An efficient technique to detect and prevent Sybil attacks in social network applications

Sybil attack is an attack where malicious users obtain multiple fake identities and access the system from multiple different modes. It is an attack wherein a reputation system is destroyed by falsifying identities in peer to peer networks. Communication between the users of networks only requires the users to be part of the same network. All kinds of distributed systems are capable of being wounded to Sybil attacks. An attacker can easily create a number of duplicate identities (called as Sybil) to impure the system with fake information and affect the exact performance of the system. In this paper, We propose an algorithm to improve the efficiency of blocking a Sybil attack by combining neighborhood similarity method and improved Knowledge Discovery tree based algorithm. This algorithm is proposed to block sybil attacks in social websites like Facebook, Twitter.