Efficient applications of expert systems to project risk management problemsare seldom, if not unusual. In this paper we overcome this lack by using the probabilisticexpert system shell SPIRIT. The rule-based shell’s power in conditioning, inference andreasoning under incomplete information...