Privacy Preserving Outsourcing for Frequent Itemset Mining

Cloud computing uses the paradigm of data mining-as-a-service. A company/store lacking in mining expertise can outsource its mining needs to a service provider (server). The item-set of the outsourced database are the private property of the data owner. To protect this corporate privacy, the data owner encrypts the data and sends to the server. Based on the mining queries sent from client side, server conducts data mining and sends encrypted pattern to the client. To get true pattern client decrypts encrypted pattern. In this paper we have studied the problem of outsourcing the frequent item-set within corporate privacy preserving framework. We have proposed an attack model based on the basic assumption, attacker knows items and support of the item, he may know the details of encryption algorithm and some pairs of item and corresponding cipher values. Based on this basic assumption we have improved the security of the system, to reduce the item and item-set based attack, and to reduce the processing time.