Mining OLAP cubes: semantic links based on frequent itemsets

This paper proposes a new approach for rules discovering relationships between OLAP cube cells. Such relationships are based on frequent itemsets computation, which was always used in the literature as a first step for association rules generation. They define semantic links between cells of the OLAP cube. An algorithm to pragmatically extracting knowledge from data warehouses is proposed. Results of this algorithm should be shown in the graphical representation of the OLAP cube. Association rules discovery was successfully used for mining huge databases.