论文部分内容阅读
通过对Web日志文件的挖掘,可以发现Web用户在Web上的访问行为和隐藏在Web日志记录中的规律,并发现不同Web用户在Web访问行为上的相似性,以及他们在Web上的访问偏好和消费习惯等知识。该文阐述了用户浏览行为的URL—UserID关联矩阵表示方法,建立了基于URL—UserID关联矩阵的Web页面聚类模型和Web用户聚类模型,为电子商务和其它基于Web信息服务系统的市场营销、客户保持、潜在客户发现和个性化服务等策略的制定提供了科学决策的依据。
Through the mining of Web log files, we can find the behavior of Web users on the Web and the hidden rules in Web log records, and find out the similarities of Web users in the Web access behaviors and their preferences on the Web And spending habits and other knowledge. This paper describes the URL-UserID association matrix representation of the user browsing behavior, and establishes a Web page clustering model and a Web user clustering model based on the URL-UserID correlation matrix for the marketing of e-commerce and other Web-based information service systems , Customer retention, potential customer discovery and personalized service strategy to provide a basis for scientific decision-making.