摘要
根据Web用户的浏览历史建立用户浏览预测模型是Web环境下实现个性化服务和开发各种浏览导航工具的关键。该文首先利用PLSA模型对Web用户进行用户聚类,然后建立基于用户类别的混合Markov链用户浏览预测模型,该模型更能准确地描述用户浏览特征。实验结果表明了该模型的优越性。
In Web environment, according to Web users' navigation history, building user browsing prediction model are keys to achieve the personalized services and develop all kinds of browsing navigation tools. In the paper, at first we produce similar user groups based on PLSA model, then according to different user category, user browsing prediction model based on mixture Markov chains is built, this model can more accurately describe user browsing characteristics. Experimental result shows the superiority of the model.
出处
《数字技术与应用》
2016年第4期54-56,共3页
Digital Technology & Application