Why do we need Ontology for Agent-Based Models?

• Pierre Livet ^[1] ; Denis Phan ^[2] ; Lena Sanders ^[3]
  1. [1] CEPERC, UMR 6059, CNRS & Université de Provence, France
  2. [2] GEMAS UMR 8598, CNRS & University Paris IV – Sorbonne, France
  3. [3] Géographie-cités, UMR 8504, CNRS & Université Paris 1 & Université Paris 7, France

  Mostrar afiliaciones +
• Localización: Complexity and Artificial Markets, 2008, ISBN 9783540705567, págs. 133-145
• Idioma: inglés
• Texto completo no disponible (Saber más ...)
• Resumen

  • The aim of this paper is to stress some ontological and methodological issues for Agent-Based Model (ABM) building, exploration, and evaluation in the Social and Human Sciences. Two particular domain of interest are to compare ABM and simulations (Model To Model) within a given academic field or across different disciplines and to use ontology for to discuss about the epistemic and methodological consequences of modeling choices. The paper starts with some definitions of ontology in philosophy and computer sciences. The implicit and different ontology which underlies the approach of a same object of interest are discussed in the case of spatial economists and geographers. Finally, using the case of Shelling’s model, we discuss the concept of “ontological test,” and raise the question of the ontological compatibility between the “model world” and the “real world.”