<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>This paper deals with an important task in legal text processing, namely reference and relation extraction from legal documents, which includes two subtasks: 1) reference extraction; 2) relation determination. Motivated by the fact that two subtasks are related and share common information, we propose a joint learning model that solves simultaneously both subtasks. Our model employs a Transformer-based encoder-decoder architecture with non-autoregressive decoding that allows relaxing the sequentiality of traditional seq2seq models and extracting references and relations in one inference step. We also propose a method to enrich the decoder input with learnable meaningful information and therefore, improve the model accuracy. Experimental results on a dataset consisting of 5031 legal documents in Vietnamese with 61,446 references show that our proposed model performs better results than several strong baselines and achieves an F<sub>1</sub> score of 99.4% for the joint reference and relation extraction task.</p>
</abstract>

This paper deals with an important task in legal text processing, namely reference and relation extraction from legal documents, which includes two subtasks: 1) reference extraction; 2) relation determination. Motivated by the fact that two subtasks are related and share common information, we propose a joint learning model that solves simultaneously both subtasks. Our model employs a Transformer-based encoder-decoder architecture with non-autoregressive decoding that allows relaxing the sequentiality of traditional seq2seq models and extracting references and relations in one inference step. We also propose a method to enrich the decoder input with learnable meaningful information and therefore, improve the model accuracy. Experimental results on a dataset consisting of 5031 legal documents in Vietnamese with 61,446 references show that our proposed model performs better results than several strong baselines and achieves an F1 score of 99.4% for the joint reference and relation extraction task.

Joint Reference and Relation Extraction from Legal Documents with Enhanced Decoder Input

Cybernetics and Information Technologies

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

This paper deals with an important task in legal text processing, namely reference and relation extraction from legal documents, which includes two subtasks:...

Joint Reference and Relation Extraction from Legal Documents with Enhanced Decoder Input

Published Online: Jun 12, 2023

Page range: 72 - 86

Received: Feb 17, 2023

Accepted: May 03, 2023

DOI: https://doi.org/10.2478/cait-2023-0014

Keywords
Reference extraction, relation extraction, legal documents, transformer, joint models

© 2023 Nguyen Thi Thanh Thuy et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Joint Reference and Relation Extraction from Legal Documents with Enhanced Decoder Input

Published Online: Jun 12, 2023

Page range: 72 - 86

Received: Feb 17, 2023

Accepted: May 03, 2023

DOI: https://doi.org/10.2478/cait-2023-0014

KeywordsReference extraction, relation extraction, legal documents, transformer, joint models

© 2023 Nguyen Thi Thanh Thuy et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Keywords
Reference extraction, relation extraction, legal documents, transformer, joint models