Web日志挖掘技术在尚品网个性化服务的应用研究

发布时间：2018-03-30 07:42

本文选题：数据挖掘　切入点：Web日志挖掘　出处：《天津大学》2014年硕士论文

【摘要】：如今,随着Web技术的不断发展,电子商务站点风起云涌,其已经成为商业领域内的新模式。电子商务网站,就技术角度来说,开发部署一个电子商务站点并不是什么难事,关键问题是如何让自己的电子商务站点能够获得更高的盈利,而Web挖掘技术正是针对这一问题,并能够有效的对其进行解决。Web日志挖掘通过对用户访问Web服务器时留在服务器上的Web日志进行分析,使用数据挖掘技术,发现隐藏在Web日志中的规律性知识,就可以得到用户的访问模式以及兴趣偏好。通过Web日志挖掘可以为构建个性化服务提供依据以及为电子商务的开发和为管理人员提供大量有利于对Web站点改进、带来经济效益的信息。本文给出了Web日志挖掘技术的应用框架。然后,以该框架以及Web日志挖掘的一般框架为基础,以理解客户为目标设计了电子商务Web日志挖掘的一种简易框架图,并把整个挖掘工作分为三个部分,即数据清理、模式挖掘以及模式分析,便于对实际应用的操作。由于Web日志挖掘的流程相对繁琐,特别是数据预处理工作。因此,本文以简化图为指导,采取了两种新的XML应用,即XGMML和LOGML,可以使预处理工作大大地简单化；以此为基础,对商品的客户购买模式进行挖掘,既包括客户的个体访问模式,也包括群体访问模式。这不仅有利于实现“一对一”的个性化服务要求,同时,也运用了分类技术对客户群体进行了划分,提升了商家的营销准确性。最后,以上述的访问模式的挖掘技术为前提,对Web挖掘模式在电子商务环境下的客户分类支持框架进行了详细地分析,以及对电子商务环境下的客户管理问题进行了重点讨论,以此来正确地制定企业营销策略,帮助企业吸引客户,提高盈利。
[Abstract]:Today, with the continuous development of Web technology, e-commerce sites have become a new model in the business field. In terms of technology, it is not difficult to develop and deploy an e-commerce site. The key question is how to make your e-commerce site more profitable, and Web mining technology is aimed at this problem. By analyzing the Web logs left on the server when the user visits the Web server, and using the data mining technology, the regular knowledge hidden in the Web log can be found. Through Web logging mining can provide a basis for building personalized services, e-commerce development, and a large number of Web site improvements for administrators. This paper gives the application framework of Web log mining technology. Then, based on this framework and the general framework of Web log mining, A simple framework diagram of Web log mining in electronic commerce is designed to understand customers, and the whole mining work is divided into three parts: data cleaning, pattern mining and pattern analysis. The flow of Web log mining is relatively cumbersome, especially the data preprocessing work. Therefore, two new XML applications are adopted under the guidance of simplified graph. In other words, XGMML and LOGML can greatly simplify the preprocessing work. On this basis, we can mine the customer purchase patterns of commodities, including the individual access patterns of the customers. This not only helps to realize the "one-to-one" personalized service requirements, but also uses the classification technology to divide the customer groups, which improves the marketing accuracy of the merchants. Based on the above access pattern mining technology, this paper analyzes the customer classification support framework of Web mining pattern in E-commerce environment in detail, and discusses the customer management problem in E-commerce environment. In order to correctly formulate enterprise marketing strategy, to help enterprises attract customers, improve profitability.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.09;TP311.13

【参考文献】