Building Dialog Systems for Tutorial Applications:
Papers from the AAAI Fall Symposium
Carolyn Penstein Rosé and Reva Freedman, Cochairs
Studies of human tutoring have argued the importance of conversation between the student and the tutor in making tutoring interactions successful, suggesting that intelligent tutoring systems will be more effective if they can engage in dialogues with students. Although building highly interactive dialogue-based systems presents a wide range of new computational challenges, recent advances in computational linguistics have made it possible to make significant strides towards the development of dialogue-based tutors in both educational and indus-trial settings. Already several large-scale efforts towards the construction of such systems are in progress at a variety of universities worldwide, many of which will be represented at the symposium. The overall focus of this symposium will be the design, implementation, and evaluation of effective dialogue-based intelligent tutoring systems (ITSs). The working notes include papers on architectures for dialogue-based ITSs, approaches to dialogue management, robust language understanding tech-niques, practical approaches to natural language generation, authoring tools, student modeling, analysis of human tutorial dialogues, and system evaluation.