A Hybrid Legal Expert System

Abstract

Legal expert systems are the nexus of Artificial Intelligence and the law. A legal expert system is system capable of performing at a level expected of a lawyer [Popple 1996, page 3]. Legal expert systems may be designed for use by legally trained people or for use by the general public (lay-people). Legal expert systems designed for use by legally trained people aim to provide a method of speeding-up the provision, and improving the accuracy, of legal research undertaken with the aim of advising the client. Designed for use by legally trained people, these systems may assume general legal knowledge. Consequently the questions asked by the system and the reports returned may be stated at a level appropriate for legally trained people. The primary benefit of this category of legal expert system is the reduction of internal cost of legal research. The flow-on benefits for clients reductions in the cost of legal services and consequently improved access to quality representation, and reduction of the time taken to resolve a legal question. Legal expert systems designed for use by lay-people aim to provide greater access to the law. This category of legal expert system is more difficult to create because no legal knowledge by the user can be assumed. The discovery of the facts of the case becomes problematic [Susskind 2001]. More research is required in the area of fact elicitation before such systems become viable. Once they are viable, access to the law should be dramatically improved. A consequential benefit may be a reduction in litigation, as potential litigants could settle their dispute by reference to the advice of a legal expert system. However, such a system would raise an important ethical question – the creators of such a system may be usurping the role of the courts in that the public may come to rely on the statements by the system as what the law is. SHYSTER-MYCIN is the legal expert system created for and discussed in this thesis. SHYSTER-MYCIN combines rule-based reasoning with case-based reasoning. The system is designed as the first category of legal expert systems described above: a legal expert system to be consulted by legally trained people. This hybrid system enables the case-based reasoner to determine open-textured concepts when required by the rule-based reasoner, MYCIN. The system operates on a reduced version of the Copyright Act 1968, including cases that define the term authorization (see Chapter 2). The Act is reasoned by a system of rules. Whereas cases are reasoned by analogy. This approach is supported by jurisprudential discussions on legal reasoning (see Chapter 3). The system was created in three progressive versions (Chapter 5). The focus of the creation of the system was the reporting of reasons for conclusions. The second and third versions were tested against three criteria: validity, conciseness and correctness (see Chapter 6). The system performed well (see Chapter 7) against those criteria, indicating that the approach taken is appropriate: that is, it is…