摘要： 根据浏览历史对用户进行有效聚类，建立基于用户聚类的用户浏览行为预测模型是Web环境下实现个性化服务的关键。该文对系统用户进行聚类，产生相似用户群，根据每个相似用户群的浏览特征，建立基于相似用户群的类Markov链用户浏览行为预测模型，实验验证了该模型的有效性。

关键词: 浏览序列, 用户聚类, Markov链

Abstract: In Web environment, clustering Web users based on the browsing behaviors and building user browsing sequences prediction model based on user cluster are keys to achieve the personalized services. This paper produces similar user groups through clustering Web users. According to browsing features of each similar user group, the classified Markov chains based on the different similar user groups are built. Experimental result shows the efficiency of the model.

Key words: browsing sequence, user clustering, Markov chain

中图分类号: 

• TP301