Analogical reasoning and case-based learning in model management systems

摘要

Developing computer-based model management systems has been a focus of recent research in decision support. In this paper, we explore the use of analogical reasoning and case-based learning in model management. Analogical reasoning and case-based learning are techniques found useful in human problem solving. They can help model builders apply their modeling experience to construct new models and improve their modeling knowledge through learning. This paper presents a feasible approach to incorporate case-based analogical reasoning in model management systems. First, the role of analogical reasoning and case-based learning in model management is described. A scheme for representing case features is presented. Then, problem and model similarities are discussed at conceptual, structural, and functional levels. This is followed by a description of the process for analogical model formulation, which includes feature mapping, transformation, and evaluation. Finally, examples are illustrated and case-based learning is discussed.