A Novel Association Rule Hiding Approach in OLAP Data Cubes

Data mining services require exact input data for their outcomes to be significant, but privacy concerns may influence users to provide fake information. We study here, with respect to mining association rules, whether or not users can be confident to provide correct information by ensuring that the mining process cannot, with any reasonable degree of certainty, breach their privacy. A data warehouse stores current and historical records consolidated from multiple transactional systems. Protecting data warehouses is of rising interest, particularly in view of areas where data are sold in pieces to third parties for data mining studies. In this case, current normal data warehouse security techniques, like data access control, may not be easy to impose and can be in effective. As an alternative, this paper proposes a data perturbation based approach, to provide privacy preserving in association rule mining on data cubes in a data warehouse. In order to conceal association rules and save the utility of transactions in data cubes, we select Genetic Algorithm to find optimum state of modification. In our approach various hiding styles are applied in different multi-objective fitness functions. To cope with the multi-objective functions, Pareto-front ranking strategy has been applied for obtaining the non-dominated solutions front. First objective of these functions is hiding sensitive rules and the second one is keeping the accuracy of transactions in data cube. After sanitization process we test the sanitization performance by evaluation of various criterions. The major feature is that the proposed strategy does not affect the functionality of the On-Line Analytical Processing system. Finally our experimental results show its effectiveness and feasibility.