基于情境偏好知识超图的移动用户细分研究  

The Model of Mobile Customer Segmentation Based on Hypergraph of Context Preference Knowledge

作　　者：胡慕海[1] 蔡淑琴[1] 

机构地区：[1]华中科技大学管理学院

出　　处：《管理学报》2011年第10期1509-1516,共8页Chinese Journal of Management

基　　金：国家自然科学基金资助项目(71071066);国家自然科学基金资助重点项目(70731001)

摘　　要：构建了一个通用的用户细分模型,定义了用户情境偏好的概念和度量方式,采用情境偏好知识超图描述了用户间基于情境偏好相似性的多元弱关联,并基于超图分割实现用户细分,避免了一般聚类操作中用户间弱关联的丢失。应用该模型进一步构建了移动服务策略矩阵。最后通过实证验证了模型在用户细分和服务策略分析、制定上的适用性。A general model of user segment model is constructed, which includes the definition of user context preferences feature , the construction of user context preference knowledge hypergraph describing multiple weak association between users and the segment operation by hypergraph partitioning to avoid the lose of weak association in general clustering techniques. Furthermore, a construction and application of mobile services strategy matrix based on the model is explored. Finally the model is verified and its practical value for the planning of marketing strategy is discussed.

关 键 词：情境 超图 营销预测 移动服务 移动用户细分 

分 类 号：C93[经济管理—管理学] F713