Reasoning as Remembering: The Theory and Practice of CBR

Kristian Hammond

Over the past few years, Case-Based Reasoning (CBR) has grown from a Yalecentric view of cognition to a solid sub-area that is supported by wide-spread academic research and industrial development. Unfortunately, the definition of what is and is not CBR remains seriously ambiguous. In this talk, I will look at different takes on CBR and suggest a definition that, oddly enough, doesn’t include the use of "cases ". In particular I will look at: CBR as nothing new. CBR as an alternative cognitive model. CBR as an approach to knowledge engineering. CBR as a new set of assumptions. CBR as a new set of modeling goals. In the end, I argue that the CBR is part of a larger model of long-term agency. This model is distinguished by the view of agents and environments as dynamic entities that change to fit each other over time. Along with arguing for a new view of autonomous agents, this model supports the main tenant of CBR, that reasoning and learning must be linked within any intelligent system.


This page is copyrighted by AAAI. All rights reserved. Your use of this site constitutes acceptance of all of AAAI's terms and conditions and privacy policy.