中文摘要：

传统电子商务推荐系统采用离线方式分析和处理客户需求信息,因而丧失了时效性。为解决相关问题,提出了基于需求线索的推荐系统。给出了系统的基本模型、数据结构、关键算法以及运行流程。该系统采用需求线索跟踪的方式收集客户当前的需求信息,从而进行需求倾向分析;并在需求VS商品叠加空间中匹配和检索需要推荐的商品与客户。在线营销仿真实验表明,系统具有较好的及时响应能力和较高的客户满意度。

英文摘要：

Conventional e-commerce recommendation system analyses and processes customer requirements offline which lacks promptness.To deal with the problem,a requirement hint based recommendation system is proposed,followed by its basic models,data structures,essential algorithms,as well as processing flows.The system gathers customers＇ contemporary requirements with the requirement hint tracing approach for requirement trend analysis,then match and search for commodities and customers from the requirement vs.commodity overlay space to recommend.Online shopping simulation results show that the novel system not only speeds up responding time but also satisfies customers better.