Resolving Legalese: A Multilingual Exploration of Negation Scope Resolution in Legal Documents

Resolving the scope of a negation within a sentence is a challenging NLP task. The complexity of legal texts and the lack of annotated in-domain negation corpora pose challenges for state-of-the-art (SotA) models when performing negation scope resolution on multilingual legal data. Our experiments demonstrate that models pre-trained without legal data under perform in the task of negation scope resolution compared to cross-domain experiments conducted between different domains such as medical data and literary texts. We release a new set of annotated court decisions in German, French, and Italian and use it to improve negation scope resolution in both zero-shot and multilingual settings. We achieve a token-level F1-score of 86.7% in our zero-shot cross-lingual experiments, where the models are trained on two languages and evaluated on the third. Our multilingual experiments, where the models were trained on all available negation data and evaluated on our legal datasets, resulted in an F1-score of 91.1%.