(424)AUTOMATED LEARNING TOPIC REFINEMENT OF EXERCISESBASED ON LOGGING DATA

摘要

For exercise-based adaptive learning systems, it is crucial that the available exercises have detailedmetadata specifying the learning topic they train. These learning topics must be detailed to allow thesystem to guarantee that all different facets of the learning matter are trained in a learning session. Aswe noticed that many exercises in such systems have a too generic learning topic specification, wedesigned a low level methodology for the automated refinement of a learning topic into severallearning subtopics. More specifically, we start from a set of exercises on the same learning topic andgroup all exercises having a common learning subtopic. Our methodology uses the learning resultsfrom multiple learners that already solved the exercises involved. In this paper we explain our threestepalgorithm and present and discuss a simulation by which we validated our algorithm.