A Hybird Case-Based Reasoning Architecture and Its Application

Abstract

This paper proposes a hybrid CBR architecture to help CBR reasoning. It hybridizes CBR, fuzzy neural networks, induction, utility-based decision theory, and knowledge-based planning technology to facilitate solutions finding. The basic mechanism is CBR which accumulates experiences as cases in the case library and proposes solutions by adapting the old cases that have successfully solved similar previous case. The distributed fuzzy neural network is introduced to perform approximate matching to tolerate potential noise in case retrieval. The induction technology along with relevance theory is used in case selection, adaptation, and retaining. Knowledge-based planning is used as a general architecture for case adaptation by creating an adaptation plan, whose execution, in turn, proposes a solution. Hybridizing these techniques in the CBR module can effectively produce a high-quality solution for a given problem.