Prefix tuning with prompt augmentation for efficient financial news summarization

Shangyang Mou (), Qiang Xue, Chen Xunquan, Jinhui Chen, Ryoichi Takashima, Tetsuya Takiguchi and Yasuo Ariki
Additional contact information
Shangyang Mou: Kobe University
Qiang Xue: Kobe University
Jinhui Chen: Wakayama University
Ryoichi Takashima: Kobe University
Tetsuya Takiguchi: Kobe University
Yasuo Ariki: Kobe University

Journal of Computational Social Science, 2025, vol. 8, issue 1, No 19, 16 pages

Abstract: Abstract In financial markets, the sentiment expressed in news articles plays a pivotal role in interpreting and forecasting market trends, which also holds true for the task of financial news summarization (FNS). Leveraging AI models to analyze social science data, this paper employs financial sentiment to improve FNS effectiveness by introducing a novel method that combines the sentiment polarity extracted from financial news with prompt augmentation techniques to ensure that the generated summaries are emotionally consistent with the source articles. Specifically, the detected sentiments are embedded into prompts and provide directive instructions to the model to generate summaries. Furthermore, to address the problem of limited large-scale datasets for FNS and ensure more tailored results, we employed prefix tuning as a fine-tuning strategy. Preliminary results indicate that our combined methodology outperforms approaches that use only prefix tuning. The experimental findings further validate the significance of sentiment analysis in FNS, which enhances the accuracy of capturing and reflecting market sentiment, thereby yielding valuable insights into financial markets. This method not only improves the accuracy and relevance of summaries but also ensures that their content is emotionally consistent with the source news, offering a new perspective on financial news summarization.

Keywords: Financial news summarization; Financial sentiment analysis; Natural language processing; Prompt augmentation; Prefix tuning (search for similar items in EconPapers)
Date: 2025
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://link.springer.com/10.1007/s42001-024-00352-w 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:jcsosc:v:8:y:2025:i:1:d:10.1007_s42001-024-00352-w

Ordering information: This journal article can be ordered from
http://www.springer. ... iences/journal/42001

DOI: 10.1007/s42001-024-00352-w

Access Statistics for this article

Journal of Computational Social Science is currently edited by Takashi Kamihigashi

More articles in Journal of Computational Social Science from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().