Some Issues in the Learning of Accurate, Interpretable User Models From Sparse Data

We discuss issues that arise when applying techniques for the learning of Bayesian networks in the user modeling context. We address the problem of sparse data that is often present in user modeling and show how we try to cope with it by introducing available a-priori knowledge into the learning procedures. Particularly, we present initial results concerning the learning of the structural part of a Bayesian network user model.