This paper examines the main methodological issues to be considered for case-based reasoning (CBR) systems. The advantages of knowledge representation in cases are discussed, giving the rationale for these systems. Many different aspects of design are considered, including user requirements where the system is intended to encourage user learning. A framework for designing such case-based learning and reasoning (CB-LR) systems is discussed. The focus is on feature calibration and case stabilisation processes, together with issues concerning implementation and evaluation of systems.
© 2001-2024 Fundación Dialnet · Todos los derechos reservados