Using LLMs to Discover Legal Factors

DOI

10.3233/FAIA241234

Authors

Morgan Gray, University of Pittsburgh
Jaromir Savelka, School of Computer Science
Wesley Oliver, Duquesne University
Kevin Ashley, University of Pittsburgh

Document Type

Conference Paper

Publication Date

1-1-2024

Publication Title

Frontiers in Artificial Intelligence and Applications

Volume

395

First Page

60

Last Page

71

ISSN

9226389

Keywords

large language models, legal factors, legal reasoning, machine learning

Abstract

Factors are a foundational component of legal analysis and computational models of legal reasoning. These factor-based representations enable lawyers, judges, and AI and Law researchers to reason about legal cases. In this paper, we introduce a methodology that leverages large language models (LLMs) to discover lists of factors that effectively represent a legal domain. Our method takes as input raw court opinions and produces a set of factors and associated definitions. We demonstrate that a semi-automated approach, incorporating minimal human involvement, produces factor representations that can predict case outcomes with moderate success, if not yet as well as expert-defined factors can.

Open Access

Hybrid_Gold

Repository Citation

Gray, M., Savelka, J., Oliver, W., & Ashley, K. (2024). Using LLMs to Discover Legal Factors. Frontiers in Artificial Intelligence and Applications, 395, 60-71. https://doi.org/10.3233/FAIA241234