AI and Law (2)
Hajime Yoshino*and Katsumi Nitta**
*Faculty of Law, Meiji Gakuin University 1-2-37 Shirokanedai, Minatoku, Tokyo 108, Japan
**Department of Computational Intelligence and Systems Science,Tokyo Institute of Technology 4259 Nagatsuta, Midori-ku, Yokohama 226, Japan
In the last issue (Vol.1, No.2), we introduced the Legal Expert System (LES) project led by Hajime Yoshino of Meiji Gakuin University, presenting six papers on the LES project. Those papers were mainly related to higher order reasoning systems such as ase-based reasoning, abductive and inductive logic programming, nonmonotonic reasoning, and analogical reasoning. The objective of the LES project was to develop a legal expert system effective for use by lawyers, so the project covers inference mechanisms, analysis of legal knowledge, and user interfaces. In this second special issue on the LES project, we present five more papers, mainly related to the analysis of legal knowledge, legal knowledge representation language, and legal reasoning system user interfaces. Hajime Yoshino analyzes the logical structure of contract law. To develop a knowledge base for the United Nations Convention on Contracts for the International Sale of Goods (CISG), he proposes a clear logical model of the contract law system, which treats relations between events and legal status such as rights and obligations. Yoshino demonstrates that legal metarules are effective in constructing deductive legal reasoning systems, and are appropriate from the viewpoint of jurisprudence. Seiichiro Sakurai discusses the logical features of the legal knowledge representation language, CPF, developed by Hajime Yoshino. CPF is a logic programming language that enhances the representation of complex data structures. CPF is a convenient tool for representing legal knowledge, yet lawyers often attempt to describe nonexecutable forms of CPF rules.Sakurai introduces a way to construct an executable knowledge base from lawyers’ CPF rules. Masato Shibasaki and Katsumi Nitta introduce a new framework to formalize nonmonotonic reasoning with dynamic priorities. The several frameworks proposed thus far to model relationships among arguments do not treat complex arguments, composed of strict rules and default rules. They show that the new framework represents such relationships and analyze these relationships for this framework and others. Takashi Miyata and Yuji Matsumoto introduce LES natural language generation using a user interface for lawyers rather than computer scientists. They describe a sentence generation system that translates logical forms provided from an inference engine into natural-language sentences, and present the unification grammar, generation algorithm and graphical debugging tool. To develop a knowledge base, the lawyers of the LES project analyze and represents the relationships between requirements (actions or events) and consequences (legal status) of legal rules in the form of logical flowcharts. Once the appropriateness of a flowchart is confirmed, they convert it to a CPF rule in their knowledge base. Koji Miyagi, Motoki Miura and Jiro Tanaka developed a flowchart editor that makes legal flowcharting easier. To make it easier to decide where to locate flowchart components and draw linens between the components, the editor possesses several algorithms.