Foundations of Rule Learning