Atze Dijkstra, Martijn Schrage
Januari 8, 1999
Transfer of knowledge is an essential ingredient of the human way of living, necessary for survival in general but even more so in our information oriented western society. Educational institutions make their best effort in offering knowledge, hoping that students will acquire this knowledge as efficient as possible. The ways and means by which knowledge is transferred do not restrict themselves to the traditional ways anymore but are becoming more varied with the growth of the variation of media capable of storing and transferring knowledge.
Within this context, research is often focussed on the receiving side: how does a student (a receiver) receive knowledge, what is a good way of presenting it, given a certain model of learning? Other issues concern the different technologies available to implement an educational presentation for a receiver: text, images, sound, movie clips, and interaction together designated by the term multimedia. However, one issue is addressed less frequently, i.e. how do we as authors (the givers) of such educational presentations construct the material. How is material organized, grouped; how is this material annotated with meta information normally only stored inside the mind of authors?
It is argued in this article that especially these last issues concerning the 'giving' side are generally not well supported in existing courseware authoring systems. Writing courseware often involves writing a presentation to be used by students, but in the process of creating such a presentation the original knowledge, concerning the design decisions is lost. We argue that in order to be able to deal with the variations in the giving and receiving ends of knowledge transfer, as well as the transfer itself, a full fledged development environment is needed.
Especially in the field of computer science, where education faces the difficulty of a rapidly evolving knowledge domain, a system that provides a road map on the existing material will help to reduce time and effort needed to update course material. Other advantages include more uniformity in produced educational material, by having the system provide templates for common patterns in the material. The meta information present in the system could also be used to create tailor made, possibly electronic, presentations of the material, taking into account student-specific properties and circumstances.
Such a development environment should support specific development oriented requirements like versioning and storing of created artifacts, but it should also cater for the specific courseware authoring needs. We will compare existing environments with a first courseware authoring and management prototype - called SKIT - that we built. The prototype system attempts to form connection between a traditional database layer and a powerful, intuitive user interface environment for creating and managing large bodies of objects and relations between these objects. Using the results of this comparison, we propose an architecture supporting the separately usable abstractions of data (i.e. content) and relationships between data (i.e. structure).