Research Of Trust Mechanism For E-commerce Transactions Behavior

The Internet and Web carries rich information across time and space, and provides a dynamic,open and convenient environment for the development of social and business. However, when tradingwith strangers in the virtualized network, it is a challenge to make sure the transaction is reliable, safe,and of good quality. In order to solve this problem, this thesis researches the trust mechanism ofe-commerce transactions behavior. Through the establishment of a platform of mutual understandingfor consumers and businesses, it is able to respond to the dynamics needs of personal or business，reduce operating costs，improve services and promote the transactions online.The main contributions of the thesis are as follows:First, this thesis briefly introduces the basic concepts of trust in e-commerce, and builds atheoretical model. The trust value calculation is divided into subjective and objective trust value, andthe proportion of the two is difficult to define. Therefore a dynamic adaptive trust mechanism isproposed in this thesis to dynamically adjust their weights, which overcomes the one-sidedness of thecommonly used subjective judgment. This thesis distinguishes the concepts of presenter trust valueand the credibility, and the adaptive adjustment recommends credibility, which makes the trust modelhave better scientific and practical applications.Second, this thesis analyzes the influence of context information for the assessment of trust, andbased on this it puts forward a recommendation trust algorithm using context information. Thealgorithm makes use of the context information of time, place, and relationship of social network andso on to reflect affecting the current transactions between users through the historical ones, whichimproves the accuracy of the recommended. The higher the similarity degree of the background andtime between the users is, the better the correlation of them will be, and also the higher the value ofinformation recommended. That avoids a huge amount of calculation and at the same time makes theevaluation results more accurate.Finally, this thesis compares and analyzes this trust mechanism model with typical trust modelthrough simulation experiments. It results that the model this thesis proposed has a good ability toadapt to the dynamics of the network, and effectively curb the irrational behaviors of malicious nodes.So it will be better used in the e-commerce network.