摘要：In this paper, we apply two-stage fuzzy comprehensive evaluation of fuzzy theory to evaluate the recommendation trust vector of P2P network nodes, which fully reflects the effect of various factors on trust evaluation of the target node, at the same time, dynamically reflects the demand of different nodes in the different transaction context. So the trust evaluation result is more truthful and reliable. And applying fuzzy comprehensive evaluation to evaluate trust vector can better embody fuzziness and subjectivity of trust itself. Finally, experiment results show that the model can prevent malicious recommendation more effectively.
