Digital Legal Lab member Antoine Louis will present his research (“Interpretable Long-Form Legal Question Answering with Retrieval-Augmented Large Language Models“), written together with Gijs van Dijck and Gerasimos Spanakis, at the AAAI2024!

About the research:

Many individuals may encounter legal disputes without understanding how to navigate the complex legal landscape, leaving them vulnerable. The advancement of natural language processing offers potential solutions to this issue through automated legal aid systems. However, existing legal question answering methods often provide limited responses. This study introduces a new approach that generates detailed answers to legal questions by first retrieving relevant information and then analyzing it. The Long-form Legal Question Answering (LLeQA) dataset, containing expert-annotated legal questions and detailed answers in French, supports this approach. Experimental results show promise, but there are areas for improvement. The release of code, data, and models aims to facilitate further research in this area and provide a benchmark for evaluating similar systems.

About Antoine:

Antoine Louis is a PHD Researcher in Artificial Intelligence & Law at the Maastricht Law & Tech Lab. His research focuses on data science and natural language processing technologies to automate the law.