基于K-MEANS聚类的电商网站用户行为分析

Analysis of E-commerce Website User Behavior Based on K-MEANS Cluster

调整网站访问日志数据,从中提取用户访问各类页面次数,考虑类别化的变量组合方式,采用K-MEANS聚类对类别化的变量进行聚类分析,理解各类别特征,描述用户行为,分析各类别与输出结果的关联性,并为制定网站经营策略提供支持和参考依据.实证研究表明,对页面访问次数占比进行K-MEANS聚类分析,可以明确各类型页面与输出结果之间的关联性.

The paper probes into the behavior analysis of E-commerce website users through the adjustment of website access log data and extraction of the number from all kinds of website pages users visit.Considering the classified category of variable combinations,the categorical variables for cluster analysis is applied via the use of variation of K-MEANS cluster in order to comprehend the characteristics of each category,describe the user behaviors,analyze the correlation between each category and output results and provide the support and the reference frame for the website operating strategies.Empirical studies show that the K-MEANS clustering analysis on the proportion of page access times can make clear the correlation between each type of pages and the output results.