ABSTRACT
This chapter describes an iterative methodology for developing these systems, along with the generic tutoring foundation. Development of an intelligent tutor, like development of any artificial intelligence (AI) system, requires several iterative cycles: Computer scientists and instructional designers first collaborate on the design and development of the system. AI programs require that a teaching expert define the knowledge to be used along with the control structures that define the way an interpreter will traverse that knowledge. Tutoring knowledge has been described both in terms of content and in terms of context. Several application systems have been described, namely those in statics, thermodynamics, and time management, along with a few tools, namely TUPITS, DACTNs, and ExGen. The network describes tutorial strategies in terms of a vocabulary of primitive discourse moves such as teach, motivate, contrast, and summarize. It is implemented in a language called TUPITS. DACTNs allow discourse control decisions to be based on a dynamic interpretation of the situation.
