@inproceedings{chen-zhang-2026-f,
title = "{F}-{G}raph{RAG}: Preserving Fine-Grained Facts in Graph Retrieval-Augmented Generation",
author = "Chen, Shenmeng and
Zhang, Kang",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.2253/",
doi = "10.18653/2026.findings-emnlp.2253",
pages = "48219--48237",
ISBN = "979-8-89176-462-0",
abstract = "Retrieval-Augmented Generation (RAG) mitigates factual hallucinations in Large Language Models but struggles with complex reasoning. Graph-based RAG (GraphRAG) addresses this gap yet can exhibit a failure pattern: semantic compression during triple extraction may reduce the preservation of fine-grained facts such as precise quantities, temporal constraints, and entity attributes. We formalize this limitation as the Fine-Grained Fact Retrieval (FGFR) cascade and introduce stage-wise diagnostic metrics, including R{\_}rep, R{\_}sel, and R{\_}end, to quantify information loss. To stress-test FGFR, we construct a diagnostic evaluation suite from classic novels that provide controlled long-context scenarios containing explicit entity attributes and temporal constraints. F-GraphRAG, our framework, features a granularity-aware extractor that preserves atomic details and an information-preserving filter that counters global importance bias. On this evaluation suite, F-GraphRAG outperforms GraphRAG by 12.6{\%} in accuracy while using 33.8{\%} fewer relationships, and achieves gains of 5.3{--}11.2{\%} across nine public benchmarks. These results suggest that improving evidence completeness can provide a more effective alternative to increasing graph connectivity."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="chen-zhang-2026-f">
<titleInfo>
<title>F-GraphRAG: Preserving Fine-Grained Facts in Graph Retrieval-Augmented Generation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Shenmeng</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kang</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>Retrieval-Augmented Generation (RAG) mitigates factual hallucinations in Large Language Models but struggles with complex reasoning. Graph-based RAG (GraphRAG) addresses this gap yet can exhibit a failure pattern: semantic compression during triple extraction may reduce the preservation of fine-grained facts such as precise quantities, temporal constraints, and entity attributes. We formalize this limitation as the Fine-Grained Fact Retrieval (FGFR) cascade and introduce stage-wise diagnostic metrics, including R_rep, R_sel, and R_end, to quantify information loss. To stress-test FGFR, we construct a diagnostic evaluation suite from classic novels that provide controlled long-context scenarios containing explicit entity attributes and temporal constraints. F-GraphRAG, our framework, features a granularity-aware extractor that preserves atomic details and an information-preserving filter that counters global importance bias. On this evaluation suite, F-GraphRAG outperforms GraphRAG by 12.6% in accuracy while using 33.8% fewer relationships, and achieves gains of 5.3–11.2% across nine public benchmarks. These results suggest that improving evidence completeness can provide a more effective alternative to increasing graph connectivity.</abstract>
<identifier type="citekey">chen-zhang-2026-f</identifier>
<identifier type="doi">10.18653/2026.findings-emnlp.2253</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.2253/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>48219</start>
<end>48237</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T F-GraphRAG: Preserving Fine-Grained Facts in Graph Retrieval-Augmented Generation
%A Chen, Shenmeng
%A Zhang, Kang
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F chen-zhang-2026-f
%X Retrieval-Augmented Generation (RAG) mitigates factual hallucinations in Large Language Models but struggles with complex reasoning. Graph-based RAG (GraphRAG) addresses this gap yet can exhibit a failure pattern: semantic compression during triple extraction may reduce the preservation of fine-grained facts such as precise quantities, temporal constraints, and entity attributes. We formalize this limitation as the Fine-Grained Fact Retrieval (FGFR) cascade and introduce stage-wise diagnostic metrics, including R_rep, R_sel, and R_end, to quantify information loss. To stress-test FGFR, we construct a diagnostic evaluation suite from classic novels that provide controlled long-context scenarios containing explicit entity attributes and temporal constraints. F-GraphRAG, our framework, features a granularity-aware extractor that preserves atomic details and an information-preserving filter that counters global importance bias. On this evaluation suite, F-GraphRAG outperforms GraphRAG by 12.6% in accuracy while using 33.8% fewer relationships, and achieves gains of 5.3–11.2% across nine public benchmarks. These results suggest that improving evidence completeness can provide a more effective alternative to increasing graph connectivity.
%R 10.18653/2026.findings-emnlp.2253
%U https://aclanthology.org/2026.findings-emnlp.2253/
%U https://doi.org/10.18653/2026.findings-emnlp.2253
%P 48219-48237
Markdown (Informal)
[F-GraphRAG: Preserving Fine-Grained Facts in Graph Retrieval-Augmented Generation](https://aclanthology.org/2026.findings-emnlp.2253/) (Chen & Zhang, Findings 2026)
ACL