Secure and Distributed Approach for Mining Association Rules

Abstract

Data mining is the process of extracting trends from data sources. Domain exerts can make use of the trends to derive business intelligence. Big organizations store data in multiple server and often data is horizontally distributed. Mining such database provides useful and actionable knowledge which can help in making well informed decisions. However, secure mining of extracting association rules can provide interesting information that can help enterprises to make expert decisions. In this paper, we propose an algorithm and have a secure mechanism in order to mine association rules for deriving knowledge. We also incorporated auditing of data in the proposed system. We built a prototype application that demonstrates the secure mining of association rules with support and confidence. The statistical measures such as support and confidence help in knowing the usefulness of the rules. The empirical results are encouraging.