OLAP recommender: supporting navigation in OLAP cubes using association rule mining

The OLAP Recommender tool automates the multidimensional data exploration process and recommends potentially interesting views on the data to the user. It integrates two data analytics methods – OLAP visualisation and data mining, the latter being represented by GUHA association rule mining. Algo-rithms implemented in the tool include automated data discretization, setup of dimensions’ commensurability, automatic design of the data mining task based on the data structure, and mapping between the mined association rules and the corresponding OLAP visualisation. The system was tested with real retail data and with EU structural funds data. The experiments indicate that complemen-tary usage of association rule mining and OLAP analysis identifies relationships in the data with higher success rate than the isolated use of both techniques