This repository contains the annotation guidelines for the annotation of the MuliLegalNeg dataset and the code to finetune negation models based on this data.
A pre-trained model is available on huggingface. See LegalNegBERT for how this model was trained and how to use it.
@misc{christen2023resolving,
title={Resolving Legalese: A Multilingual Exploration of Negation Scope Resolution in Legal Documents},
author={Ramona Christen and Anastassia Shaitarova and Matthias Stürmer and Joel Niklaus},
year={2023},
eprint={2309.08695},
archivePrefix={arXiv},
primaryClass={cs.CL}
}