On the development of Intelligent CAI system on the Internet

构建了一个基于相似性度量的模糊法律专家系统(FLES)，为法律领域的计算机辅助教学(CAI)奠定了基础。自由贸易区的目标法律是《联合国国际货物销售合同公约》(《销售公约》)。FLES的目的是对《公约》中模糊的法律概念进行分类，支持国际法初学者的教育。FLES由模糊法律案例推理(FLCBR)模块和模糊法律论证(FLA)模块组成。前者提供了对模糊法律概念的初步研究。后者是前者的延伸，可以在原告和被告之间进行辩论。这两个模块在知识表示和相似度度量中考虑了模糊性和上下文敏感效应。在案例推理中，引入了一个层次模糊框架来表示由问题、特征和案例规则组成的案例。基于Hausdorff距离的相似性度量是检索和推理案例的基础，FLA采用模糊因素层次来表示由问题、抽象因素和原子因素组成的案例。将基于Hausdorff距离的相似度、基于特征的相似度的模糊扩展和新提出的基于上下文的相似度相结合，构建了反映原告和被告观点的法律论据模型--基于因子的相似性度量，并通过多名法学院学生对所构建的系统进行了评估，证明了该研究项目有助于实现法学领域的计算机辅助教学系统。

A fuzzy legal expert system (FLES) based on similarity measure is constructed as a basis for the computer aided instruction (CAI) in the law field. The target law of the FLES is the United Nations Convention on Contracts for the International Sale of Goods (CISG). The aim of the FLES is to classify vague legal concepts in the CISG and support the education for the beginners of international law.The FLES is composed of fuzzy legal case-based reasoning (FLCBR) module and fuzzy legal argument (FLA) module. The former provides a primary study on vague legal concepts. The latter is the extension of the former, and can make an argument between plaintiff and defendant. The fuzziness and context-sensitive effects are taken into account in the knowledge representation and similarity measures in these two modules. In FLCBR, a hierarchical fuzzy frame is introduced to represent the case that is composed of issues, features and case rules. The similarity measure in the case of retrieval and inference is based on the newly introduced Hausdorff distance-based similarity measure.In FLA, a fuzzy factor hierarchy is studied to represent the case that is composed of issues, abstract factors and atomic factors. The legal argument consisting of claim, objection and rebuttal, that reflect the viewpoints of plaintiff and defendant, is modeled by the factor-based similarity measure that is a structural similarity measure integrated the Hausdorff distance-based similarity, the fuzzy extension of the feature-based similarity and the newly proposed context-based similarity.The constructed system is evaluated by several students in law school, and proved that this research project leads to the realization of CAI system in the law field.