DFM-IA:面向B2C电子商务的多源用户兴趣数据采集机制

发布时间：2018-07-25 17:26

【摘要】：用户兴趣模型是电子商务个性化推荐服务的基础,用户兴趣数据的获取则是构建用户兴趣模型的核心环节。传统的轮询采集方法存在数据源不够全面、在线应用可扩展性差的不足,会导致同一业务分析所需数据采集时间跨度大、先采集数据可能失效等情况,使得最终业务分析结果出现偏差。针对上述问题,对B2C电子商务用户兴趣数据进行了深入分析,提出了一种基于智能Agent的多源用户兴趣数据采集机制DFM-IA(Data Fetching Mechanism based on Intelligent Agent)。DFM-IA以用户Session为基本处理单元,设计了四种智能Agent(Fetching Agent、Watching Agent、Sort Agent、Logical Agent)和三条排序规则,对七类用户兴趣数据(浏览行为、关键词搜索、收藏行为、购物车行为、订单行为、支付行为、评价行为)进行排序与合并处理,从而在丰富数据采集源的同时大幅提高了在线数据采集效率,有助于解决推荐服务的数据稀疏性问题。仿真实验表明了该机制的高效性。
[Abstract]:User interest model is the basis of e-commerce personalized recommendation service, and the acquisition of user interest data is the core of building user interest model. The traditional polling data collection method is not comprehensive in data source and poor in the scalability of online application, which will lead to the long time span of data acquisition required for the same business analysis, and the possibility of failure of the first data acquisition, etc. The result of the final business analysis is biased. In order to solve the above problems, the user interest data of B2C e-commerce is analyzed in depth, and a multi-source user interest data acquisition mechanism based on intelligent Agent is proposed. DFM-IA (Data Fetching Mechanism based on Intelligent Agent). DFM-IA takes user Session as the basic processing unit. Four kinds of intelligent Agent (Fetching agents sort logical Agent) and three collation rules are designed to sort and merge seven kinds of user interest data (browsing behavior, keyword searching, collecting behavior, shopping cart behavior, order behavior, payment behavior, evaluation behavior). Thus, the efficiency of online data acquisition is greatly improved while enriching the data collection sources, and it is helpful to solve the problem of data sparsity of recommendation services. The simulation results show that the mechanism is efficient.
【作者单位】：四川师范大学计算机科学学院;匹兹堡大学Katz商学院;四川师范大学图书信息中心;合肥工业大学管理学院;
【基金】：国家自然科学基金资助项目(71202165);国家自然科学基金重点资助项目(71331002) 四川省哲学社会科学规划资助项目(SC13C019)
【分类号】：F713.36;TP391.3

【相似文献】