个性化需求预测面临的挑战及解决方法研究  

Challenges and Methods of Individual Demands Forecasting

作　　者：唐中君[1] 陈晓红[1] 

机构地区：[1]中南大学商学院,湖南长沙410083

出　　处：《预测》2009年第5期48-54,共7页Forecasting

基　　金：国家自然科学基金资助项目(70772043);中南大学博士后基金资助项目

摘　　要：本文首先运用计划行为理论和信息交换理论分析得出,个性化需求预测在获取预测所用信息阶段的主要挑战是"如何帮助顾客将其拥有的与需求相关的隐性信息转换成显性信息",并提出了应对该挑战的方法"开发需求相关隐性信息转换系统";然后运用"自我推理"和"错误一致性效应"等理论分析得出,在理解需求相关信息阶段的主要挑战是"制造商在理解顾客需求相关信息时,如何避免自我推理和错误一致性效应",基于认知图作用的分析,提出了应对该挑战的方法"充分运用描述顾客价值系统的认知图理解顾客需求相关信息";将顾客需求相关信息转换成预测信息阶段的主要挑战是"如何提高预测精度",运用锚定-调整过程理论、信念更新范式理论、状态依赖性理论和记忆可接近性优势理论分析得出,制造商采取以顾客前一次购买订单所含信息为起始点的净变预测方式实现基于属性的预测,将有助于提高个性化需求预测的预测精度。This research focuses on uncovering challenges to realise individual demand forecasting and on proposing methods to address the challenges uncovered by using existing theories and research. To share demand-related information, one challenge may be faced, which is whether consumers have abilities to present their demand-related tacit information. To aid consumers to promote the abilities, an exogenously provided tool named demand-related tacit information transformation system is proposed. To understand the information shared, a critical challenge is how to avoid self-refer- encing and false consensus effect. An application of the theory of cognitive map shows that cognitive map may aid personnel of a manufacturer to avoid the self-referencing and false consensus effect. To translate the information understood into forecasts, how to improve forecasting performance becomes more serious in the context of individual demand forecas- ting than traditional forecasting methodologies. To improve forecasting performance, two methods of attribute-based forecasting and net change forecasting and feeding back forecasts to consumers are advanced.

关 键 词：个性化需求预测 即时顾客化定制 定制-响应困境问题 

分 类 号：F273[经济管理—企业管理]