The linking of knowledge bases in the context of representation of human knowledge states

THE LINKING OF KNOWLEDGE BASES IN THE CONTEXT OF HUMAN KNOWLEDGE STATES Faleh H. Alkaisi The aim of the investigation described here is to contribute to the development of ideas underlying interactions between complex knowledge bases and interactions between human beings and complex knowledge bases. The primary aim is to test the hypothesis that interaction between knowledge bases with similar structural schemas can be managed by simple mechanisms. Mathematics and Computer Science knowledge have been selected as representative examples of complex knowledge systems. An "Expert - Non-Expert" model simulating the teaching and learning of Mathematics and Computer Science is used to characterise the nature of an interaction process between structures representing the knowledge of the expert (i.e the teacher) and the non-expert (i.e the student). The main steps of the investigation were: Firstly, the literature related to knowledge representation and knowledge based systems using interaction mechanisms was studied. Secondly, the "Expert - Non-Expert" model was used as the basis for specification of a general interaction mechanism. Thirdly, specification of knowledge representation and an interaction mechanism were demonstrated in a prototype system. Finally, the prototype system was evaluated. This was achieved by, comparing and contrasting the expected objectives from using the system with the actual achievements, comparing and contrasting the system with other knowledge based systems, and elaborating comments and suggestions made by other users of the system. The investigation has produced a mechanism for interaction between knowledge bases with similar structural schemas in the context of teaching students. The research has also led to the production of a demonstrator system for managing the interaction mechanism and the knowledge bases, and has outlined ideas for the future implemention of a knowledge based system with geometrical knowledge representation. The overall study has produced a suitable method of interaction between knowledge bases. The study also suggests possiblities for structuring knowledge for the development of intelligent knowledge acquisition and expert systems.