Dynamic Hypertext Generation for Reusing Open Corpus Content

Abstract

Adaptive hypermedia systems traditionally focus on providing personalised learning services for formal or informal learners. The learning material is typically sourced from a proprietary set of closed corpus content. A fundamental problem with this type of architecture is the need for handcrafted learning objects, enriched with considerable amounts of metadata. The challenge of generating adaptive and personalised hypertext presentations from open source content promises a dramatic improvement of the choice of information shown to the learner. This paper proposes an architecture of such a dynamic hypertext generation system and its use in an authentic learning environment. The system is evaluated in terms of educational benefit, as well as the satisfaction of the users testing the system. Concluding from this evaluation, the paper will explore the future work necessary to further enhance the system performance and learning experience.