Social commerce is a new generation of e-commerce. Compared to traditional commerce, social commerce, treating social media and social network as a carrier, promotes online trading and information exchange related to commercial activities. Information sharing in social commerce affects buying behavior of consumers in all aspects. Therefore, research on information sharing in social commerce has important implications for further investigation in consumer behavior and business marketing strategy. In this paper, the authors propose a method of multi-agent simulation based on FIRE model, to discuss the influence of sharing evaluation information on consumer buying behavior in social commerce. This study established a simulation model of consumer behavior based on the information, took full account of the effect given by situational factors and information sources, set up rules on consumer decision of purchase on the basis of consumer sub-model and merchant sub-model, realized the simulation for two different information sharing situation by using Repast, and drew relevant conclusions according to simulation results about the total number of transactions and trading volatility of each type of merchants.