多Markov链用户浏览预测模型

Modeling User Navigation Sequences Based on Multi-Markov Chains

建立有效的用户浏览预测模型 ,对用户的浏览做出准确的预测 ,是开发各种浏览导航工具的关键 .传统的Markov链模型是一种简单而有效的预测模型 ,但它存在预测准确率低 ,存储复杂度高等缺点 .通过对该模型的扩展 ,该文提出并建立了一种基于用户分类的新模型———多Markov链模型 .实验表明 ,该模型能更准确地描述用户在Web上的浏览特征 ,在预测准确率和存储复杂度方面都显著地优于传统的Markov链模型 .

Modeling users' navigation in the Web is the key to build tools which can help user navigate the Web efficiently. The Markov chain is a simple and practical model, but it gives low prediction accuracy and requires high space complexity. In this paper, we propose an new approach to modeling user navigation sequences based on multi-Markov chains. This approach is shown to be superior to existing Markov chain based approaches. In particular, it is more accurate in making prediction and yet has lower space complexity.