面向在线交易日志的用户购买行为可视化分析

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，本文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.

Visual Analysis of Consumer Purchasing Behavior for Online Transaction Log

Online transaction log is a set of commodity trading records generated by Electronic Commerce (E-Commerce) platform. It incorporates information of the consumers, commodities, sellers and transactions that reflect consumer purchasing behavior. The existing visualization methods can not fully combine the time series, hierarchical, geospatial and multi-dimensional features of online transaction log to analyse consumer purchasing behavior in muti-aspact. Combined with mutiple features of online transaction log, we propose a composite temporal visualization method based on the radial layout and a timeline visualization method incorporated with spatial information. Also we design extreme color mapping method and identifiable color mapping method to support our analysis. UPB-VIS is designed and implemented based on the methods above to realize the comprehensive analysis of consumer purchasing behavior. We prove the usability of the system and the validity of the visualization methods by using JD online transaction log.