摘要
基于顾客购物篮的分析提出一种新的基于虚拟变量的数据挖掘方法。该方法结合因子分析、聚类和关联规则挖掘技术。首先通过因子分析方法从众多的实变量中标识出影响顾客购买决定的少数几个虚拟变量,其次利用聚类分析对顾客划分成若干簇,最后运用关联规则分析获取每一簇中对象之间隐藏的模式。通过在一个零售业的案例中进行实施,表明了该方法的有效性。
This paper presents a new data mining method based on virtual variables in light of market basket analysis.This method consists of factor analysing,clustering and association rules mining.Firstly,people exploit factor analysis to identify a few effective virtual factors which influence customers＇ purchasing choices from many original variables.And then they utilise clustering analysis to divide the customers into clusters.Finally,they use association rules to acquire the hidden patterns in each cluster.A case study in a retailer store demonstrates the effectiveness of their proposed method.
出处
《计算机应用与软件》
CSCD
2011年第12期156-158,共3页
Computer Applications and Software
关键词
数据挖掘
购物篮分析
虚拟变量
关联规则
Data mining Market basket analysis Virtual variables Association rules
