Handling and analysis of fake multimedia contents threats with collective intelligence in P2P file sharing environments

In this paper, we discuss the question of removing fake multimedia content files, intentionally manipulated by a group of attackers. By employing the un-biased collective intelligence of participating P2P (peer-to-peer) nodes, we identify and remove fake multimedia content files based on the reputation management. The proposed scheme determines the reputation value according to the trustworthiness (along with confidence) of multimedia content files, which are statistically drawn by collectively relating the decision making of individual peers about each multimedia content file. To verify this, we simulate the detection and recovery of the proposed reputation management that employs K-means and LBG clustering algorithms over colluded attackers.