Research On Trust Multiple-evaluation Of C2C Platform E-commerce

With the further development of the Intemet,C2C platform e-commerce has gained a big following in the Chinese people for the convenience of shopping.But due to the virtuality of the network environment and the characteristics of acrossing time and space,the information become asymmetric.And the trust problems of partial businessman still can't be settled effectively,such as malicious fraud,vilified reputation.How to improve the detection ability of various participants'trust issues is the key to soluting the development of C2C platform e-commerce on the whole.Building a trust evaluation mechanism of multple perspctives based on the participants' characteristics is of great significance to guide the trade subject of C2C platform e-commerce to choose honesty trading partners and be in good faith transactions.Nowadays the trust evaluation mechanismused by many e-commerce is not comprehensive to consider the influencing factors of trust and it is not objective and authentic to reflect the status of trust.There are few articles can systematically consider factors that affect the trust evaluation of each participant based on the different types of C2C platform e-commerce of participants.The computation of trust value is only the direct trust between the individual which based on the past history of interaction data or probability bayesian method as little consideration of direct trust and recommendation trust.On the basis of existing research at home and abroad,this paper respectively study on the trust evaluation of buyers,sellers,e-commerce and logistics enterprises through the method of theoretical analysis and empirical research,which includes the common factors of direct trust for four types of participants,the particular factor based on their own behavior characteristics or qualities and the factors of recommendation trust based on the path problems,the degree of trust and credibility of social network under the condition of following rules.Finally,the dynamically weighted trust evaluation model of C2C platform e-commerce from multiple perspectives is forming in this paper.In this paper,MATLAB R2011b has been selected to the simulation experiment to set up an emulated environment.Observe the change of these four types of participants' direct trust value within a month.Take buyers for example,comparing the calculation errors of trust value on this model with the other three models in the case of the malicious fraud users of random and steady types on collusion and no collusion,which verify the direct trust model has the feasibility and effectiveness;On the condition of different proportion of malicious fraud,this paper uses the method of experimental control group to compare the identification accuracy of malicious fraud users,average path span and recommendation span,which verify the feasibility and effectiveness of the recommendation trust model;This paper use the simulation to detect malicious fraud of these four types of participants,which verify the trust evaluation model from mutiple perspwctives has the ability to fight for malicious fraud behaviors.It is helpful for trade subject to against the dishonest behavior of malicious trading partners.