Privacy-Preserving Access Control Mechanism for Social Statistical Data considering Accuracy Constrained

Abstract

In the current era of digitization, data volume is increasing rapidly. In order to store and manage these data for different purposes, there exist various standards and policies. Since data storing and managing is a tedious task, some organization may adopt data outsourcing to third party in order to reduce their load and to concentrate on quality of services. It becomes a privacy breach of users when third party cannot be trusted. In order to ensure privacy of users, different encryption approaches are used such as k-anonymity, l-diversity, variance diversity etc. There are two basic approaches to ensure the privacy of users. First is encrypt the whole database before storing at third party side and whenever needed import, decrypt and use them. Second approach is data fragmentation, in which whole data is divided in to n blocks and stored in different databases. We can use different fragments of different database at same time in order to process a query. Both approaches have their own pros and cons. Besides this, access control policy also plays vital role in order to ensure the privacy of user. We can achieve privacy using k-anonymity, l-diversity or variance diversity at the cost of data imprecision. Higher the level of imprecision, lower the accuracy will be. In this paper we are concerning about user privacy especially data stored on social media sites like facebook, twitter and others. Personal information of users is still susceptible to privacy breach, although they have their own privacy protection mechanism and access control policies with limitations of their own. In this paper we are providing privacy preserving access control mechanism for social statistical data considering accuracy constrained. It ensures the privacy of user data and provides data imprecision with more accuracy. Keywords- Access control policy, Privacy protection mechanism, k-anonymity, l-diversity