面向在线交易日志的用户购买行为可视化分析

发布时间：2018-05-14 08:13

本文选题：可视化分析 + 用户购买行为　；参考：《软件学报》2017年09期

【摘要】：在线交易日志,即用户通过电商平台购买商品产生的交易记录,包括用户、商品、交易及商家的相关信息,反映了用户的购买行为.现有的可视化方法未能充分结合在线交易日志的时序、层次、地理、多维等特征,实现对用户购买行为的多角度分析.对此,结合交易日志的多个特征,提出了基于径向布局的复合时序可视化方法和融合空间信息的时间轴可视化方法,设计了颜色极值映射方法和规律映射方法;并基于上述方法,设计并实现了面向在线交易日志的用户购买行为可视化分析系统UPB-VIS,从而完成了单个用户和用户群体购买行为的全方位分析;最后,通过在京东商城在线交易日志数据集上的实验,证明了系统的易用性和相关可视化方法的有效性.
[Abstract]:The online transaction log (ITL), which is the transaction record generated by users buying goods through e-commerce platform, including users, commodities, transactions and related information of merchants, reflects the purchase behavior of users. The existing visualization methods fail to fully integrate the temporal, hierarchical, geographical and multidimensional features of the online transaction log to achieve multi-angle analysis of the user's purchase behavior. Based on the multiple features of the transaction log, the combined temporal visualization method based on radial layout and the time axis visualization method based on spatial information are proposed, and the color extremum mapping method and the law mapping method are designed. Based on the above methods, a visual analysis system UPB-VIS-oriented to the online transaction log (ITL) is designed and implemented to analyze the purchasing behavior of individual users and user groups. Through the experiment on the online transaction log data set of JingDong Mall, it is proved that the system is easy to use and the related visualization method is effective.
【作者单位】：四川大学计算机学院(软件学院);RMIT
【基金】：四川省科技支撑计划(2013GZ0015) 四川省学术带头人培养基金(2013DTPY0010)~~
【分类号】：F713.55;F724.6;TP311.13

【相似文献】