摘要
在基于Agent的个性化网页推荐中,目前主要有两种过滤方法:基于内容的过滤和基于多Agent合作的过滤。在分析了单独使用这两种方法存在的不足之后,提出了基于改进蚁群算法的聚类分析建立用户模型的算法,并给出了结合两种方法优点的用户模型主动学习算法,最后给出了个性化信息推荐模型及相关算法。
In personal information service field, there are two classical methods： content-based and collaborative-based information faltering method. After analyzing the shortcomings of each method, this paper presents the algorithm of constructing user models with cluster analysis based on parallel ant colony adaptive algorithm, and the algorithm of user models active study combining the advantages of the two methods. Lastly, it provides the model of personal informational recommendation, as well as the corresponding algorithm.
出处
《信息技术》
2006年第9期83-87,共5页
Information Technology
