A Semantic Framework for Content-Based Access Controls

Abstract

Social networking sites provide role-or group-based access controls to help users specify their privacy settings. However, information posted on these sites is often intentionally or unintentionally leaked and has caused harm or distress to users. In this paper, we investigate possible improvements to existing implementations by introducing content-based access control policies using Linked Data. Users are able to specify the type of content in the form of tags or keywords in order to indicate which information they wish to protect from certain roles (for example employment), groups or individuals. Providing all possible keywords matching a specific topic may be too time consuming and prone to error for users. Hence using Linked Data we enrich the provided keywords by identifying other meaningful and related concepts. This paper presents the implementation and challenges of developing such a semantic framework. We have qualitatively evaluated this framework using 23 participants. Feedback from participants suggests that such a framework will help ease privacy concerns while posting and sharing social network content.