DRTEMBB: Dynamic recommendation trust evaluation model based on bidding

Fraud or cheating of malicious nodes were restrained by reducing the trust value of malicious nodes in the past trust model, but few recommendation nodes for cooperation cheating are punished, so it has had only limited malicious recommendation or fraudulent action again and again. Due to lack of punishment for malicious recommendation of recommendation nodes, enormous malicious recommendation nodes remain and search for next criminal opportunity. By borrowing idea of bidding and the introduction of competition and the mechanism of rewards and penalty, this paper proposes a dynamic recommendation trust evaluated model based on bidding in E-Commerce environment. By greatly increasing the "criminal cost" of malicious recommendation nodes, with recommendation optimization algorithm based on Markov chain, the model ensures that the recommendation service is true and more objective, which will stimulate the enthusiasm of nodes objective recommendation and reach the goal that nodes give up malicious recommendation and cooperative cheating on its own initiative by objective method so as to restrain malicious recommendation and cooperative cheating effectively. In simulation experiment, the model also shows good restraint on malicious recommendation.