Artificial Intelligence
Artificial intelligence and tutoring systems: computational and cognitive approaches to the communication of knowledge
Causal reasoning about quantities
Readings in qualitative reasoning about physical systems
Generating causal explanation from a cardio-vascular simulation
IJCAI'89 Proceedings of the 11th international joint conference on Artificial intelligence - Volume 1
| Hi-index | 0.00 |
Student modelling is not typically concerned with representing the deep mental models a student employs in dealing with the world around him/her. In this research we discuss an intelligent tutoring system, PRESTO, whose goal is to understand the mental model a student has of a physical device, and then use this mental model in providing help to overcome misunderstandings related to the functioning of the device. The mental model is extracted from the student by asking questions about the relationships of variables affecting the physics of the device. The mental model is represented using deKleer and Brown's qualitative confluence equations. The mental model can then be compared to a set of confluences representing a correct perspective on how the device functions. A variety of pedagogical choices can be made: to explain contradictions implicit in the student's understanding of the device, to show the student a simpler physical device that by analogy illustrates anomalies in the student's understanding, or to let the student witness his/her version of the device in operation so the misunderstandings become obvious. Experiments in running PRESTO with a number of students shows this approach to mental modelling to be promising.