通过SOFM神经网络能根据数据的相似性对e-供应链中海量的客户数据进行客观的、科学的聚类分析,将客户划分成不同群类,并针对这些群类的不同特征和客户的重要性,采取有针对性的营销策略,不仅能提高客户的满意度,而且能实现e-供应链效益最大化.采用了改进的SOFM神经网络进行e-供应链中的客户聚类分析,运行效果较好. SOFM neural network can be based on the similarity of data e-supply chain massive customer data objective and scientific cluster analysis, the customer is divided into different groups, and for the different characteristics of these groups and the importance of customers And take a targeted marketing strategy, not only can improve customer satisfaction, but also to maximize e-supply chain efficiency.Using the improved SOFM neural network for e-supply chain customer cluster analysis, the operating results better.