Treillis Relationnel : Une Structure Algébrique pour le Data Mining Multi-Dimensionnel

Constrained multidimensional patterns differ from the well known frequent patterns from a conceptual and logical points of view because they are structured and support various types of constraints. Classical data mining techniques are based on the power set lattice and, even extended, are not suitable when addressing the discovery of multidimensional patterns. In this paper we propose a foundation for various multidimensional data mining problems by introducing a new algebraic structure called relational lattice which characterizes the search space to be explored. We take into consideration monotone and/or anti-monotone constraints enforced when mining multidimensional patterns. In addition, we propose range representations of the constrained relational lattice which is a convex space and present an algorithm for computing them. Finally, we place emphasis on advantages of the relational lattice when compared to the power set lattice in multidimensional data mining. MOTS-CLÉS : Data mining multi-dimensionnel, Théorie des bases de données, Treillis.