一种新颖的电子商务模式:高效消费者响应系统  

A Novel Electronic Commerce Mode:Efficient Consumer Response System

作　　者：杨辉华[1,2] 王行愚[1] 

机构地区：[1]华东理工大学信息学院,上海200237 [2]桂林电子工业学院计算机系

出　　处：《计算机工程与应用》2004年第19期23-28,161,共7页Computer Engineering and Applications

基　　金：国家重点基础研究发展规划项目(编号:2002CB312200);教育部科学技术研究重点项目(编号:00053);国家自然科学基金项目(编号:69974014)

摘　　要：ECR系统是降低快速消费品行业物流成本的有效途径。该文简要分析了ECR系统实现的关键及难点:以著名学者T.Poggio和S.Smale提出的学习理论的一个关键算法(Akeyalgorithm)为基础,进一步提出一种在预测中集成行业背景知识与高阶推理的方法,成功解决销售滚动预测问题;提出了ECR系统的最优化发货策略及其实现方法。基于上述方案,以某大型乳业公司和某大型连锁超市之间业务关系为背景,构建了一种无纸、自动化的新型电子商务模式,能够有效降低物流成本。Efficient Consumer Response(ECR)system is on e of the most effective approaches to cut off the logistics costs in the field of fast moving consumer goods.The paper elaborates the difficulties and key poi nts in actualizing E-CR.A machine learning algorithm recently named″A Key Alg orithm″by T.Poggio and S.Smale is used to tackle the tough problem of forecas ting the rolling daily sales of each stock keeping unit,and a method to integra te corresponding background information and high order reasoning is presented fo r the purpose of boosting prediction performance.More-over,in order to achiev e highest profit in ECR,the optimal goods delivery policy is discussed and adop ted.Based on the scheme presented in the paper,an ECR system is actualized ba sed on the business relationship between a large dairy corporation and another s upermarket chain,and as a result,a no-paper,automatic and novel electronic c ommerce mode come s to truth,and it is proved that the logistics cost can be re duced significantly.

关 键 词：高效消费者响应 关键学习算法 销售预测 供应商管理库存 电子商务数据挖掘 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]