The key issues for case retrieve are the similarity measurement and matching process. Classical methods for case
retrieve can’t meet the demands of those complex systems, which are of special features, such as textual or fuzzy attributes,
so that case-based reasoning (CBR) system can’t go further in judging, evaluating, deciding, and reasoning. A method, multigalois
lattice (MGL), is presented to do the similarity measurement for numerical, symbolic, fuzzy concept, and fuzzy interval
attributes. Experiment results show that MGL can not only simplify the matching process, but also resolve the problems of
similarity measurement for multi-attributes cases.