Performance Evaluation of Secure Association Rule Mining Method over Distributed Databases Using FDM, Secure Scalar Dot Product, Data Perturbation and RSA Techniques

Association patterns generated from association rule mining (ARM) process may contain sensitive data and revealing it will cause potential privacy threats. This problem evolves even poorer if ARM is performed on distributed databases. There exists many ways by which sensitive data leakage can be handled and each method employs its own style to gain privacy, and each method is having its own pros and cons. We perform the performance analysis of secure association rule mining process, here in this paper, using four different strategies, viz. Fast Distributed Mining method, Secure Scalar Dot Product method, Data Perturbation method and Fast Distributed Mining with RSA algorithm. From experimental results, it is clear that Fast Distributed Mining of association rules along with cryptographic technique viz. RSA algorithm is flexible and ensures privacy on different global support counts.