Knowledge-based tutors : an artificial intelligence approach to education.

A vehicle is suggested for bringing information technology into education. Knowledge-based systems are proposed as a way to explore, reason about, and synthesize large knowledge bases. These systems utilize resources such as artificial intelligence, multimedia, and electronic communication to reason about what, with whom, and how they should teach in order to tailor knowledge and communication to individual students. Teaching material does not consist of a repertoire of prespecified responses; rather, reasoning about the student and the complexity of the subject matter informs the system's response so that inferences made by the machine become key features of the system's response. Currently, such systems can reason about a student's presumed knowledge, can solve the problems given to the student, and can begin to recognize plausible student misconceptions. This document provides a practical hands-on guide for people who are considering building knowledge-based systems. It identifies the requisite resources, personnel, hardware and software and describes artificial intelligence methodologies and tools that might become available. The document is directed both at increased production of knowledge-based systems and also at improving the dialogue among computer scientists, educators, researchers, and classroom practitioners around the issue of information technology in the schools.