Headline Token-based Discriminative Learning for Subheading Generation in News Article

Joonwon Jang, Misuk Kim


Abstract
The news subheading summarizes an article’s contents in several sentences to support the headline limited to solely conveying the main contents. So, it is necessary to generate compelling news subheadings in consideration of the structural characteristics of the news. In this paper, we propose a subheading generation model using topical headline information. We introduce a discriminative learning method that utilizes the prediction result of masked headline tokens. Experiments show that the proposed model is effective and outperforms the comparative models on three news datasets written in two languages. We also show that our model performs robustly on a small dataset and various masking ratios. Qualitative analysis and human evaluations also show that the overall quality of generated subheadings improved over the comparative models.
Anthology ID:
2023.findings-eacl.159
Volume:
Findings of the Association for Computational Linguistics: EACL 2023
Month:
May
Year:
2023
Address:
Dubrovnik, Croatia
Editors:
Andreas Vlachos, Isabelle Augenstein
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
2128–2135
Language:
URL:
https://aclanthology.org/2023.findings-eacl.159
DOI:
10.18653/v1/2023.findings-eacl.159
Bibkey:
Cite (ACL):
Joonwon Jang and Misuk Kim. 2023. Headline Token-based Discriminative Learning for Subheading Generation in News Article. In Findings of the Association for Computational Linguistics: EACL 2023, pages 2128–2135, Dubrovnik, Croatia. Association for Computational Linguistics.
Cite (Informal):
Headline Token-based Discriminative Learning for Subheading Generation in News Article (Jang & Kim, Findings 2023)
Copy Citation:
PDF:
https://aclanthology.org/2023.findings-eacl.159.pdf
Dataset:
 2023.findings-eacl.159.dataset.zip
Video:
 https://aclanthology.org/2023.findings-eacl.159.mp4