Explanatory Visualization with LLMs for Employment Law

Abstract

This paper presents EmployODR, an AI-enhanced online tool to help improve access to employment law dispute resolution. The system combines a chat bot powered by a Large Language Model (LLM) with explanatory visualizations that aim to communicate the system’s reasoning. Our tool integrates GPT-3.5 Turbo with structured legal data through a Retrieval-Augmented Generation (RAG) pipeline. The system guides users through complex employment law scenarios using network maps visualized using node-link diagrams, and plain-language summaries. Through our ongoing research, an early framework has emerged that we then apply for scalable AI-driven legal visualization that we hope can aid in bridging the gap between complex legal procedures and public accessibility, placing explainable AI as an essential component for democratizing access to justice. This work supports the developing integration of legal technology and explainable AI by exploring how visualization principles can promote transparency in AI-driven legal decision-making systems, aiming to formulate methodologies applicable across various legal domains to enhance public access to justice.

Downloads

PDF

Poster

See below for embedded view

Citation

E. G. Ogbonda, S. Nason, J. C. Roberts, and P. W. S. Butcher, “Explanatory Visualization with LLMs for Employment Law,” in Posters presented at the IEEE Conference on Visualization (IEEE VIS 2025), Vienna, Austria, 2025.

Bibtex

@inproceedings{Ogbonda-et-al-Poster-VIS2025,
  author = {Ogbonda, Ebube Glory and Nason, Sarah and Roberts, Jonathan C. and Butcher, Peter W.S.},
  title = {{Explanatory Visualization with LLMs for Employment Law}},
  year = {2025},
  month = nov,
  booktitle = {Posters presented at the IEEE Conference on Visualization (IEEE VIS 2025), Vienna, Austria}
}