Privacy Preserving Approach for Association Rule Mining in Horizontally Partitioned Data using MFI and Shamir’s Secret Sharing

Preserving privacy while collaboratively mining association rules has caught the attention of researchers due to privacy concern by collaborative participants. Existing public key based homomorphic encryption schemes achieve the privacy but increases the computation and the communication cost. This paper explores the non-public key based collision resistance technique called shamir’s secret sharing for preserving privacy for distributed association rule mining (PPDARM) in horizontal partitioned data. We use the concept of MFI (Maximal Frequent Itemset) to reduce the communication cost. The theoretical and experimental analysis of the proposed algorithm, which is conducted with a real dataset shows that it performs superlative with respect to communication and computation compared to existing PPDARM techniques. Additionally, the proposed algorithm is analyzed in terms of the security, privacy and correctness respectively.