论文标题:Web数据挖掘在电子商务中的应用研究
Application of Web Data Mining in E-Commerce
论文作者
论文导师 贾作皆；张兆臣,论文学位 硕士,论文专业 软件工程
论文单位 山东科技大学,点击次数 160,论文页数 73页File Size3140K
2006-05-01论文网 http://www.lw23.com/lunwen_17767587/
Data warehouse; data mining; web data mining; e-commerce
随着计算机和Internet应用的普及，信息化的进程不断加速，使传统的商务模式面临巨大的变革考验，电子商务得到迅速推广。作为一种与传统模式不同的交易方式，电子商务具有简便、成本低、信息丰富等诸多优点，更是一种被年轻人青睐的时尚购物方式。近年来，我国的电子商务也发展迅速，不仅受到政府和投资商的关注，也吸引了许多的研究者。 数据挖掘是当前研究的热点，其中Web数据挖掘在电子商务中的应用尤其引人注目。Web数据挖掘的目的就是从互联网上浩如烟海的半结构数据中，找到有用的信息和模式，从而为市场分析和决策提供有力的支持。Web数据挖掘的对象可以是Web页面内容、页面之间的结构、用户访问信息、电子商务信息在内的各种Web数据。 本文研究了Web数据挖掘的基础理论，探讨了其中的关键技术，对Web访问信息挖掘进行了深入细致的研究，并在以上工作的基础上设计了一个基于客户浏览行为分析的电子商务推荐系统的结构框架，将设计的挖掘实验应用于此框架，构成一个完整的推荐系统，并分析了该系统的功能，对用户的兴趣程度进行了矩阵量化和计算，在应用过程中根据已有算法的缺陷提出了一种新的聚类算法，并将其用于页面聚类分析，以帮助改进页面归类；对用户的行为进行算法分析；将关联规则挖掘技术运用到电子商务中进行用户访问模式分析，帮助电子商务的经营者改进网站的设计；对电子商务的用户进行分类，针对不同类型的用户进行不同的页面推荐，实现了电子商务的个性化服务。
With the popularization of computer and Internet, and the quick development of information, the traditional commercial models faces great challenges for change. E-commerce becomes popular rapidly. As a different way of transaction from the traditional models, E-commerce has the advantages of simple, low cost, and informative, favored by young people as a fashion shopping way. In the recent years, E-commerce developed rapidly too in our country, gaining attentions not only from government and investors, but also from researchers.Data mining is the hot topic in the current research, among which the application of Web data mining in E-commerce draws special attentions. The purpose of Web data mining is to find useful information and models from the semi-structured data which are so much in the internet, so to provide strong support For market analysis and decision making. The objects of Web data mining can be all kinds of Web data including Web page"s content, the structures between web pages, users" visiting information, and E-commerce information.This thesis researched on the basic theories of Web data mining, and explored its key techniques, by making thorough researches on Web visiting information mining. Furthermore, based on the above researches, a structure of E-commerce recommended system based on Customer browsing behavioral analysis has been designed, and two mining experiments designed were applied to this structure and formed a complete recommended system. And has analyzed this system function, has carried on the matrix quantification and the computation to user"s interest degree .In application, a new cluster algorithm was put Forward on the basis of the drawbacks of the existing algorithm, and used it in the web page cluster analysis to help improve page classification; applied association rules mining technique to E-commerce For model analysis of users visiting, helping managers of E-commerce improve the design of website; classified users of E-commerce recommended different web pages to different kinds of users, so as to personalize the service of E-commerce.
|
