Extracting Student Models for Intelligent Tutoring Systems

John C. Stamper, Tiffany Barnes, Marvin J. Croy

Abstract: Intelligent Tutoring Systems (ITSs) that adapt to an individual student’s needs have been shown to be extremely effective, showing significant improvement in achievement over non-adaptive instruction (Murray 1999). The most successful of these systems require the construction of complex cognitive models that are applicable only to a specific tutorial in a specific field, requiring the time of experts to create and test these models on students. In order to achieve the benefits that ITSs provide, we must find a way to simplify their creation. Therefore, we are creating a framework to automate the generation of ITS student models. The goal is to provide a simple way to allow developers of computer-based training (CBT) to add adaptive capabilities with minimal work while still maintaining the effectiveness of a true ITS.