Study on multi-attribute dynamic trust evaluation model based on network transaction

Aiming at the randomness and fuzziness of complex network environment, a multi attribute dynamic trust evaluation model for complex network transactions is proposed, including direct trust, recommendation trust and trust risk function, is used to evaluate trust relationships in complex network transactions. And the method of entropy weight is proposed to ensure the objectivity of the trust value calculation. And service quality is introduced into risk assessment calculation, including description of goods, service attitude and delivery speed. Through the electronic commerce experiment, it is proved that the model can effectively prevent credit speculation and periodic deception.