Researches on Case-Representation in Case-Based Reasoning System
Xiuzhen Xie · Journal of Hefei University · 2007
Case-based Reasoning(CBR) as a cognitive model suggests that people learn it the best from storing former problem-solving cases that they solve new problems,and plays an important role in Machine Learning methods.Cases are the contexts that remindings of right cases are fodder for generalization,and present the key experience in CBR process.They may be represented in semi-structured or unstructured models,even in natural language text,that is strongly interrelated with the system efficiency.Analyzing the most case-representation methods,the paper presents different strategies to approach case-representation in the status of different conditions,and suggests choosing the most suitable style for the construction of case-base.