An Efficient Algorithm Research to Web User Access Prediction

In order to enhance accuracy of the algorithms on user access prediction. The prediction algorithm based on Markov chain and association rule(PAMA), and Markov prediction model with feedback(MPMF) were proposed. The PAMA integrates the advantage of Markov chain and association rule well. It corrects the Markov prediction result on forward and reverse perspective, and gets the last prediction page. The MPMF adjusts the prediction algorithm dynamically according to user feedback mechanism and history prediction information, and then gets the prediction page at last. Theoretical analysis proves these two prediction methods with linear time complexity. Experiments result shows that the accuracy of PAMA and MPMF is good, so the prediction efficiency is also meeting the requirement.