Sisyphus-VT: A CommonKADS solution

This article contains a CommonKADS description of the Sisyphus ’93 problem : the VT elevator design . The term ‘‘CommonKADS’’ is used to refer to KADS as it has been developed within ESPRIT project P5248 KADS-II . The goal of this project was to arrive at a de facto European standard for KBS development . As the term ‘‘KADS’’ has become strongly associated with the University of Amsterdam , it is worthwhile to point out that CommonKADS is the result of a cooperative ef fort of a number of partners , in particular also (in the context of this paper) the Free University of Brussels and the Netherlands Energy Research Centre ECN . A central theme of CommonKADS is the idea that various perspectives or models are important in the KBS development process : models of the organization , of the overall task , of the agents involved , of the required communication , of the expertise , and of the design of the final artifact . In this article the focus is on expertise modelling and its relation to design and implementation . An overview of CommonKADS can be found in Schreiber , Wielinga , de Hoog , Akkermans and Van de Velde (1994 b ) ; the textual and graphical notations used are described in Schreiber , Wielinga , Akkermans , Van de Velde and Anjewierden (1994a) . † In Section 2 a brief characterization is given of the VT data set and of the application task . Section 3 describes the structure of the VT domain knowledge in the form of two ontologies . In Section 4 a top-down description is given of tasks and inferences involved . The methods underlying this decomposition are characterized . System design and implementation are addressed in Section 5 . The process of using the VT domain theory within our application is discussed also in this context . Section 6 discusses the possible contributions of this paper , as well as its limitations . This article assumes that the reader is familiar with the VT data set and the associated ONTOLINGUA ontologies .