Semi-Supervised Policy Recommendation for Online Social Networks

Abstract

Fine grain policy settings in social network sites is becoming a very important requirement for managing user's privacy. Incorrect privacy policy settings can easily lead to leaks in private and personal information. At the same time, being too restrictive would reduce the benefits of online social networks. This is further complicated with the growing adoption of social networks and with the rapid growth in information uploading and sharing. The problem of facilitating policy settings has attracted numerous access control, and human computer interaction researchers. The solutions proposed range from usable interfaces for policy settings to automated policy settings. We propose a fine grained policy recommendation system that is based on an iterative semi-supervised learning approach that uses the social graph propagation properties. Active learning and social graph properties were used to detect the most informative instances to be labeled as training sets. We implemented and tested our approach using real Facebook dataset. We compared our proposed approach to supervised learning and random walk approaches. Our proposed approaches provided high accuracy and precision when compared to the other approaches.

The 'Retailio' Privacy Wizard: Assisting Users with Privacy Settings for Intelligent Retail Stores

“…There have been several approaches to overcome this user burden by automatizing privacy setting maintenance. One of the most commonly used techniques is to use machine learning and a subset of labeled friends to predict the privacy settings of the remaining users [17], [18], [6]. Fang and LeFevre [6] proposed a semi-supervised machine learning technique to infer privacy settings of a user's social network (SN) friends: The user is asked to label several of her friends on the SN with privacy privileges.…”

Fine-Grained Privacy Setting Prediction Using a Privacy Attitude Questionnaire and Machine Learning

“…There are also other systems that use machine learning for the prediction of privacy settings, for example by labeling some of the friends with privacy permissions and using a supervised learning approach [6,7,2]. Other approaches additionally take the post content into account, by using latent Dirichlet allocation (LDA) and maximum entropy to predict settings for a new post based on the privacy settings chosen in earlier posts [8].…”

Detecting Privacy Preferences from Online Social Footprints: A Literature Review

“…In particular, semi-supervised techniques, which have the advantage of utilizing fewer labeled data to achieve better predictions, can be a potential research avenue to explore further. A graph-based semi-supervised method has been proposed to effectively capture privacy preference (Shehab & Touati, 2012), and other methods such as Expectation and Maximization (EM), topic modeling, and co-training need to be investigated further. For instance, co-training has been successfully used to detect users' latent personal attributes in social networks (Mo, Wang, Li, Hong, & King, 2010).…”

“…In (Shehab & Touati, 2012), active learning and the properties of the social graph are first used to detect a set of the most informative contacts to be labeled as training samples. In the labeling process, the user specifies whether he/she is willing to share a specific data item with the selected contact.…”