Authors: Kyoungyeon Cho, Seungkum Han, Young Rok Choi, Wonseok Hwang

Published on: September 08, 2023

Impact Score: 8.22

Arxiv code: Arxiv:2309.04146

Summary

• What is new: NESTLE, a no-code tool for statistical analysis of legal documents, leveraging a Large Language Model (LLM) for customizable information extraction.
• Why this is important: Lack of a unified, no-code solution for selecting, structuring, and visualizing large legal text corpora for statistical analysis.
• What the research proposes: The introduction of NESTLE—a tool that employs a Large Language Model and a custom IE system to enable customizable and code-free analysis of legal texts.
• Results: Successfully applied to 15 Korean precedent information extraction tasks and 3 legal text classification tasks, achieving GPT-4 comparable results with minimal human supervision.

Technical Details

Technological frameworks used: NESTLE, powered by a Large Language Model

Models used: Custom internal IE system, GPT-4

Data used: 15 Korean legal precedents, LexGLUE dataset

Potential Impact

Legal analytics, Legal-tech startups, Law firms, Corporate legal departments

Want to implement this idea in a business?

We have generated a startup concept here: LegalEagle Analytics.