Skip to Main Content
Data mining is to discover previously unknown, potentially useful and nontrivial knowledge, patterns or rules. Because databases may have some sensitive information which should not be leaked out, it is nontrivial to study data mining techniques without neglecting sensitive information, i.e., privacy-preserving data mining. In this paper, a new technique has been proposed for privacy-preserving mining of sequential patterns based on data perturbation. Experimental results show that the reconstructing support of frequent sequences can achieve a rather high level of accuracy.