LRDM: Local Record-Driving Mechanism for Big Data Privacy Preservation in Social Networks

Abstract

The popularity of big data has resulted in the pro-liferation of applications and services with a lot of conveniences in social networks. Unfortunately, the diverse data sets contain users' privacy-sensitive information, which brings about privacy concerns potentially if the information is released or shared to the service provider. The existing practical techniques for big data privacy protection aims at anonymizing data via generalisation to satisfy a given privacy model. However, most widely-adopted privacy-preserving schemes ignore the privacy of the original data, which are generated from the initial organization or individual. In this paper, we formulated a general architecture of big data privacy to illustrate the basic structure of big data privacy. Specifically, we designed a Local Record-Driving Mechanism (LRDM) to achieve privacy-protecting for big data, which contains a methodology for organization or individual to find a privacy-preserving scheme. At the same time, we proposed a privacy metric to measure their privacy degree in big data scenario. Finally, thorough analysis and evaluation results show the effectiveness and efficiency of our proposed strategies and solution.