@inproceedings{wu-etal-2025-susgen,
title = "$SusGen-GPT$: A Data-Centric {LLM} for Financial {NLP} and Sustainability Report Generation",
author = "Wu, Qilong and
Xiang, Xiaoneng and
Hejia, Huang and
Wang, Xuan and
Wei Jie, Yeo and
Satapathy, Ranjan and
Filho, Ricardo Shirota and
Veeravalli, Bharadwaj",
editor = "Chiruzzo, Luis and
Ritter, Alan and
Wang, Lu",
booktitle = "Findings of the Association for Computational Linguistics: NAACL 2025",
month = apr,
year = "2025",
address = "Albuquerque, New Mexico",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-naacl.66/",
doi = "10.18653/v1/2025.findings-naacl.66",
pages = "1184--1203",
ISBN = "979-8-89176-195-7",
abstract = "The rapid growth of the financial sector and the increasing focus on Environmental, Social, and Governance (ESG) considerations have created a pressing need for advanced natural language processing (NLP) tools. Despite recent advancements, there is still a notable absence of open-source Large Language Models (LLMs) that are proficient across both general finance and ESG domains, such as generating ESG reports. To address this gap, we introduce $SusGen$-$30k$, a high-quality, category-balanced dataset comprising seven financial NLP tasks. In addition, we propose $TCFD$-$Bench$, a benchmark designed to improve the evaluation of sustainability report generation. Our data-centric approach led to the development of a suite of models, $SusGen$-$GPT$, trained on the curated dataset. These models were evaluated across six adapted tasks and two off-the-shelf tasks, showing state-of-the-art performance, surpassing all other models except GPT-4. Remarkably, $SusGen$-$GPT$ achieved an average score only 0.02 below GPT-4, despite using models with only 7-8B parameters compared to much larger GPT-4. This demonstrates the efficiency of our approach in delivering high performance with significantly fewer resources, addressing existing challenges and fostering further advancements in the financial and ESG research community."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="wu-etal-2025-susgen">
<titleInfo>
<title>SusGen-GPT: A Data-Centric LLM for Financial NLP and Sustainability Report Generation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Qilong</namePart>
<namePart type="family">Wu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiaoneng</namePart>
<namePart type="family">Xiang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Huang</namePart>
<namePart type="family">Hejia</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xuan</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yeo</namePart>
<namePart type="family">Wei Jie</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ranjan</namePart>
<namePart type="family">Satapathy</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ricardo</namePart>
<namePart type="given">Shirota</namePart>
<namePart type="family">Filho</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bharadwaj</namePart>
<namePart type="family">Veeravalli</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-04</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: NAACL 2025</title>
</titleInfo>
<name type="personal">
<namePart type="given">Luis</namePart>
<namePart type="family">Chiruzzo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alan</namePart>
<namePart type="family">Ritter</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lu</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Albuquerque, New Mexico</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-195-7</identifier>
</relatedItem>
<abstract>The rapid growth of the financial sector and the increasing focus on Environmental, Social, and Governance (ESG) considerations have created a pressing need for advanced natural language processing (NLP) tools. Despite recent advancements, there is still a notable absence of open-source Large Language Models (LLMs) that are proficient across both general finance and ESG domains, such as generating ESG reports. To address this gap, we introduce SusGen-30k, a high-quality, category-balanced dataset comprising seven financial NLP tasks. In addition, we propose TCFD-Bench, a benchmark designed to improve the evaluation of sustainability report generation. Our data-centric approach led to the development of a suite of models, SusGen-GPT, trained on the curated dataset. These models were evaluated across six adapted tasks and two off-the-shelf tasks, showing state-of-the-art performance, surpassing all other models except GPT-4. Remarkably, SusGen-GPT achieved an average score only 0.02 below GPT-4, despite using models with only 7-8B parameters compared to much larger GPT-4. This demonstrates the efficiency of our approach in delivering high performance with significantly fewer resources, addressing existing challenges and fostering further advancements in the financial and ESG research community.</abstract>
<identifier type="citekey">wu-etal-2025-susgen</identifier>
<identifier type="doi">10.18653/v1/2025.findings-naacl.66</identifier>
<location>
<url>https://aclanthology.org/2025.findings-naacl.66/</url>
</location>
<part>
<date>2025-04</date>
<extent unit="page">
<start>1184</start>
<end>1203</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T SusGen-GPT: A Data-Centric LLM for Financial NLP and Sustainability Report Generation
%A Wu, Qilong
%A Xiang, Xiaoneng
%A Hejia, Huang
%A Wang, Xuan
%A Wei Jie, Yeo
%A Satapathy, Ranjan
%A Filho, Ricardo Shirota
%A Veeravalli, Bharadwaj
%Y Chiruzzo, Luis
%Y Ritter, Alan
%Y Wang, Lu
%S Findings of the Association for Computational Linguistics: NAACL 2025
%D 2025
%8 April
%I Association for Computational Linguistics
%C Albuquerque, New Mexico
%@ 979-8-89176-195-7
%F wu-etal-2025-susgen
%X The rapid growth of the financial sector and the increasing focus on Environmental, Social, and Governance (ESG) considerations have created a pressing need for advanced natural language processing (NLP) tools. Despite recent advancements, there is still a notable absence of open-source Large Language Models (LLMs) that are proficient across both general finance and ESG domains, such as generating ESG reports. To address this gap, we introduce SusGen-30k, a high-quality, category-balanced dataset comprising seven financial NLP tasks. In addition, we propose TCFD-Bench, a benchmark designed to improve the evaluation of sustainability report generation. Our data-centric approach led to the development of a suite of models, SusGen-GPT, trained on the curated dataset. These models were evaluated across six adapted tasks and two off-the-shelf tasks, showing state-of-the-art performance, surpassing all other models except GPT-4. Remarkably, SusGen-GPT achieved an average score only 0.02 below GPT-4, despite using models with only 7-8B parameters compared to much larger GPT-4. This demonstrates the efficiency of our approach in delivering high performance with significantly fewer resources, addressing existing challenges and fostering further advancements in the financial and ESG research community.
%R 10.18653/v1/2025.findings-naacl.66
%U https://aclanthology.org/2025.findings-naacl.66/
%U https://doi.org/10.18653/v1/2025.findings-naacl.66
%P 1184-1203
Markdown (Informal)
[SusGen-GPT: A Data-Centric LLM for Financial NLP and Sustainability Report Generation](https://aclanthology.org/2025.findings-naacl.66/) (Wu et al., Findings 2025)
ACL
- Qilong Wu, Xiaoneng Xiang, Huang Hejia, Xuan Wang, Yeo Wei Jie, Ranjan Satapathy, Ricardo Shirota Filho, and Bharadwaj Veeravalli. 2025. SusGen-GPT: A Data-Centric LLM for Financial NLP and Sustainability Report Generation. In Findings of the Association for Computational Linguistics: NAACL 2025, pages 1184–1203, Albuquerque, New Mexico. Association for Computational Linguistics.