Price Movement Prediction in Financial Markets from News Information Based on Contrastive Learning between Financial Time-Series and Text [in Japanese]

Ryoya Yoshida, Ryota Ozaki, Kentaro Imajo, Masanori Hirano

The 36th meeting of Special Interest Group on Financial Informatics of Japanese Society for Artificial Intelligence, pp. 92-98, Mar. 21, 2026

Conference

The 36th meeting of Special Interest Group on Financial Informatics of Japanese Society for Artificial Intelligence (SIG-FIN)

Abstract

This paper proposes a price prediction model for financial market price movement using contrastive learning between news information and financial time-series data. By focusing on the correlation between news and actual market data and utilizing contrastive learning to narrow the gap between their respective feature spaces, the accuracy of price movement prediction is improved. Experiments using the Nikkei index demonstrate that the proposed method achieves superior predictive accuracy compared to conventional techniques.

Keywords

Stock Price Prediction; Contrastive Learning; LLM; Financial Time Series;

doi

10.11517/jsaisigtwo.2026.FIN-036_92

bibtex

@inproceedings{Yoshida2026-sigfin36,
  title={{Price Movement Prediction in Financial Markets from News Information Based on Contrastive Learning between Financial Time-Series and Text [in Japanese]}},
  author={Ryoya Yoshida and Ryota Ozaki and Kentaro Imajo and Masanori Hirano},
  booktitle={The 36th meeting of Special Interest Group on Financial Informatics of Japanese Society for Artificial Intelligence},
  pages={92-98},
  doi={10.11517/jsaisigtwo.2026.FIN-036_92},
  year={2026}
}