@inproceedings{ji-etal-2026-medragchecker,
title = "{M}ed{RAGC}hecker: Claim-Level Verification for Biomedical Retrieval-Augmented Generation",
author = "Ji, Yuelyu and
Kwak, Min Gu and
Zhang, Hang and
Wu, Xizhi and
Li, Chenyu and
Wang, Yanshan",
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.399/",
pages = "8431--8454",
ISBN = "979-8-89176-462-0",
abstract = "Biomedical retrieval-augmented generation (RAG) produces long-form answers that mix supported, under-evidenced, and contradicted claims, demanding fine-grained verification. We propose MedRAGChecker, a claim-level diagnostic framework that combines textual natural language inference (NLI) with a biomedical knowledge-graph (KG) compatibility signal. Its asymmetric selective-gating rule preserves the textual prediction for most claims and applies KG adjustment only to confident textual entailments assigned low plausibility by the KG back-end. We distill GPT-4.1 claim decompositions and NLI pseudo-labels into compact student models and evaluate three biomedical generators on PubMedQA, MedQuAD, LiveQA, and MedRedQA. The student-ensemble results show that selective BioPortal gating increases macro-averaged faithfulness from 60.8{\%} to 64.4{\%} and reduces the contradiction rate from 19.9{\%} to 15.0{\%}. On 172 annotations of 86 claims whose decisions change after fusion, agreement with human evidence judgments increases from 63.4{\%} to 69.8{\%}. A comparison with DRKG shows that fusion behavior varies substantially with KG coverage, entity linking, score construction, and calibration; the comparison therefore characterizes back-end effects rather than isolating a causal effect of coverage."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ji-etal-2026-medragchecker">
<titleInfo>
<title>MedRAGChecker: Claim-Level Verification for Biomedical Retrieval-Augmented Generation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yuelyu</namePart>
<namePart type="family">Ji</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Min</namePart>
<namePart type="given">Gu</namePart>
<namePart type="family">Kwak</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hang</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xizhi</namePart>
<namePart type="family">Wu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chenyu</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yanshan</namePart>
<namePart type="family">Wang</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>Biomedical retrieval-augmented generation (RAG) produces long-form answers that mix supported, under-evidenced, and contradicted claims, demanding fine-grained verification. We propose MedRAGChecker, a claim-level diagnostic framework that combines textual natural language inference (NLI) with a biomedical knowledge-graph (KG) compatibility signal. Its asymmetric selective-gating rule preserves the textual prediction for most claims and applies KG adjustment only to confident textual entailments assigned low plausibility by the KG back-end. We distill GPT-4.1 claim decompositions and NLI pseudo-labels into compact student models and evaluate three biomedical generators on PubMedQA, MedQuAD, LiveQA, and MedRedQA. The student-ensemble results show that selective BioPortal gating increases macro-averaged faithfulness from 60.8% to 64.4% and reduces the contradiction rate from 19.9% to 15.0%. On 172 annotations of 86 claims whose decisions change after fusion, agreement with human evidence judgments increases from 63.4% to 69.8%. A comparison with DRKG shows that fusion behavior varies substantially with KG coverage, entity linking, score construction, and calibration; the comparison therefore characterizes back-end effects rather than isolating a causal effect of coverage.</abstract>
<identifier type="citekey">ji-etal-2026-medragchecker</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.399/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>8431</start>
<end>8454</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T MedRAGChecker: Claim-Level Verification for Biomedical Retrieval-Augmented Generation
%A Ji, Yuelyu
%A Kwak, Min Gu
%A Zhang, Hang
%A Wu, Xizhi
%A Li, Chenyu
%A Wang, Yanshan
%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 ji-etal-2026-medragchecker
%X Biomedical retrieval-augmented generation (RAG) produces long-form answers that mix supported, under-evidenced, and contradicted claims, demanding fine-grained verification. We propose MedRAGChecker, a claim-level diagnostic framework that combines textual natural language inference (NLI) with a biomedical knowledge-graph (KG) compatibility signal. Its asymmetric selective-gating rule preserves the textual prediction for most claims and applies KG adjustment only to confident textual entailments assigned low plausibility by the KG back-end. We distill GPT-4.1 claim decompositions and NLI pseudo-labels into compact student models and evaluate three biomedical generators on PubMedQA, MedQuAD, LiveQA, and MedRedQA. The student-ensemble results show that selective BioPortal gating increases macro-averaged faithfulness from 60.8% to 64.4% and reduces the contradiction rate from 19.9% to 15.0%. On 172 annotations of 86 claims whose decisions change after fusion, agreement with human evidence judgments increases from 63.4% to 69.8%. A comparison with DRKG shows that fusion behavior varies substantially with KG coverage, entity linking, score construction, and calibration; the comparison therefore characterizes back-end effects rather than isolating a causal effect of coverage.
%U https://aclanthology.org/2026.findings-emnlp.399/
%P 8431-8454
Markdown (Informal)
[MedRAGChecker: Claim-Level Verification for Biomedical Retrieval-Augmented Generation](https://aclanthology.org/2026.findings-emnlp.399/) (Ji et al., Findings 2026)
ACL