A Framework for Building Intelligent Learning Environments in Ill-defined Domains

A critical problem of instruction in ill-structured and complex domains has been how to help students attain a deep understanding of a complex concept. Solutions for this problem are usually very costly. For example, the practical course in medical education often requires one-to-one assistance of the expertteacher for the student to be able to account for a great diversity of complex and real clinical situations. Intelligent learning environments could provide significant help for instruction in ill-defined domains. In this paper, we show how to exploit advanced technologies such as temporal Bayesian networks and computer-based simulations to help the student in advanced learning of complex concepts such as the sacro-iliac screw fixation in orthopedic surgery.