A Rule-Based/ Neural Network Hybrid Legal Expert System: a Prototype for Providing Legal Advice on Criminal Cases Under Ethiopian Law

No Thumbnail Available

Date

2005-06

Journal Title

Journal ISSN

Volume Title

Publisher

Addis Ababa University

Abstract

Expert system s have been applied in various do ma in s in o rd e r to solve pro blest hat require Expert knowledge e. Law is one such domain area that can receive a lot of help from ex pert systems in the attempt to co me up with afire and efficient judicial system. Legal expert systems contribute to this endeavor by embedding g expert knowledge with in the m in various forms and providing expert advice .approaches abound in the design o f legal ex pert systems rang in g from rule based legal perpetrate systems that represent the know ledge acquired from experts in the form o f IF-TH EN rules to legal expert systems employ in g neural networks. The approach adopted in this research in the design of a prototype legal expert system HyRIlNLES (Hybrid Rule based Neural network Legal Ex pert System) is a combination of a rule based approach and a neural network model in a n attempt to reap the best o f each . The rule based module of the system works with rules that were derived from legal experts regarding a specific domain of criminal law in order to arrive at a judgment. The neural network part of the prototype Hy Ru NLES system is needed to h an dl e a special type of open texture (re la tin g to sentencing discretion) that the rule based part can't handle . The neural net-work part works based on a modeled enveloped through training an ANN us in g past court cases. The test result o f the performance of the system showed that the model of legal reasoning use d in this research and adopted in the design of the prototype legal e x pert system is functional and Could be encouraging to implement in the form of a full y functional legal e x pert system.

Description

Keywords

Information Science

Citation