StockGenChaR: A Study on the Evaluation of Large Vision-Language Models on Stock Chart Captioning

Le Qiu; Emmanuele Chersoni

StockGenChaR: A Study on the Evaluation of Large Vision-Language Models on Stock Chart Captioning

Department of Language Science and Technology

Research output: Chapter in book / Conference proceeding › Conference article published in proceeding or book › Academic research › peer-review

Abstract

Technical analysis in finance, which aims at forecasting price movements in the future by analyzing past market data, relies on the insights that can be gained from the interpretation of stock charts; therefore, non-expert investors could greatly benefit from AI tools that can assist with the captioning of such charts.
In our work, we introduce a new dataset StockGenChaR to evaluate large vision-language models in image captioning with stock charts. The purpose of the proposed task is to generate informative descriptions of the depicted charts and help to read the sentiment of the market regarding specific stocks, thus providing useful information for investors.

Original language	English
Title of host publication	Proceedings of the 10th Workshop on Financial Technology and Natural Language Processing (FinNLP)
Editors	Chung-Chi Chen, Genta Indra Winata, Stephen Rawls, Anirban Das, Hsin-Hsi Chen, Hiroya Takamura
Publisher	Association for Computational Linguistics
Pages	33-46
Publication status	Published - Nov 2025
Event	The 10th Workshop on Financial Technology and Natural Language Processing (FinNLP) - Suzhou International Expo Centre (SuzhouExpo), Suzhou, China Duration: 9 Nov 2025 → 9 Nov 2025 https://sigfintech.github.io/finnlp.html

Conference

Conference	The 10th Workshop on Financial Technology and Natural Language Processing (FinNLP)
Country/Territory	China
City	Suzhou
Period	9/11/25 → 9/11/25
Internet address	https://sigfintech.github.io/finnlp.html

More information

Cite this

Qiu, L., & Chersoni, E. (2025). StockGenChaR: A Study on the Evaluation of Large Vision-Language Models on Stock Chart Captioning. In C.-C. Chen, G. I. Winata, S. Rawls, A. Das, H.-H. Chen, & H. Takamura (Eds.), Proceedings of the 10th Workshop on Financial Technology and Natural Language Processing (FinNLP) (pp. 33-46). Association for Computational Linguistics. https://arxiv.org/pdf/2412.04041

Qiu, Le ; Chersoni, Emmanuele. / StockGenChaR: A Study on the Evaluation of Large Vision-Language Models on Stock Chart Captioning. Proceedings of the 10th Workshop on Financial Technology and Natural Language Processing (FinNLP). editor / Chung-Chi Chen ; Genta Indra Winata ; Stephen Rawls ; Anirban Das ; Hsin-Hsi Chen ; Hiroya Takamura. Association for Computational Linguistics, 2025. pp. 33-46

@inproceedings{bf912cd8e7304983838617c1f23c56f3,

title = "StockGenChaR: A Study on the Evaluation of Large Vision-Language Models on Stock Chart Captioning",

abstract = "Technical analysis in finance, which aims at forecasting price movements in the future by analyzing past market data, relies on the insights that can be gained from the interpretation of stock charts; therefore, non-expert investors could greatly benefit from AI tools that can assist with the captioning of such charts. In our work, we introduce a new dataset StockGenChaR to evaluate large vision-language models in image captioning with stock charts. The purpose of the proposed task is to generate informative descriptions of the depicted charts and help to read the sentiment of the market regarding specific stocks, thus providing useful information for investors.",

author = "Le Qiu and Emmanuele Chersoni",

year = "2025",

month = nov,

language = "English",

pages = "33--46",

editor = "Chung-Chi Chen and Winata, \{Genta Indra\} and Stephen Rawls and Anirban Das and Hsin-Hsi Chen and Hiroya Takamura",

booktitle = "Proceedings of the 10th Workshop on Financial Technology and Natural Language Processing (FinNLP)",

publisher = "Association for Computational Linguistics",

address = "United States",

note = "The 10th Workshop on Financial Technology and Natural Language Processing (FinNLP) ; Conference date: 09-11-2025 Through 09-11-2025",

url = "https://sigfintech.github.io/finnlp.html",

}

Qiu, L & Chersoni, E 2025, StockGenChaR: A Study on the Evaluation of Large Vision-Language Models on Stock Chart Captioning. in C-C Chen, GI Winata, S Rawls, A Das, H-H Chen & H Takamura (eds), Proceedings of the 10th Workshop on Financial Technology and Natural Language Processing (FinNLP). Association for Computational Linguistics, pp. 33-46, The 10th Workshop on Financial Technology and Natural Language Processing (FinNLP), Suzhou, China, 9/11/25. <https://arxiv.org/pdf/2412.04041>

StockGenChaR: A Study on the Evaluation of Large Vision-Language Models on Stock Chart Captioning. / Qiu, Le; Chersoni, Emmanuele.
Proceedings of the 10th Workshop on Financial Technology and Natural Language Processing (FinNLP). ed. / Chung-Chi Chen; Genta Indra Winata; Stephen Rawls; Anirban Das; Hsin-Hsi Chen; Hiroya Takamura. Association for Computational Linguistics, 2025. p. 33-46.

Research output: Chapter in book / Conference proceeding › Conference article published in proceeding or book › Academic research › peer-review

TY - GEN

T1 - StockGenChaR: A Study on the Evaluation of Large Vision-Language Models on Stock Chart Captioning

AU - Qiu, Le

AU - Chersoni, Emmanuele

PY - 2025/11

Y1 - 2025/11

N2 - Technical analysis in finance, which aims at forecasting price movements in the future by analyzing past market data, relies on the insights that can be gained from the interpretation of stock charts; therefore, non-expert investors could greatly benefit from AI tools that can assist with the captioning of such charts. In our work, we introduce a new dataset StockGenChaR to evaluate large vision-language models in image captioning with stock charts. The purpose of the proposed task is to generate informative descriptions of the depicted charts and help to read the sentiment of the market regarding specific stocks, thus providing useful information for investors.

AB - Technical analysis in finance, which aims at forecasting price movements in the future by analyzing past market data, relies on the insights that can be gained from the interpretation of stock charts; therefore, non-expert investors could greatly benefit from AI tools that can assist with the captioning of such charts. In our work, we introduce a new dataset StockGenChaR to evaluate large vision-language models in image captioning with stock charts. The purpose of the proposed task is to generate informative descriptions of the depicted charts and help to read the sentiment of the market regarding specific stocks, thus providing useful information for investors.

M3 - Conference article published in proceeding or book

SP - 33

EP - 46

BT - Proceedings of the 10th Workshop on Financial Technology and Natural Language Processing (FinNLP)

A2 - Chen, Chung-Chi

A2 - Winata, Genta Indra

A2 - Rawls, Stephen

A2 - Das, Anirban

A2 - Chen, Hsin-Hsi

A2 - Takamura, Hiroya

PB - Association for Computational Linguistics

T2 - The 10th Workshop on Financial Technology and Natural Language Processing (FinNLP)

Y2 - 9 November 2025 through 9 November 2025

ER -

StockGenChaR: A Study on the Evaluation of Large Vision-Language Models on Stock Chart Captioning

Abstract

Conference

More information

Fingerprint

Cite this