Mimosa: Using Ontologies for Modelling and Simulation
Jahrestagung der Gesellschaft für Informatik
Modeling is the shared activity for both modeling and simulation, and knowledge representation. Howeverknowledge representation is focused on building adescription, usually static, of the reality in terms of named categories, properties, relations and objects, while the modeling and simulation community aims at putting together behavioral components [MB05] as farabstractions of the dynamical structure of the reality.T oface the complexity of the systems we are trying to model and to simulate
... adays, the challenge addressed in this paper is to mix both approaches in ac ommon framework. Trying to use modeling and simulation as away to combine both expert and profane knowledge in support to negotiation activities[Aa03], our aim is to coverthe whole modeling activity going from knowledge elicitation and representation up to running simulation models for exploring prospective options. Fordoing that, we first describe separately the recent advances made in the modeling and simulation community as well as in the knowledge representation community,e specially the ontologies with the arrivalofthe semantic web[LG04]. We showthat, despite the same goal to describe areality,oratleast apart of it, it results in very different, although related concepts. From the analysis of the differences and similarities, we propose an architecture which is being tested in amodeling and simulation platform: Mimosa. The outcome is a formal waytopavethe path from conceptual to running models which is sketched in this paper.T he achievements and the perspectivesare discussed in the conclusion. 2M odeling and simulation Modeling and simulation has been primarily used to model dynamical processes, initially with differential equations. In industrial contexts, the theory of control needed more sophisticated representations, essentially by composing transfer functions. This need gave rise to anumber of formalisms for structuring modeling which were finally unified within awidely recognized framework called DEVS for Discrete EVent System [ZKP00]. DEVS defines ac lear operational semantics and provesi ts closure under composition. However,the manipulation of very complexmodels called for even more structured modeling paradigm among which object-oriented and multi-agent systems.