Secure Outsourcing Association Rule Mining in Horizontally and Vertically Partition Database Using Eclat and Double Encryption Technique

Abstract

Cloud computing uses the ideal model of information mining-as-an service, utilizing these it seems to be an obvious choice for companies saving on the cost of contributing to secure, manage and keep up an IT infrastructure. An association or company who is lacking in mining capacity can outsource its mining needs to third party service providers. Be that as it may, each the association rules and item-set of the outsourced database are seen as private property of the association (company). The data owner encrypt the data and sends to the server to protect the corporate security. Data owner or client transfers its mining queries to server, and afterward server conducts mining task & encrypt rules and sends generated association rules to the data owner or client. To get genuine pattern client decrypts the received rules. Paper focuses on the issue of outsourcing the rule mining task inside a corporate privacy preserving framework. It additionally shows the core idea of privacy preserving association rule mining on vertically partitioned data with utilization of enhanced cryptographic technique. The strategies incorporate cryptographic techniques to minimize the data shared, while adding minimal overhead to the mining task. This research tries to propose desirable algorithm for both vertically as well as horizontally partitioned data. A technique for solving a main problem of privacy preserving association rule mining in two party databases is proposed. To improve the performance of system horizontal partitioning as well as vertical partitioning of data is performed, also double encryption technique is used to increase the security of dataset which includes homomorphic encryption algorithm followed by asymmetric algorithm