Building knowledge based tutors

Abstract

We have described existing tools for representing, acquiring, and reasoning about tutoring knowledge. Tutoring knowledge has been described both in terms of content (i.e., topics, questions, and examples) and in terms of context (i.e., tutoring strategies and discourse interventions). Current tools provide a generic framework which allows development of multiple tutors. Several application systems have been described, namely those in statics and thermodynamics, along with a few tools, namely TUPITS, DACTNs, and the Response Matrix. The tools and applications provide a test-bed for exploring new issues and testing new tutoring functionality. Ultimately, we expect to build systems which can reason about the choice of tutoring strategy based on a clear representation of a student's cognitive knowledge. No extant system yet has this capability.

This work was supported in part by a grant from the National Science Foundation, Materials Development Research, No. 8751362. It was also supported in part by the Air Force Systems Command, Rome Air Development Center, Griffiss AFB, New York, 13441 and the Air Force Office of Scientific Research, Boiling AFB, DC 20332 under contract No. F30602-85-0008 which supported the Northeast Artificial Intelligence Consortium (NAIC). Partial support was also received from University Research Initiative Contract No. N00014-86-K-0764.

A longer version of this paper appeared in Intelligent Tutoring Systems: Evolutions in Design, H. Burns, J. Parlett and C. Redfield (Eds.), Lawrence Erlbaum: NJ., 1991.