An Improved Approach to High Level Privacy Preserving Itemset Mining

Abstract

Privacy preserving association rule mining has triggered the development of many privacy preserving data mining techniques. A large fraction of them use randomized data distortion techniques to mask the data for preserving. This paper proposes a new transaction randomization method which is a combination of the fake transaction randomization method and a new per transaction randomization method. This method distorts the items within each transaction and ensures a higher level of data privacy in comparison to the previous approaches. The pertransaction randomization method involves a randomization function to replace the item by a random number guarantying privacy within the transaction also. A tool has also been developed to implement the proposed approach to mine frequent itemsets and association rules from the data guaranteeing the antimonotonic property.