The rapid development of Internet technologies is promoting e-commerce. Mobile agents are mobile, personalized, autonomous, and adaptive. These qualities make them useful for information-rich and communication-rich environments such as e-commerce. Online shopping market commerce sites have two main drawbacks: 1. Because of the different product data format in the database and representation, it is difficult to exchange information between two online markets. 2.Consumers must search and filter product information by browsing a lot of shopping sites and compare product prices by themselves. 3. It's hard to accumulate consumer loyalty. Therefore, the aim is to extend the e-commerce platform developed by our agent-based e-commerce research group and build an agent-bated consumer recommendation mechanism. On behalf of consumers, agents can trade on the e-commerce platform, record consumer preference and produce appropriate product recommendation information according to the consumer's preference.