摘要： 提出了一种新的基于智能Agent的用户兴趣发现和更新方法，主要介绍了智能Agent需要跟踪用户的哪些行为，如何发现用户真正感兴趣的文献，如何通过分析提取兴趣特征词及权重计算等内容。该方法主要利用了用户浏览文献页面的信息，有效地解决了用户兴趣的自适应变化。

关键词: 智能Agent, 用户模型, 兴趣特征词

Abstract: A method based on intelligent agent to discover and update user's interest is described in this paper. It mainly introduces what behaviors need to be tracked, how to find user's indeed interested literature and how to pick up interested words and compute relative weight. This method effectively solves the selfadapting changes of user's interest by using information of viewing literature.

中图分类号: 

• TP393.03