SBIR-STTR Award

Icai systems contain content domain knowledge and teaching knowledge. * both content domain and teaching experts are needed to build a complete icai system. Two separate knowledge acquisition processes must be performed -- one for content, one for teaching strategies. In this project, we systematically capture and represent the teaching knowledge in an instructional diagnostic module. The teaching - strategies represented in this module are content-independent. The captured teaching strategies are based on a well-researched educational theory of instruction that has been applied in numerous computer-based education lessions. Because content domain and teaching knowledge are separately represented, either knowledge base can be updated or improved independently. This allows the icai system to be improved as the content domain changes, or teaching theory evolves.

During phase i of this research, gist built and demonstrated an intelligent tutoring system prototype that implemented mastery instruction in simulation training. Domain and instructional knowledge were represented separately in the system; they communicated through a generic interface called an instructional calculus model. An expert teacher model implemented direct instruction strategies (engelman & carnine, 1982) that have been used in numerous computerbased lessons. During phase ii of this research, we will analyze actual simulations that have been designed for a high-technology weapon system. We will design and implement an expert teacher model that teaches the simulations to mastery. This will allow weapon systems operators to practice and maintain their skill levels in the field.