A Comparision Of Two Approaches To Model-Based Knowledge Acquisition
Buchtitel: Proceedings of the European Knowledge Acquisition Workshop (EKAW-94), Hoegaarden, Belgium, September 26-29, 1994
Reihe: Lecture Notes in Artificial Intelligence (LNAI)
Verlag: Springer Verlag, Berlin
This paper discusses and compares two different approaches to model-based knowledge acquisition. That is, we regard the Model-based and Incremental Knowledge Engineering (MIKE) approach and the Configurable Role-limiting Method approach (CRLM). MIKE is based on the distinction of different phases in the software development process. It uses the formal and operational knowledge specification language KARL allowing a precise and unique description of a model of expertise which is the outcome of the analysis phase. CRLM is based on the role-limiting method approach. Role limiting shells are implementations of strong problem-solving methods and substantially simplify knowledge acquisition through guidance by predefined models of problem-solving and by sophisticated graphical user interfaces. The main disadvantages, namely inflexibility and brittleness, are to some degree overcome by the CRLM where the problem-solving methods are split into smaller parts, which can then be reconfigured allowing the integration of new methods or other method combinations. Although these two approaches are often discussed as contradictory, we, however, experienced that both approaches complete each other very well. As an outcome of our comparison, we outline topics of future research for both approaches.
Weitere Informationen unter: Link