Privacy Preserving Association Rules by Using Greedy Approach
Chieh-Ming Wu, Yin‐Fu Huang, Jianying Chen · 2009
Data mining techniques have been developed in many applications. However, they also cause a threat to privacy. In this paper, we proposed a greedy method for hiding the number of sensitive rules. The experimental results showed that the undesired side effects can be avoided in the rule hiding process by use of our approach. The results also revealed that in most cases, all the sensitive rules are hidden without generating spurious rules. First, the good scalability of our approach in terms of database sizes is achieved by using an efficient data structure FCET to store solely maximal frequent itemsets rather than the entire frequent itemsets. Furthermore, we proposed a new framework for enforcing the privacy in mining association rules, that combine the techniques for efficiently hiding sensitive rules and the transaction retrieval engine based on the FCET index tree. In particular, four strategies are implemented in the sanitized procedure, for hiding a group of association rules characterized as sensitive or artificial rules.