基于购买次数的用户群体聚类分析  

Clustering analysis of user groups based on the number of purchases

作　　者：顾强[1] Gu Qiang(China Mobile Group Jiangsu Co.,Ltd.,Jiangsu,Nanjing 21000,China)

机构地区：[1]中国移动通信集团江苏有限公司,江苏南京210000

出　　处：《计算机时代》2022年第9期83-86,91,共5页Computer Era

基　　金：2021年企业数据中心开发项目(B22302310CBG001);2020年机器学习应用软件项目(R20302310231106)。

摘　　要：为了对线上用户群体进行精准细分,提出一种基于购买次数的用户群体聚类分析方法,可以显示每个用户的群体划分以及不同群体用户的购买偏好。提出聚类分析算法的总体设计,详述了通过优化Keams++方法得到聚类中心,再进行聚类分析的具体实现过程。仿真实验结果表明,该聚类分析算法的正确性、有效性均可得到验证,且该算法的求解性能较好,可以应用于商城管理系统的用户群体聚类分析中。In order to accurately segment the online customer group,a user group clustering analysis method based on the number of purchases is proposed,which can display the group division of each user and the purchase preferences of different groups of users.The overall architecture of the clustering analysis algorithm is proposed,then the cluster centers obtained by optimizing the Keams++method are described in detail,and then the specific implementation process of the cluster analysis is performed.The simulation experiment results show that the correctness and effectiveness of the clustering analysis algorithm can be verified,and the algorithm has a good solution performance,which can be applied to the user group clustering analysis of the mall management system.

关 键 词：聚类算法 聚类中心 购买次数 用户细分 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]