IJCTT-International Journal of Computer Trends and Technology
With the rapid development of online reputation systems, manipulations against such systems are evolving quickly. In this paper, the authors propose a cluster based approached theory to protect reputations. Tested against users attack data taken from a cyber-competition, the proposed system has achieved a better performance in terms of accurately identifying unwanted users. It also describes a great conceivable to effectively remove dishonest ratings and keeps the online reputation system a secure and fair marketplace. They collect all the information of the user and depend on their rating they get the dishonest users.