A privacy settings recommender system for Online Social Networks

There is a rapid growth in the Online Social Networks (OSNs) in recent years. Privacy settings in OSNs provide its users an option to control their online data sharing but managing the privacy settings is a confusing and a time consuming task and hence there is a need for a system that could measure and compare the privacy settings ofthe target users and help them to customize their privacy settings. In this paper we have proposed a context based personalized privacy settings recommender system. We have used homophily to group the target user´s friends according to a context (context based) and collaborative filtering mechanism to quantify the user´s profile privacy to provide meaningful recommendations with respect to their friend list (personalized). We have validated our solution using the data extracted from Facebook for the objects like the photos, videos, notes and links shared by target users and their friends.