Saved in:
Bibliographic Details
Main Authors: Bizzaro, Pietro Giovanni, Della Valentina, Elena, Napolitano, Maurizio, Mana, Nadia, Zancanaro, Massimo
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2402.14457
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914812408102912
author Bizzaro, Pietro Giovanni
Della Valentina, Elena
Napolitano, Maurizio
Mana, Nadia
Zancanaro, Massimo
author_facet Bizzaro, Pietro Giovanni
Della Valentina, Elena
Napolitano, Maurizio
Mana, Nadia
Zancanaro, Massimo
contents In this paper, we propose a new annotation scheme to classify different types of clauses in Terms-and-Conditions contracts with the ultimate goal of supporting legal experts to quickly identify and assess problematic issues in this type of legal documents. To this end, we built a small corpus of Terms-and-Conditions contracts and finalized an annotation scheme of 14 categories, eventually reaching an inter-annotator agreement of 0.92. Then, for 11 of them, we experimented with binary classification tasks using few-shot prompting with a multilingual T5 and two fine-tuned versions of two BERT-based LLMs for Italian. Our experiments showed the feasibility of automatic classification of our categories by reaching accuracies ranging from .79 to .95 on validation tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2402_14457
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Annotation and Classification of Relevant Clauses in Terms-and-Conditions Contracts
Bizzaro, Pietro Giovanni
Della Valentina, Elena
Napolitano, Maurizio
Mana, Nadia
Zancanaro, Massimo
Computation and Language
In this paper, we propose a new annotation scheme to classify different types of clauses in Terms-and-Conditions contracts with the ultimate goal of supporting legal experts to quickly identify and assess problematic issues in this type of legal documents. To this end, we built a small corpus of Terms-and-Conditions contracts and finalized an annotation scheme of 14 categories, eventually reaching an inter-annotator agreement of 0.92. Then, for 11 of them, we experimented with binary classification tasks using few-shot prompting with a multilingual T5 and two fine-tuned versions of two BERT-based LLMs for Italian. Our experiments showed the feasibility of automatic classification of our categories by reaching accuracies ranging from .79 to .95 on validation tasks.
title Annotation and Classification of Relevant Clauses in Terms-and-Conditions Contracts
topic Computation and Language
url https://arxiv.org/abs/2402.14457