Privacy settings recommender for Online Social Network

In recent years Relationship Based Access Control (ReBAC) has become the reference paradigm for controlled information sharing in Online Social Network (OSN) scenarios. Nevertheless, many of the most popular OSN providers do not implement in their platforms an access control model fully compliant with ReBAC. This fact, thus, limits the capability of OSN users to define customized and fine-grained access control policies. Moreover, average users might have difficulties in properly setting, potentially, complex access control policies. As results, many users give up in defining proper privacy setting, simply accepting the default setting proposed by OSN provider. To cope with this problem, we see the need of tools in support of policy specification. At this aim, in this paper we present a recommendation system that, exploiting an association rules mining process, learns OSN users' habits in releasing resources in online social networks, and exploit them to suggest customized access control policies. We also prove the feasibility of the presented techniques by illustrating an experiment which has been conducted on 30 human users by building customized access control policies from the data learnt from each of them.