Research On Key Problems Of Trust Evaluation In E-commerce Transaction Process

With the advantage of cross space,low cost and flexible,electronic commerce has quickly become a new way of economic growth.However,in the process of rapid development,electronic commerce trust crisis hinders electronic commerce development seriously.Analyzing the reason of trust problems,existing unreasonable trust evaluation mechanism has become a key reason for trust crisis.Weather the website trust evaluation or seller trust evaluation,the existing trust evaluation mechanism has some vulnerabilities and defects,with the performance of deceptive website threatens the safety of consumers,low fraud cost and poor reference of sellers'trust value to consumers.Therefore,how to help consumers establish trust on websites and sellers is needed to be solved urgently.From the perspective of transaction flow,this paper takes key problems of electronic commerce trust evaluation as the research subject,which are website trust evaluation and seller trust evaluation.Website trust evaluation is evaluating a website according to six factors,such as usefulness,customer service,credibility and safety,which can provide a theoretical reference for consumers to choose a trustworthy website and for regulators to rank the website trust value.The seller trust evaluation refers to the trust evaluation to a seller,combining transaction satisfaction score with transaction attributes to establish an evaluation model,which is an improvement of the existing trust evaluation mechanism and can effectively deal with some malice behavior such as malicious collusion evaluation and concussion trade,etc..Main contents of paper are as follows:(1)Firstly,based on the existing research,this paper gives a description of the key concepts such as electronic commerce trust.And from the perspective of transaction flow,this paper divides the E-commerce trust evaluation into two key problems,which are website trust evaluation and seller trust evaluation.Then this paper analyzes the influencing factors of two key problems.(2)For the website trust evaluation,firstly,this paper collects data and processes data.Secondly,this paper selects key evaluation factors by using Mean Impact Value.Then,compared with BP neural network,we find MIV-BP neural network model is more accurate.So this paper uses MIV-BP neural network model to evaluate website trust value.Finally,we collect website cases data into the MIV-BP neural network to evaluate website trust value and to evaluate the website trust levels.(3)For the seller trust evaluation,this paper takes C2C electronic commerce as an example.Firstly this paper constructs seller trust evaluation model based on transaction satisfaction,transaction time,transaction amount,reliability and penalty factor.Then this paper uses C#to achieve the algorithm and interface of sellers trust evaluation model.Finally,through collecting the data from a Taobao seller,this paper proves this model is more effective in dealing with malicious collusion evaluation,concussion trade and other trust crisis problems.This paper uses methods of data statistics,qualitative and measurable methods to have a research on key problems of electronic commerce trust evaluation,which are the website trust evaluation and seller trust evaluation.This paper contributes to electronic commerce trust evaluation research and provides some method reference for solving e-commerce market common trust issues,and the whole study has strong application value.