Murray, R. C., VanLehn, K., & Mostow, J. (2001). A decision-theoretic approach for selecting tutorial discourse actions. In E. Horvitz, T. Paek, & C. Thompson (Eds.), Proceedings of the NAACL Workshop on Adaptation in Dialogue Systems (pp. 41-48). New Brunswick, NJ: Association for Computational Linguistics.

We propose a decision-theoretic architecture for selecting tutorial discourse actions.  DT Tutor, an action selection engine which embodies our approach, uses a dynamic decision network to consider the tutor's objectives and uncertain beliefs in adapting to and managing the changing tutorial state.  It predicts the effects of the tutor's discourse actions on the tutorial state, including the student's internal state, and appears to be unique in explicitly predicting outcomes and their utilities, DT Tutor selects the tutorial action with maximum expected utility.  We illustrate our approach with prototype applications for diverse domains: calculus problem-solving and elementary reading.  Feasibility evaluations assess DT Tutor's ability to select optimal actions quickly enough to keep the student engaged.

For a PDF full article version, click here (761KB).