SWARM OPTIMIZER BASED ON THE SENSITIVE RULE HIDING WITH THE CONSTRAINTS MINIMIZATION FOR THE DATA PUBLISHING

Main Article Content

P.Tamil Selvan
Dr.S. Veni

Abstract

Recently, motivating the demand for the privacy and secure data mining research is the expansion of techniques that include the privacy and security along with the effective data publishing. Most of the research work is developed for the data distribution with the privacy. However, the protocols used in the homomorphic encryption which increased the computational costs and communication. In order to overcome the limitations, a Swarm optimization and Iterative Privacy Rule Preservation (SIPRP) method is designed in the paper to improve the efficiency of the privacy preserving association rule mining with the constraint minimization. Initially, SIPRP method generates the association rules for the privacy preserving distribution database based on the support and confidence threshold. Then, the sensitive rules associated with the optimal sensitive items which are hidden are evaluated. After that, the sensitive rules are subjected to the Particle Swarm Optimization (PSO) for hiding and preserving highly confidential privacy rules. The constraints arise on preserving the high confidential privacy rules which are minimized by the iterative generation rules for the different sensitive sets of items. Finally, the SIPRP method obtains the sensitive sets of items for generating the specific sensitive. It is hidden with the less effect on the privacy being exposed during the data distribution across the multiple users. The Proposed SIPRP method uses adult data sets from the University of California’s Irvine data repository for conducting the experimental work. Experimental evaluation of the SIPRP method is done with the performance metrics such as the number of sensitive rules, processing time, number of hidden rules, and the rate of privacy. Experimental analysis shows that the SIPRP method is able to improve the privacy rate by 10.5% and also increases the number of hidden rule generated by 26.5 % when compared to the state-of-the-art works.

 


Keywords: computational cost, association rule mining, sensitive item sets, sensitive rules, principle component analysis

Downloads

Download data is not yet available.

Article Details

Section
Articles