Post-Mining of Generalized Association Rules from Data Cubes

Data warehouses provide decision makers with the necessary tools to help them understand their data. The structural complexity of the data is maintained through multidimensional data views commonly called data cubes. Mining such data is still an open problem as few approaches really take the specificities of this framework into account (e.g. multidimensionality, hierarchies, measures). In this paper, we introduce a novel approach that post-mines generalized association rules by exploiting the dimension hierarchies that feature data warehouses. We define the concepts related to our problem as well as the associated algorithm. Carried out experiments show the significance of our approach.