Privacy preserving optimized rules mining from decision tables and decision trees

Abstract

With the swelling amount of data mining projects and information sharing, preserving privacy is a challenging issue which needs higher priority to guarantee security. This paper is an extension of Qureshi et al. (2010) with the addition of security features to the existing classification rules to preserve privacy. Experimental results show secure, accurate, sanitized and optimized rules which we achieve by using the concept of genetic algorithm (GA). © Indian Society for Education and Environment (iSee).