Most research in social networks has focused on the assumption that unknown entities are malicious and thus the traditional approach was to detect them and deny their access to sensitive data. In this paper, we propose a new computational model that helps users predict security risks associated with their information sharing on social networks. The model is based on the assumption that a risk indicator value can be predicted by assessing a number of risk attributes using a neuro-fuzzy technique. A disclosure decision is made based on this risk indicator value. The approach was tested in a real prototype of a social mobile service at a university campus. Further, we show how the model can be implemented in a popular social rating site. Results obtained show the relevance and effectiveness of the proposed approach in predicting risks and in deciding up on it about disclosure decisions in social pervasive applications.
- Neuro-fuzzy systems
- Security risks
- Social Network Services (SNS)
- Social networks
- Social pervasive applications