A Symposium on Artificial Intelligence and Research in Japan

Abstract

This symposium was planned to introduce research activities on AI and Law in Japan. It includes five papers. Research into AI and Law in Japan began in 1980, when Hajime Yoshino (Meiji Gakuin University) started the A-project. Since then, he has recruited several lawyers and computer scientists, and developed some legal reasoning systems which handle the civil code. His research activities gained a high evaluation rating, with the result that the Ministry of Education and Science decided to support his ‘Legal Expert’ project for five years (1993–1998). Currently, about 30 lawyers and 10 computer scientists have joined the project and have developed legal reasoning systems. The main research objective has been to develop logic-based legal knowledge bases. Yoshino selected the United Nations Convention for International Sales of Goods (CISG) as the target domain, to maximise the possibilities for international collaboration. Further research arose from the Fifth Generation Computer Systems Project (1982–1995), which developed a wide range of logic-based parallel processing technologies from hardware to application software. While working for the FGCS project, Katsuma Nitta (Electrotechnical Laboratory and Tokyo Institute of Technology) identified legal reasoning systems as a promising application domain. He organized a Legal Reasoning Research Group consisting of eight researchers, and developed legal reasoning systems. Since the FGCS project ended in 1995, Nitta has been continuing the development of these systems in the Electrotechnical Laboratory and Tokyo Institute of Technology. The papers of this collection were selected from the Legal Reasoning System Workshop held during the International Symposium of FGCS 1994, and revised by appending recent research results. All papers are research results of Yoshino’s ‘Legal Expert’ project or of the FGCS project. Yoshino introduces a logic-based knowledge representation language – Compound Predicate Formula (CPF). He investigated the necessary functions of legal knowledge representation languages and formulated a CPF which has been used in legal reasoning systems. The CPF is a conservative extension of the standard 1st