Computational Methods and Techniques for Case-Based Reasoning (CBR)

Case-based reasoning (CBR) is a computational model of problem solving. It uses a representation of specific episodes of problem solving to learn to solve a new problem. CBR uses the experience of past problem solving when solving a new problem. CBR systems store past experience as individual problem solving episodes. These two parts of case-based reasoning are modeling as a computational model. In this context, we propose to present a methodology allowing the performance of the CBR system. The methodology consists of designing experiments to control the variability in the behavior of the CBR system and obtaining empirical performance data.