DefinePK

Abstract

Privacy preservation has become a major issue during data publishing. The simplest solution is not to disclose the information. This would be of no use since it will hinder the process of data analysis. Instead, the data can first be modified so that it can guarantee privacy and, at the same time, it retains sufficient utility and can be released to other parties safely. The existing system uses a hybrid approach of sampling and generalization for data modification. An alternative system using cross over and mutation of genetic algorithm for data modification, without the use of a trusted third party is proposed. In addition, shearing technique and double pass matrix based encryption technique are implemented to improve the confidentiality of the data.