Improved privacy-preserving using elliptic curve diffie hellman and k-anonymity methodology

Abstract

Within a corporate privacy-preserving system, the issue of outsourcing the association rule mining task is taken into account. Therefore, privacy-preserving data mining is a research topic that is concerned with the security established from personally identifiable information when taken into account for data mining. We create a secure comparison technique and an effective homomorphic encryption scheme to guarantee data privacy. then suggests a frequent item set mining approach supported by the cloud that is used to create an association rule mining approach. The solutions are made for external databases that let several data owners communicate their information securely and productively without sacrificing data privacy. Compared to the majority of other solutions, the solutions leak less information about the raw data. The Rob Frugal encryption method is suggested as a solution to the security issue created by authorised users using external datasets. This system's proposed Rob algorithm incorporates fictitious patterns in cyphers for objects stored in the database. The false patterns contained in the outsourced data may increase the capacity overhead. We include weighted support in the original support of items to address this issue by lowering the number of false patterns and raising the security level for outsourced data with less complexity. To reduce storage requirements, the fictitious transaction table data is transformed into a matrix format. Outsourced data is more secure in the proposed work because process attacks based on items and item sets are not feasible.