Affiliations: School of Computing, Leeds University, LS2 9JT, UK.
[email protected], http://www.comp.leeds.ac.uk/vania/
Abstract: There is a strong argument in Artificial Intelligence in Education
which advocates that computer-based learning systems need to adapt to the needs
of learners if they are to provide for effective personalised instruction
(Self, 1999a). Diagnosing a learner's cognitive capacity is a crucial issue in
building adaptive systems. We have explored an interactive open learner
modelling (IOLM) approach which conceives diagnosis as an interactive process
involving both a computer system and a learner that discuss and together
construct the learner model. This paper outlines the architecture of an
interactive open learner modelling system and illustrates the method in a
terminological domain. We discuss an evaluative study of an IOLM demonstrator
– a system called STyLE-OLM. The results from the study demonstrate potential
benefits of the method for improving the quality of the learner model and
providing a means for fostering reflective thinking. We argue that IOLM is a
fruitful approach which may be employed in intelligent learning environments
both for obtaining a better model of a learner's cognitive state and
engaging learners in reflective activities.