Abstract
During the past decades, many methods have been developedfor the creation of Knowledge-Based Systems (KBS).For these methods, probabilistic networks have shown to bean important tool to work with probability-measured uncertainty.However, quality of probabilistic networks dependson a correct knowledge acquisition and modelation.KAMET is a model-based methodology designed tomanage knowledge acquisition from multiple knowledgesources that leads to a graphical model thatrepresents causal relations. Up to now, all inference methodsdeveloped for these models are rule-based, andtherefore eliminate most of the probabilistic information.We present a way to combine the benefits of Bayesiannetworks and KAMET, and reduce their problems. Toachieve this, we show a transformation that generates directedacyclic graphs, the basic structure of Bayesian networks, and conditional probability tables, from KAMET models. Thus, inference methods for probabilistic networksmay be used in KAMET models.