In order to investigate the formation of e-commerce market network, this paper describes an analytical framework from a complex network point of view which contains three steps-definition of network, analysis of network topology and analysis of network environment. Then, an innovative model for explaining the evolutionary process is proposed, with several original factorsgrowth-factor, select-order-factor, preferential attachment mechanism and global-local-factor. Our research reveals that the attraction mechanism impacts evolutionary trend and network structure to some extent, and also reveals that the global-local-factor and select-order-factor impact the evolutionary structure of the network, the smaller the probability, the smaller the concentration of networks and the more obvious the randomness are. In order to analyze the impact of edge-increasing mechanism on network evolutionary trend, a contrast test is designed with two models. The test reveals that the edge-increasing mechanism makes the network become a small world with a lower average distance and a higher clustering coefficient, and makes it be like a scale-free network with a lower power-law exponent and a higher centralization.