摘要
随着电子商务的发展,以"客户为中心"已成为电子商务企业的经营策略,而任何高效的客户关系管理都是以扎实的客户分类为基础。然而电子商务中所搜集到的客户信息往往具有海量、高维度和不完备等特点,如何对其正确、高效地分类是一个难题。根据电子商务客户信息的特点,构建B2C客户分类模型,提出了先对客户信息进行主成分分析以消除属性之间的依赖性,而后用朴素贝叶斯算法进行分类的新方法。实验表明了该方法的有效性。
With the development of e-commerce,"customer-eentred" has become the operating strategy of e-commerce companies,and any efficient customer relationship management is based on solid customer classification. However, customers' information collected in e-commerce is always huge,muhi-dimension and incomplete, how to classify the information properly and efficiently is a hard problem. According to the characteristics of e-connneree customers,a model of B2C customer elassification was established in the paper. A new method was put forward as well,in it the principal components analysis on customer information is performed first for removing dependence between the attributes,and then the customers are classified with ha? ve Bayes algorithm. Experiments illustrate the efficiency of the method.
出处
《计算机应用与软件》
CSCD
2009年第6期72-74,95,共4页
Computer Applications and Software
基金
重庆市自然科学基金重点项目(2008BA2017)
关键词
主成分分析
电子商务
B2C
客户分类
朴素贝叶斯
Principal components analysis (PCA) E-commerce Business to customer(B2C) Customer classification Naive Bayes