UML as an approach to Modelling Knowledge in Rule-based Systems
2001 (English)In: ES2001-The Twenty-first SGES International Conference on Knowledge Based Systems and Applied Artificial Intelligence., 2001Conference paper (Refereed)
The Knowledge Acquisition (KA) process within Knowledge Based System (KBS) development has always been a difficult task. Some of the difficulties are connected to the range of different users, from the developers to the end users of a system. These users might be interested in different domain knowledge, which has to be preserved in the KBS. This can be facilitated, e.g. through diagrams from the Unified Modelling Language (UML), used for visualising, specifying, constructing and documenting object-oriented systems. UML is a modelling language whose vocabulary and rules focus on a conceptual representation of a system. It is appropriate for modelling systems ranging from enterprise information systems to Web-based applications and real time embedded systems. UML is applicable to a variety of programming languages and it might be quite straightforward applied for developing frame-based systems. Our interest, though, is to apply the UML diagrams to support rule-based KBS development, resulting in modified UML diagrams.
This article suggests an improvement of the KA process from a knowledge representation perspective. The domain knowledge is modelled with modified UML diagrams from both the perspectives of the design user and the end user. These diagrams are used to insert knowledge in the knowledge base as well as generate knowledge about static and dynamic domain knowledge and inform the users about the system’s processing etc. The diagrams should be presented by using graphical objects rather than text code, since these objects can support the graphics in UML’s diagrams and simplify the understanding of the system.
Place, publisher, year, edition, pages
Knowledge Acquisition, Modelling, UML
IdentifiersURN: urn:nbn:se:uu:diva-43780OAI: oai:DiVA.org:uu-43780DiVA: diva2:71685