A Dynamic Trust Algorithm For C2C E-Commerce Environment

In C2C (Consumer to Consumer) e-commerce systems, the transactions areanonymous and dynamic, the partners lack of the basic trust foundations, so there existsrisks in the process of the transactions. It is one of the efficient ways to reduce thetransaction risk that the trust algorithm is used to evaluate sellers’ trustworthiness and helpbuyers make decision-buying.In order to compute trust quantitatively and objectively, some concepts of trust andtrust relationships are defined in terms of human cognitive methods and behaviors insociety, and a formal description model is raised based on the trust network. Consideringthe reputation propagation discipline, optimization algorithms for the trust network are putforward which can effectively simplify the complex network relationship. And a visualgeneration tool is developed for the trust network, which can reduce the complexity of thetrust computing algorithm, and build the theoretical principle for studying on trustpropagation mechanism and trust evaluation model.Then, the dynamic trust algorithm CDTA (C2C Dynamic Trust Algorithm) isconstructed for C2C e-commerce environment in detail. The algorithm first computes thedirect trustworthiness of the buyer to the seller by the transaction experience between thebuyer and the seller. Secondly, the reference trustworthiness is computed from the buyer’sfriends in the trust network. Finally, the trust of the buyer to the seller is the integration ofthe direct trustworthiness and the reference trustworthiness by means of the trust adjustingfactor. In order to reduce the attack from malicious act and make the trustworthiness moreaccurately, a trust-factor is defined in the algorithm.The experiments show that both the transaction multi-attributes and their relativitiesare taken into account in CDTA， the granularity of the trust evaluation is morefine-grained and the evaluation result is more objective. On the other hand, the trustcorrelation coefficient can help the buyer sift out the reference nodes meeting with thebuyer’s individuation, make the reference trustworthiness more accurately, and curb the impact on the CDTA from some malicious nodes.