In legal case-based reasoning (CBR), there exist problems concerning fuzziness, e.g., representation of precedents, their retrieval, and similarity measures. In our proposed fuzzy legal CBR system, the issues and features of precedent are characterized on the basis of the facts of precedent and statute rule. The case rule that is used for interpreting the court judgment, which cannot be obtained from the statute rule directly, is made by experts. Fuzziness is represented by membership functions. Features and case rules, written in terms of Compound Predicate Formula (CPF) and frame, are stored in a case base. Cases similar to a new case are retrieved by issues and features, and an inference is made by case rules. A user interface is devised for this system. The system proposed here will be used for law education, where the target law of the system is contract, especially as it relates to the United Nations Convention on Contracts for the International Sale of Goods (CISG).
