摘要
提出一种过程完整的针对消费数据挖掘的客户细分新方法。设计了包含3种类型10个指标的客户细分模型,并采用因子分析法从中提取细分变量,再使用基于划分的聚类算法进行客户细分。通过对某大型纸巾生产企业100万销售数据的计算分析,得出了有效客户类别,表明了本方法具有更强的客户细分能力和客户行为特征的解释能力。
This paper put forward a new multi-indicator RFM （ recency, frequency, monetary） method for customer segmenta- tion by consuming data mining. Firstly it constructed a muhi-indicator RFM model containing 10 indicators in 3 groups, and extracted inherent segment variables from these indicators by factor analysis method, then applied partitioning based clustering algorithm for customer segmentation. Through analysis of the sales data of a large Chinese paper towel producer with 1 million ＋ data records, it effectively got the customer types. The use case demonstrates that the method has higher detection rate and stronger interpretation ability on customer behavior.
出处
《计算机应用研究》
CSCD
北大核心
2013年第10期2944-2947,共4页
Application Research of Computers
基金
国家社科基金资助项目(13BGL063)
关键词
客户细分
消费行为
数据挖掘
聚类
customer segmentation
consuming behavior
data mining
clustering
