Title: Enhanced t-closeness for balancing utility and privacy

Authors: Korra Sathya Babu; Rajesh Pillelli; Sanjay Kumar Jena

Addresses: Department of Computer Science and Engineering, National Institute of Technology Rourkela, Rourkela 769008, India ' Department of Computer Science and Engineering, National Institute of Technology Rourkela, Rourkela 769008, India ' Department of Computer Science and Engineering, National Institute of Technology Rourkela, Rourkela 769008, India

Abstract: Driven by mutual benefits or guidelines, the collected data from various sources like hospitals, government agencies and private corporations are published in the internet. Publishing data without anonymising violates individual privacy, even if they are onymous or pseudonymous. A model that works on preserving privacy is t-closeness. This model overcomes attribute disclosure but is vulnerable to identity disclosure. The distance metric used in t-closeness does not satisfy all distance metrics like probability scaling. This article discusses two issues. First, on improving the utility of the data by incorporating Bhattacharya distance metric and secondly, to overcome the identity disclosure issue of t-closeness with a proposed algorithm. In the first issue the result show that utility of the dataset is improved without degrading the privacy. The result for second issue shows that the discernibility metric cost for proposed method is better than (n, t)-closeness.

Keywords: anonymity; data publishing; privacy preservation; privacy protection; t-closeness; data utility; identity disclosure; security.

DOI: 10.1504/IJTMCC.2013.056427

International Journal of Trust Management in Computing and Communications, 2013 Vol.1 No.3/4, pp.230 - 242

Received: 02 Dec 2012
Accepted: 21 Dec 2012

Published online: 12 Jul 2014 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article