Research On Agent-Swarm Based Simulation For Credit Optimization Analysis Of Cross-border E-commerce Transaction

With the rapid development of Internet technology, the trends in international trade activities through the network have become more prevalent, especially in2009after developing cross-border E-commerce in China. The emerging form of trade continues to rise and it has become an indispensable power in international trade. However,compared to traditional trade patterns, cross-border E-commerce has some deficiencies, due to the presence of virtual E-commerce market, and compared to traditional trade, its main trading information has a greater asymmetry, which makes the transaction subject more inclined to dishonesty in the course of trade. A greater asymmetry, which makes both parties more inclined to untrustworthy behavior in the course of trade. Since leaving marks mechanisms exist in the cross-border E-commerce market,buyers and sellers could constantly modify their own behavior through multiple dynamic repeated game and choose the trustworthy behavior. The presence of traces of the mechanism reduces the transaction cost when transaction subjects are choosing credit transaction behavior, thus making the dealers more willing to maintain their trustworthy behavior.The paper breaks the traditional modeling methods, using more popular methods of Simulation economics. The using of complex adaptive systems theory modeling ideas optimizes cross-border E-commerce transactions for credit research, CAS should be a bottom-up system of the body, it will be seen as a system that can adapt to the environment with the ability to learn the subject and bottom-up analysis of complex social behavior among subjects in the way. The paper uses the theory of Agent-Swarm based simulation analysis of cross-border E-commerce most influential factor in the development of cross-border E-commerce credit how traces of the credit mechanism to achieve the optimization. In this paper, we establish two models, namely static game model and dynamic game model.First,we make a theoretical formulation of trading behavior of the main electricity supplier based on the static and dynamic game, and then use Swarm simulation model for static and dynamic trading in case the main game chooses to model the behavior of credit,coming to the corresponding conclusions:when the information is not completely symmetric in static game model, the transaction entities prefer dishonesty, but because of cross-border E-commerce traces the role of the mechanism, making body established trading dynamic game model repeated modified, so that it constantly accumulates credit value, due to the behavior of credit-driving, transaction subjects are more inclined to choose trustworthy behavior, dishonest traders will be wiped out of the transaction. It also results in a credit which continues to optimize the process of trading twin prefer0-0-0(completely trustworthy behavior), reflecting traders " winner takes all" nature, in a way that reflects the lock theory.