About the Project
Project title: Generative Aritificial Intelligence in Legal Translation
Project acronym: GenAI-LT
Project number: IP-2025-02-3140
Project financing: Croatian Science Foundation
Project duration: 2025. – 2028.
Project Leader
Name and surname: Assoc. Prof. Martina Bajčić
E-mail: martina.bajcic@uniri.hr
Project Team Members
Assoc. Prof. Giorgio Maria Di Nunzio, University of Padua
Assoc. Prof. Slavko Žitnik, University of Ljubljana
Bruno Nahod, PhD, Institute for the Croatian Language
Assoc. Prof. Adrijana Martinović, University of Rijeka
Asst. Prof. Dejana Golenko, University of Rijeka
Project Description
The impact of artificial intelligence on interlingual communication is indisputably significant. Neural models have dramatically improved the quality of machine translation, while new generative AI (GAI) technologies and tools, such as large language models, are emerging. Despite the growing interest of both researchers and the general public in the application of these tools, their use in the field of legal translation is underresearched. Considering the unprecedented interest in the application of GAI on the one hand, and the importance of legal translation and understanding legal terminology for all citizens, as well as for the future of European integration and the single market on the other hand, it is instrumental to investigate the application of GAI to legal translation by virtue of empirical research. With this in mind, this proposal aims to explore the possibility of deploying GAI in the translation of legal texts by integrating contemporary methods and approaches from three key areas: automated legal translation, corpus-informed terminology, and (legal) information science (natural language processing; legal information). Computational techniques will be applied to integrate domain knowledge with language models, and to introduce enhancements to the models with an eye to legal terminology. These methods and approaches will be adapted to the characteristics of EU law.
Key words:
- legal terminology, natural language processing (NLP), large language models, legal translation, multilingualism
Project Results