The subject of intelligent tutoring draws on skills from the fields of artificial intelligence, psychology, education and learning. A recent trend has also seen intelligent tutoring systems embracing hypertext as a presentation medium.
There are many benefits and also problems associated with the use of hypertext. A solution proposed for these problems is hypertext adaptation. Hypertext adaptation moulds a hyperspace to the needs of a specific user, much as an intelligent tutoring system will adapt a course of instruction to a particular student.
In order to tutor effectively an intelligent tutoring system must understand the current skill levels and cognitive state of the student it is tutoring. This can only be based on an analysis of a students interaction with the system. Within a hypertext system the range of interactions that can be analysed is largely limited to an analysis of a user’s browsing patterns. By comparison with intelligent tutoring systems there has been relatively little research conducted in this area.
This task is essentially one of classification which is an area where machine learning algorithms have traditionally been successful and a number of machine learning algorithms and techniques that have been used in intelligent tutoring systems
Authored by Serengul Smith
E-mail to:
serengul1@mdx.ac.uk
School of Computing Science Middlesex University
Revised: September 1998