Selective Markov Models for Predicting Web-Page Accesses

Mukund Deshpande and George Karypis
ACM Transactions on Internet Technology, Vol. 4, Issue 2, pp. 163 - 184, 2004
Download Paper
The problem of predicting a user’s behavior on a Web site has gained importance due to the rapid
growth of the World Wide Web and the need to personalize and influence a user’s browsing experience.
Markov models and their variations have been found to be well suited for addressing
this problem. Of the different variations of Markov models, it is generally found that higher-order
Markov models display high predictive accuracies on Web sessions that they can predict. However,
higher-order models are also extremely complex due to their large number of states, which
increases their space and run-time requirements. In this article, we present different techniques
for intelligently selecting parts of different order Markov models so that the resulting model has a
reduced state complexity, while maintaining a high predictive accuracy.
This is an expanded version of the SDM01 paper.
Research topics: Bioinformatics | Classification | Data mining