Knowledge Discovery in Databases, pp. 229–248. AAAI/MIT Press (1991)Google Scholar
  13. 13.
    Plasse, M., Niang, N., Saporta, G., Leblond, L.: Une comparaison de certains indices de pertinence des règles d’association. In: Ritschard, G., Djeraba, C. (eds.) EGC, Cépaduès-Éditions. Revue des Nouvelles Technologies de l’Information, vol. RNTI-E-6, pp. 561–568 (2006)Google Scholar
  14. 14.
    Sebag, M., Schoenauer, M.: Generation of rules with certainty and con dence factors from incomplete and incoherent learning bases. In: Boose, M.L.J., Gaines, B. (eds.) European Knowledge Acquisistion Workshop, EKAW 1988, pp. 28–1–28–20 (1988)Google Scholar
  15. 15.
    Tan, P.-N., Kumar, V., Srivastava, J.: Selecting the right interestingness measure for association patterns. In: KDD, pp. 32–41. ACM Press, New York (2002)Google Scholar
  16. 16.
    Vaillant, B., Lenca, P., Lallich, S.: A clustering of interestingness measures. In: The 7th International Conference on Discovery Science, October 2004, pp. 290–297 (2004)Google Scholar
  17. 17.
    Vaillant, B., Meyer, P., Prudhomme, E., Lallich, S., Lenca, P., Bigaret, S.: Mesurer l’intérêt des règles d’association. In: EGC, pp. 421–426 (2005)Google Scholar
  18. 18.
    Wille, R.: Ordered sets. In: Restructuring lattice theory: an approach based on hierachies of concepts, pp. 445–470. Reidel, Dordrecht (1982)Google Scholar
  19. 19.
    Zaki, M.J.: Generating non-redundant association rules. In: KDD 2000, pp. 34–43 (2000)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Céline Hébert
    • 1
  • Bruno Crémilleux
    • 1
  1. 1.GREYC, CNRS – UMR 6072, Université de CaenCaenFrance

Personalised recommendations