GiusBERTo: A Legal Language Model for Personal Data De-identification in Italian Court of Auditors Decisions

Recent advances in Natural Language Processing have demonstrated the effectiveness of pretrained language models like BERT for a variety of downstream tasks. We present GiusBERTo, the first BERT-based model specialized for anonymizing personal data in Italian legal documents. GiusBERTo is trained on...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2024-06
Main Authors: Salierno, Giulio, Bertè, Rosamaria, Attias, Luca, Morrone, Carla, Pettazzoni, Dario, Battisti, Daniela
Format: Article
Language:English
Subjects:
Decisions
Natural language processing
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Recent advances in Natural Language Processing have demonstrated the effectiveness of pretrained language models like BERT for a variety of downstream tasks. We present GiusBERTo, the first BERT-based model specialized for anonymizing personal data in Italian legal documents. GiusBERTo is trained on a large dataset of Court of Auditors decisions to recognize entities to anonymize, including names, dates, locations, while retaining contextual relevance. We evaluate GiusBERTo on a held-out test set and achieve 97% token-level accuracy. GiusBERTo provides the Italian legal community with an accurate and tailored BERT model for de-identification, balancing privacy and data protection.
ISSN:2331-8422