Web浏览预测的Markov模型综述

Web Navigation Prediction Based on Markov Model—A Survey

Web访问模式挖掘研究的一个重要议题是Web浏览预测,Markov模型是一种经典的Web浏览预测模型。本文首先介绍了基本Markov浏览预测模型,包括基本Markov浏览行为模型,模型的学习训练及其在Web浏览预测问题中的应用;然后重点分析了扩展的Markov浏览预测模型,包括一序组合预测模型、高序模型、混合模型、隐Mark-ov模型、连续时间Markov模型等,综述了各种扩展模型所考虑的浏览预测问题的本质出发点、模型的学习方法及预测方法,最后分析了Markov浏览预测模型有待进一步研究的问题。

Web navigation prediction is one of the most important topics for discussion in research area of Web naviga tion pattern mining. Markov is one kind of traditional Web navigation prediction model. This paper first introduces basic Markov Web navigation prediction models, which include basic Markov Model of navigation behaviors, its training method and its application in Web navigation prediction problem. Then several extended Markov Web navigation prediction models are introduced, which include one-order combined prediction models, higher order models, mixture Markov models, hidden Markov models, continuous-time Markov models and so on. Then the essences of each kind of extended models and their learning and predicting methods are summarized. Finally some problems of Markov Web navigation prediction models are pointed out for further research.