@inproceedings{barry-etal-2025-graphrag,
title = "{G}raph{RAG}: Leveraging Graph-Based Efficiency to Minimize Hallucinations in {LLM}-Driven {RAG} for Finance Data",
author = "Barry, Mariam and
Caillaut, Gaetan and
Halftermeyer, Pierre and
Qader, Raheel and
Mouayad, Mehdi and
Le Deit, Fabrice and
Cariolaro, Dimitri and
Gesnouin, Joseph",
editor = "Gesese, Genet Asefa and
Sack, Harald and
Paulheim, Heiko and
Merono-Penuela, Albert and
Chen, Lihu",
booktitle = "Proceedings of the Workshop on Generative AI and Knowledge Graphs (GenAIK)",
month = jan,
year = "2025",
address = "Abu Dhabi, UAE",
publisher = "International Committee on Computational Linguistics",
url = "https://aclanthology.org/2025.genaik-1.6/",
pages = "54--65",
abstract = "This study explores the integration of graph-based methods into Retrieval-Augmented Generation (RAG) systems to enhance efficiency, reduce hallucinations, and improve explainability, with a particular focus on financial and regulatory document retrieval. We propose two strategies{---}FactRAG and HybridRAG{---}which leverage knowledge graphs to improve RAG performance. Experiments conducted using Finance Bench, a benchmark for AI in finance, demonstrate that these approaches achieve a 6{\%} reduction in hallucinations and an 80{\%} decrease in token usage compared to conventional RAG methods. Furthermore, we evaluate HybridRAG by comparing the Digital Operational Resilience Act (DORA) from the European Union with the Federal Financial Institutions Examination Council (FFIEC) guidelines from the United States. The results reveal a significant improvement in computational efficiency, reducing contradiction detection complexity from $O(n^2)$ to $O(k \cdot n)${---}where $n$ is the number of chunks{---}and a remarkable 734-fold decrease in token consumption. Graph-based retrieval methods can improve the efficiency and cost-effectiveness of large language model (LLM) applications, though their performance and token usage depend on the dataset, knowledge graph design, and retrieval task."
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="barry-etal-2025-graphrag">
<titleInfo>
<title>GraphRAG: Leveraging Graph-Based Efficiency to Minimize Hallucinations in LLM-Driven RAG for Finance Data</title>
</titleInfo>
<name type="personal">
<namePart type="given">Mariam</namePart>
<namePart type="family">Barry</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Gaetan</namePart>
<namePart type="family">Caillaut</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Pierre</namePart>
<namePart type="family">Halftermeyer</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Raheel</namePart>
<namePart type="family">Qader</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mehdi</namePart>
<namePart type="family">Mouayad</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Fabrice</namePart>
<namePart type="family">Le Deit</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dimitri</namePart>
<namePart type="family">Cariolaro</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Joseph</namePart>
<namePart type="family">Gesnouin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-01</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Workshop on Generative AI and Knowledge Graphs (GenAIK)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Genet</namePart>
<namePart type="given">Asefa</namePart>
<namePart type="family">Gesese</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Harald</namePart>
<namePart type="family">Sack</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Heiko</namePart>
<namePart type="family">Paulheim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Albert</namePart>
<namePart type="family">Merono-Penuela</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lihu</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>International Committee on Computational Linguistics</publisher>
<place>
<placeTerm type="text">Abu Dhabi, UAE</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>This study explores the integration of graph-based methods into Retrieval-Augmented Generation (RAG) systems to enhance efficiency, reduce hallucinations, and improve explainability, with a particular focus on financial and regulatory document retrieval. We propose two strategies—FactRAG and HybridRAG—which leverage knowledge graphs to improve RAG performance. Experiments conducted using Finance Bench, a benchmark for AI in finance, demonstrate that these approaches achieve a 6% reduction in hallucinations and an 80% decrease in token usage compared to conventional RAG methods. Furthermore, we evaluate HybridRAG by comparing the Digital Operational Resilience Act (DORA) from the European Union with the Federal Financial Institutions Examination Council (FFIEC) guidelines from the United States. The results reveal a significant improvement in computational efficiency, reducing contradiction detection complexity from O(n²) to O(k · n)—where n is the number of chunks—and a remarkable 734-fold decrease in token consumption. Graph-based retrieval methods can improve the efficiency and cost-effectiveness of large language model (LLM) applications, though their performance and token usage depend on the dataset, knowledge graph design, and retrieval task.</abstract>
<identifier type="citekey">barry-etal-2025-graphrag</identifier>
<location>
<url>https://aclanthology.org/2025.genaik-1.6/</url>
</location>
<part>
<date>2025-01</date>
<extent unit="page">
<start>54</start>
<end>65</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T GraphRAG: Leveraging Graph-Based Efficiency to Minimize Hallucinations in LLM-Driven RAG for Finance Data
%A Barry, Mariam
%A Caillaut, Gaetan
%A Halftermeyer, Pierre
%A Qader, Raheel
%A Mouayad, Mehdi
%A Le Deit, Fabrice
%A Cariolaro, Dimitri
%A Gesnouin, Joseph
%Y Gesese, Genet Asefa
%Y Sack, Harald
%Y Paulheim, Heiko
%Y Merono-Penuela, Albert
%Y Chen, Lihu
%S Proceedings of the Workshop on Generative AI and Knowledge Graphs (GenAIK)
%D 2025
%8 January
%I International Committee on Computational Linguistics
%C Abu Dhabi, UAE
%F barry-etal-2025-graphrag
%X This study explores the integration of graph-based methods into Retrieval-Augmented Generation (RAG) systems to enhance efficiency, reduce hallucinations, and improve explainability, with a particular focus on financial and regulatory document retrieval. We propose two strategies—FactRAG and HybridRAG—which leverage knowledge graphs to improve RAG performance. Experiments conducted using Finance Bench, a benchmark for AI in finance, demonstrate that these approaches achieve a 6% reduction in hallucinations and an 80% decrease in token usage compared to conventional RAG methods. Furthermore, we evaluate HybridRAG by comparing the Digital Operational Resilience Act (DORA) from the European Union with the Federal Financial Institutions Examination Council (FFIEC) guidelines from the United States. The results reveal a significant improvement in computational efficiency, reducing contradiction detection complexity from O(n²) to O(k · n)—where n is the number of chunks—and a remarkable 734-fold decrease in token consumption. Graph-based retrieval methods can improve the efficiency and cost-effectiveness of large language model (LLM) applications, though their performance and token usage depend on the dataset, knowledge graph design, and retrieval task.
%U https://aclanthology.org/2025.genaik-1.6/
%P 54-65
Markdown (Informal)
[GraphRAG: Leveraging Graph-Based Efficiency to Minimize Hallucinations in LLM-Driven RAG for Finance Data](https://aclanthology.org/2025.genaik-1.6/) (Barry et al., GenAIK 2025)
ACL
- Mariam Barry, Gaetan Caillaut, Pierre Halftermeyer, Raheel Qader, Mehdi Mouayad, Fabrice Le Deit, Dimitri Cariolaro, and Joseph Gesnouin. 2025. GraphRAG: Leveraging Graph-Based Efficiency to Minimize Hallucinations in LLM-Driven RAG for Finance Data. In Proceedings of the Workshop on Generative AI and Knowledge Graphs (GenAIK), pages 54–65, Abu Dhabi, UAE. International Committee on Computational Linguistics.