一种改进的基于浏览行为的用户兴趣模型

摘要

This article mainly discussed and established a kind of improved user profile based on the user browsing behavior. In the user profile, we overall take into account the user browsing web pages behaviors, the level of interest to keywords, the user's short-term interests and long-term interests.We realize the WIPISES demo system under the. NET platform, the improved user profile is embedded in the system. The results of the simulation experiments indicate that the retrieval effects which use the improved user profile provided in this article surpass the general search engines. And with the web user's preference is stronger, the superiority of the WIPISES system is more obvious.%主要讨论并建立了一种改进的基于用户浏览行为的用户兴趣模型,模型中综合考虑了Web用户对页面的浏览行为、对页面关键词的感兴趣程度、用户的短期兴趣和长期兴趣.在.NET平台下实现了WIPISES演示系统,在系统中嵌入本文所提出的改进的用户兴趣模型.仿真实验结果表明:使用本文改进的用户兴趣模型进行信息检索,其搜索效果优于目前主流的搜索引擎,而且Web用户的偏好性越强,WIPISES系统的优势也就越显著.