earticle

영어

In order to predict network users’ access behavior accurately, this paper proposes a new prediction model based on Web access behavior. To improve prediction performance and reduce the state-space complexity, the model uses hybrid-order Markov chain structure and stores the sequences which share the same prefix. The problems that lower-order models have poor prediction performance and higher-order models result in high state-space complexity are solved in this paper. Simulation results have shown that the prediction model based on Web access behavior can improve the precision and recall to some extent.