Constructivist foundations of modeling?A Kantian perspective
Marco Cesare Bettoni · International Journal of Intelligent Systems · 1997
By linking Knowledge Engineering to Kantian Philosophy, this paper attempts to elaborate a potential theoretical foundation for understanding the nature of expertise and the processes of modeling. The established way of modeling is criticized for presuming the relation between the real objective world and the AI software models being a mapping (abstraction, copy, representation, etc.) relation. The paper argues that this aspect of the foundations of modeling in Knowledge Engineering could be improved by employing the standpoint presented in Kant's Critique of Pure Reason. Two hypotheses and ten principles of a constructivist modeling paradigm based on Kant's work are proposed. The term Constructivist here refers to the hypothesis that a model cannot correspond to reality but merely be viable in (i.e., fit into) reality.