Grouping friends to improve privacy on Social Networking Sites

Social Networking Sites (SNS) such as Facebook and MySpace have attracted millions of users because of their ability to combine individuals by social graphs. Nonetheless, considerable amount of users are suffering from their poor privacy settings. It is partly due to their lack of the awareness of privacy. To most normal users, the too exhausting privacy settings on SNS is another important reason. Moreover, even friends cannot be trusted all the time, and the privacy issues involving friends have actually been serious. In this paper¹, we propose a novel approach of grouping friends to improve the privacy policy. We introduce a closeness degree model to quantify relationship between one user and his friends, based on which all the friends can be categorized automatically and dynamically. With some further statistical inference and analysis, our approach reflects the users´ tendency of disclosing personal data as well as their original privacy preferences.