MODEL-K(ID:4187/mod013)Specification language for both knowledge level and symbol level Behaviour Related languages
References: MODEL-K, a language that allows both, to specify KADS conceptual models and to extend them to operational systems. This leads to systems that correctly reflect their underlying conceptual model by retaining structural correspondence between model and implementation. The resulting systems are more perspicuous, easier to communicate and to explain to the user, and better to maintain. Since both, modelling and operationalizing may be interleaved, MODEL-K supports the integration of prototyping in model-based methodologies like KADS as early as possible and on the highest level possible. We illustrate the MODEL-K language with a conceptual model for assignment tasks. We demonstrate how the model, originally developed for the domain of office room allocation, can be reused for an application assigning gates to airplanes External link: page at cite-seer MODEL-K, a language that allows both, to specify KADS conceptual models and to extend them to operational systems. This leads to systems that correctly reflect their underlying conceptual model by retaining structural correspondence between model and implementation. The resulting systems are more perspicuous, easier to communicate and to explain to the user, and better to maintain. Since both, modelling and operationalizing may be interleaved, MODEL-K supports the integration of prototyping in model-based methodologies like KADS as early as possible and on the highest level possible. We illustrate the MODEL-K language with a conceptual model for assignment tasks. We demonstrate how the model, originally developed for the domain of office room allocation, can be reused for an application assigning gates to airplanes interleaved, early prototyping is supported at the highest level. Systems written in MODEL-K contain their conceptual model, making them more transparent, easier to communicate to the expert, to explain to the user, and to maintain by the knowledge engineer. The strategy layer of KADS is supposed to control and possibly repair the activities being modeled by the lower layers. MODEL-K views this kind of strategic reasoning as a meta-activity. In the REFLECT project, we came to view meta-activities like resource-management or competence assessment as ordinary problem solving methods, that in turn can be described using KADS. Correspondingly, we extended MODEL-K to model and operationalize such meta-activities. Abstract: in David, J.P., Krivine, J.P.Simmons R. (eds), ""Second Generation Expert Systems"" Springer-Verlag, 1993. view details in David, J.P., Krivine, J.P.Simmons R. (eds), ""Second Generation Expert Systems"" Springer-Verlag, 1993. view details External link: Online copy Extract: FORKADS FORKADS has been developed at the IBM Germany Scientific Center in Heidelberg. It was one of the first published approaches to formal KADS models. Contrary to the languages described so far, FORKADS aims not only at operationalizing KADS models, but also at giving formal foundations to KADS models. Furthermore, the aim is to use this language as a (or perhaps even the only) communication medium between the people responsible for knowledge acquisition and those responsible for system design. For this purpose, the main foundation of FORKADS is a first-order logical language which is extended with notions of concept heterarchies and procedural attachment. In many respects FORKADS is rooted in the LLILOG language Extract: (ML)2 (ML)2 is a language for formalizing KADS models of expertise. It was developed in the course of the ESPRIT Projects "REFLECT" and "KADS-II" and a bilateral research project of the Netherlands Energy Research Foundation ECN and the University of Amsterdam. Although a subset of (ML)2 can be operationalized to allow explorative prototyping it is mainly introduced as a formalization language. Extract: MODEL-K MODEL-K is a language for operationalizing KADS models of expertise. It has been developed at the German National Research Center for Computer Science, GMD in Bonn. It is again implemented in BABYLON [CPV89]. The main point of MODEL-K is the use of an existing AI shell, i.e. BABYLON, to operationalize KADS models of expertise. This is done by introducing KADS-specific modelling primitives in BABYLON. An operational model can then be built by attaching BABYLON code to these modelling terms. Similar to OMOS, MODEL-K allows the evaluation of the model of expertise by prototyping. Additionally, MODEL-K is considered as a language for the implementation of a final expert system. The incremental development of an expert system which reflects the conceptual structure of the knowledge level description of the expertise becomes possible. The conceptual structure thus becomes a tool for improving understandability and maintenance. Another important feature of MODEL-K is its possibility of modelling reflective problem solving. This can be regarded as a step in the direction of a strategic layer. in David, J.P., Krivine, J.P.Simmons R. (eds), ""Second Generation Expert Systems"" Springer-Verlag, 1993. view details |