A Ontology-Based Semantic Reputation Evaluation Method in P2P Network

In view of some critical defects existing in P2P network such as stability and security issues, a novel approach called OSREMP is proposed to evaluate the reputation based on semantic ontology, the existing reputation models mostly predict the performance of peer based on transaction history records, lacking of comprehensive consideration on the credibility of resources and various malicious acts. We introduce trust ontology tree to understand the current peers´ resources and give a mathematical method of integration reputation based on recommendation reputation and content reputation, the two kinds of reputations based on domain ontology. Simulation experiments and analysis demonstrate that this research can distinguish malicious acts such as forged document, cooperative cheating and so on, OSREMP can enhance security and transaction successful rate of system.