Management of Experience Data for Rapid Adaptation to New Preferences Based on Bayesian Significance Evaluation

In a teaching and learning environment Bayesian network fits well because it can adjust its structure as per data presented to it. When a Bayesian network learns with a huge number of data, its belief value is updated even if the change in belief is not significant. This causes a problem when the user’s preference changes over time. The learning process cannot catch up rapidly enough to handle a new user preference. This problem is addressed in this work.