Privacy Preserving Association Rules based on Compression and Cryptography (PPAR-CC)

Abstract

Privacy-Preserving Data Mining (PPDM) is a modern technique through which data is mined while maintaining the confidentiality and privacy of sensitive information from unauthorized persons. The Privacy-Preserving Association Rules Mining (PPARM) is the most important technique for privacy-preserving data mining. PPARM means the mining of association rules with preserving the non-disclosure of sensitive correlations among items or features for competitors or the public, especially data of sensitive organizations such as financial organizations and others. In this paper, we propose an approach to hiding association rules after performing the mining process and obtaining knowledge through vertical and horizontal compressing then encoded the compressing form by using cryptography methods. The proposed approach is resistant to many known attacks and is undetectable because it includes three stages of compression and encryption in which the basic representation and size of the data change dramatically. The proposed approach significantly reduces storage space, maintains knowledge security, reduces transmission time, and facilitates the transmission of knowledge over any network.