Teresa del Soldato
The implementation of motivational theories in tutoring systems requires the insertion of a motivational planner into the system's teaching expertise. This thesis describes an instructional planner able to make decisions (about the next task to do, whether to provide hints, etc.) in order to achieve two goals: traversing the domain - domain-based planning - and maintaining the learner's optimal motivational state - motivational planning. The twin activities of "detecting the state" and "reacting appropriately" are extended in the current work to take account of motivational factors, adding a motivational state and motivational planning to the traditional ITS architecture. In order to clarify the behaviour of the system, the teaching primitives were mapped to an example domain (Prolog debugging) and a basic tutor was designed and implemented including both motivational and domain-based instructional planners and mechanisms to negotiate between these two goals. The tutor offers buggy Prolog programs to be tested and corrected by the student.
This paper is not available online