USING GENETIC ALGORITHM AND PARTICLE SWARM OPTIMIZATION IN LEGAL SYSTEM

Abstract

In contrast to the theory of jurisprudence, pragmatic reasoning was developed as a model for legal case analysis and interpretation. Specific legal cases are thought of as having a set of contextual information surrounding it. The combination of optimization techniques namely as Genetic Algorithms (GA) and Particle Swarm Optimization (PSO) gathered notable attention in legal systems, where the reduction of court cases is a pressing concern. The reasoning derived from these cases leads to solutions that are closely linked to particular situations. In order to generalize these solutions, machine learning methods are applied. GA helps in feature selection by identifying the most meaningful case attributes, while PSO fine-tunes model parameters to achieve higher accuracy. One of the techniques to induce reasoning from cases is considered ‘examples’ with a ‘classification’ in a certain area. In this paper, a genetic algorithm approach to inductive feature learning of cases from a legal database resulting in case-based reasoning taxonomy is presented. The acquired knowledge can then be used to reason from cases which are by their features and classifications ‘interesting’ and predict the classification of other cases.