Data-Augmentation Method for BERT-based Legal Textual Entailment Systems in COLIEE Statute Law Task
Yasuhiro Aoki (),
Masaharu Yoshioka () and
Youta Suzuki ()
Additional contact information
Yasuhiro Aoki: Hokkaido University
Masaharu Yoshioka: Hokkaido University
Youta Suzuki: Hokkaido University
The Review of Socionetwork Strategies, 2022, vol. 16, issue 1, 175-196
Abstract:
Abstract A legal textual entailment task is a task to recognize entailment between a law article and its statements. In the Competition on Legal Information Extraction/Entailment (COLIEE), this task is designed as a task to confirm the entailment of a yes/no answer from the given civil code article(s). Based on the development of deep-learning-based natural language processing tools such as bidirectional encoder representations from transformers (BERT), many participants in the task used such tools, and the best performance system of COLIEE 2020 was a BERT-based system. However, because of the limitation of the size of training data provided by the task organizer, training such tools to adapt to the variability of the questions is difficult. In this paper, we propose a data-augmentation method to make training data using civil code articles for understanding the syntactic structure of the questions and articles for entailment. Our BERT-based ensemble system, which uses this augmentation method, achieves the best performance (accuracy = 0.7037) in Task 4 of COLIEE 2021. We also introduce the results of additional experiments to discuss the characteristics of the proposed method.
Date: 2022
References: View complete reference list from CitEc
Citations:
Downloads: (external link)
http://link.springer.com/10.1007/s12626-022-00104-0 Abstract (text/html)
Access to the full text of the articles in this series is restricted.
Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.
Export reference: BibTeX
RIS (EndNote, ProCite, RefMan)
HTML/Text
Persistent link: https://EconPapers.repec.org/RePEc:spr:trosos:v:16:y:2022:i:1:d:10.1007_s12626-022-00104-0
Ordering information: This journal article can be ordered from
https://www.springer ... ystems/journal/12626
DOI: 10.1007/s12626-022-00104-0
Access Statistics for this article
The Review of Socionetwork Strategies is currently edited by Katsutoshi Yada, Yasuharu Ukai and Marshall Van Alstyne
More articles in The Review of Socionetwork Strategies from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().