基于Web日志挖掘的网络用户行为交互式可视化分析  

Interactive visualization analysis of user network behavior based on Web log mining

作　　者：刘翼[1] 高明 田亮亮 严都力 韦雨萌 李凯茵 张博欣 LIU Yi;Gao Ming;TIAN Liangliang;YAN Duli;WEI Yumeng;LI Kaiyin;ZHANG Boxin(College of Mathematics and Computer Science,Yan’an University,Yan’an 716000,China)

机构地区：[1]延安大学数学与计算机科学学院,陕西延安716000

出　　处：《延安大学学报（自然科学版）》2023年第3期78-85,共8页Journal of Yan'an University：Natural Science Edition

基　　金：国家自然科学基金项目(61962059);陕西省教育厅自然科学专项(22JK0622);延安市科技计划项目(2022SLGYGG-007);陕西省大学生创新创业计划项目(S202010719023,S202010719099);延安大学大学生创新创业计划项目(D2020045)。

摘　　要：高校在校大学生用户群体的网络行为分析是一个有着普遍需求和意义的重要课题,具有丰富的现实意义与社会价值。本文提出网络用户行为交互式可视化分析方法,研究分析高校大学生用户群体网络行为的特征和兴趣。首先获取某高校真实的网络用户访问日志;其次通过清洗缺省值、去重、规范化等数据预处理方法,选取日期、时间、用户账号和访问地址作为特征属性,并在数据集中标记出每个URL的主题分类和每个用户所属的学院等属性;然后从用户、主题分类、时序3个维度分析得出这些属性的分布情况,发现网络用户行为的区别和联系;最后利用ECharts交互式可视化工具展示分析结果。实验结果反映出当前大学生用户的网络行为特征及兴趣喜好,揭示了大学生用户的网络行为模式,为教学管理者的决策提供了数据支持。The network behavior analysis is an important subject,both in terms of practical significance and social value.In this paper,an interactive visualization analysis method of network user behavior based on Web usage mining was proposed to explore college students' network behavior characteristics and interesting.This method first captured the user's web access log form a real network of a university.Secondly,data preprocessing methods such as cleaning default values,deduplication,and normalization were used for raw data.At the same time,date,time,user account and access address as characteristic attributes were selected,along with the topic classification of each URL and the user's college in the dataset.Then,the distribution of these attributes from three dimensions:user,topic classification,and time series were marked to explore the differences and connections in user behavior on Internet.Finally,the ECharts were used to interactively visualize the analysis results.The experimental results reflect the network behavior characteristics and interests of current college students,as well as reveal their network behavior patterns,and provide data support for teaching managers' decision-making.

关 键 词：交互式可视化分析 网页日志挖掘 网络用户行为 主题分类 高校学生 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]