Bayesian networks for user modeling: Predicting the user´s preferences

Bayesian networks tend to be increasingly used for the management of uncertainty in user modeling and provide a simple and effective approach for constructing and manipulating probabilistic models. In the Web context, user models used by such adaptive applications contain personal information such as knowledge, preferences, objectives which are required for learning personalized process. In this paper, we investigate and present user´s preferences in our Web interface using Bayesian inference so that the user model is able to infer user´s preferences.