📄 Paper     |     🤗 Dataset    |     💻 Code    |     🎥 Online presentation

It was a deeply fulfilling experience to present my very first paper at the Association for Computational Linguistics (ACL) conference, the best conference in the world for research on natural language processing.

I had the opportunity of presenting my brand new dataset of 1100+ legal questions labeled with relevant legislation from 22600+ law articles. With questions such as “What do I risk if I violate professional confidentiality?” or “Can I miss work to visit the doctor?”, this dataset represents the first high-quality, expertly annotated resource for evaluating and developing models capable of retrieving law articles relevant to a short legal question posed in layperson’s terms.

This serves as a robust foundation for advancements in citizen-centric legal information retrieval, which is a key step towards the democratization of access to legal information.