A framework to customize privacy settings of online social network users

Abstract

Privacy is one of the most important concerns in an online social network. Online social network data is big data as millions of users are a part of it. The personal information of every user in an online social network is an asset that may be traded by a third party for its benefits. Individuals should be aware of how much of their personal information could be shared without risk. Different people have different requirements to share a profile item hence measuring privacy of such huge and diverse population is a challenging and complicated task in itself. In this paper we have proposed a framework that ensures privacy of individuals by allowing them to measure their privacy with respect to some specific people of their choice rather than measuring it with the entire population on online social networks. We have suggested a method to choose the best model to fit the real world data and to calculate the sensitivities of various profile items. The framework gives specific labels to users that indicates their profile privacy strength and enable them to customize their privacy settings so as to improve the privacy quotient. The users can also act as advisers to their online friends whose privacy quotients are low and thus spread privacy awareness in social networks.