Sybil Belief: A Semi- Creation New Approach for Structure –Based Sybil Detection

Sybil attacks are a fundamental threat to the security of distributed system. There has been a growing interest in leveraging social network to mitigate Sybil attacks. We introduce Sybil belief a semi supervised learning framework to detect Sybil nodes. Sybil Belief takes a social network of the nodes in the system, a small set of known benign nodes, and, optionally, a small set of known Sybil’s as input. We show that Sybil Belief is able to accurately identify Sybil nodes with low false positive rates and low false negative rates. Sybil Belief is resilient to noise in our prior knowledge about known benign and Sybil nodes. Sybil accounts in online social networks are used for criminal activities such as spreading spam or malware stealing other users’ private information and manipulating web search results. Sybil defenses require users to present trusted identities issued by certification authorities. However, such approaches violate the open nature that underlies the success of these distributed systems.