Bayesian Belief Networks (BBNs) are compact and intuitive graphical models for supporting decision making under uncertainty. A BBN can be used as a knowledge integration tool to represent information from diverse sources. As such it can be used to compute posterior beliefs of evidence given observations on other events as this examples illustrates.

Every day in the morning you go to work by car. One day the car does not start. What is the most likely root cause of the car not starting?

The figure on the left shows the structure of a BBN that can be used to support reasoning about why the car does not start. The nodes represent variables and the directed edges connecting nodes describe probabilistic dependence relations between the variables.

Insert, for instance, the observation that the car is not starting and observe how the probabilities of the root causes changes. In addition, you may enter evidence that lights are working and that the fuel gauge shows out of fuel.

Below is a set of HUGIN widgets for interacting with the model (click on the probability bar to instantiate a node or remove evidence):