Unlike the systems already discussed, KN-AHS is an adaptive hypertext browsing system, rather than a tutoring system. Student modelling facilities are provided via the BGP-MS user modelling shell system (Kobsa 1990). This is an off-the-shelf package which uses stereotypes, arranged into a hierarchy. The stereotypes implemented within KN-AHS are an ‘any person’ stereotype, which holds general information, and more specialised stereotypes such as ‘hypertext user’ or ‘PC user’.
The initial stereotype or types for a user are selected on the basis of an initial computer based interview. The user model is subsequently updated according to a set of heuristics, on the basis of selections the user makes within the hypertext. For example, if a user deselects an item then familiarity is assumed. If an explanation or definition is requested then unfamiliarity is assumed. The adaptation is KN-AHS is presentational, and occurs as a new page is prepared for presentation. If the student model indicates unfamiliarity with the associated concept, an explanation is inserted, whereas if the user is believed to be familiar with the concept then additional information can be provided.
Authored by Serengul Smith
E-mail to:
serengul1@mdx.ac.uk
School of Computing Science Middlesex University
Revised: September 1998