A Study on Trust of E-commerce Market Based on Multi-agents Model

The process of the trust production and transfer is both stochastic and rational. In this article, we analyze the characters of this process in consumer’s social networks, by dividing consumers and merchants into two types: Risk-aversion type and Risk-neutral type for consumers, high-grade and low-grade for merchants respectively. Based on multi-agent NETLogo system, three kinds of virtual social networks are constructed, which are the risk-neutral without institution, risk-aversion without institution, risk-aversion but with good institution. The rules of the production and transfer of trust in consumer-to-consumer and merchant-to-consumers are defined. Based on a series of computational experiments and statistical comparison, we draw a conclusion that the characters of consumers do not change the trend of e-business market scale, but only change the equilibrium value. On the other hand, a risk-aversion with good institutional society can boost market to have a highly upward trend and be self-stable. Consequentially, this article indirectly proved of the essentials for establishing the institution-based trust.