@inproceedings{kocbek-etal-2026-graph,
title = "Graph Retrieval in Multimodal {RAG} for Biomedical Documents: A Pharmacokinetics and Pharmacodynamics Case Study",
author = "Kocbek, Primoz and
Me{\v{s}}trovi{\'c}, Katarina and
Huong, Nguyen Thu and
Fajmut, Ales and
Stiglic, Gregor",
editor = "T.Y.S.S., Santosh and
Agarwal, Amit and
Gupta, Vivek and
Kementchedjhieva, Yova and
Garimella, Aparna and
Hou, Yufang",
booktitle = "The Proceedings of Workshop on Document Intelligence and Understanding",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.docinsights-1.14/",
pages = "186--195",
ISBN = "979-8-89176-477-4",
abstract = "Introduction: Biomedical PDFs distribute answer evidence across prose, tables, figures, and cross-document relations, complicating the retrieval of answer-sufficient context. We examine whether graph and multimodal retrieval provide complementary answers and whether direct concatenation exploits this complementarity. Methods: We screen 13 local generators on 205 pharmacokinetics/pharmacodynamics multiple-choice questions and evaluate 11 retrieval conditions over ten runs for five selected models. We quantify questions answered correctly by either or both retrieval routes and compare direct fusion with each standalone route using paired question-level bootstrap intervals. A separate ten-run panel tests answer order, irrelevant context, paraphrasing, generated distractors, and option removal. Results: Multimodal retrieval alone answers 11.5{--}16.2{\%} of questions correctly that graph retrieval misses, while graph retrieval alone answers 7.9{--}13.9{\%} that multimodal retrieval misses. Counting a question as correct whenever either route succeeds yields an oracle upper bound of 69.5{--}77.7{\%}, exceeding direct concatenation by 8.4{--}13.5 percentage points. Direct fusion improves on the stronger standalone route by at most 0.6 points; paired intervals include zero in all ten model{--}depth comparisons. In repeated evaluation, Gemma-4-31B with multimodal retrieval achieved the highest accuracy (66.6{\%}, 95{\%} CI 65.4{--}67.8). Discussion: The central finding is substantial answer-level complementarity that direct concatenation does not reliably exploit. The oracle bound measures retrospective selection, not an implemented routing method. These findings motivate evaluating source-aware routing or reranking to select between complementary retrieval routes."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="kocbek-etal-2026-graph">
<titleInfo>
<title>Graph Retrieval in Multimodal RAG for Biomedical Documents: A Pharmacokinetics and Pharmacodynamics Case Study</title>
</titleInfo>
<name type="personal">
<namePart type="given">Primoz</namePart>
<namePart type="family">Kocbek</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Katarina</namePart>
<namePart type="family">Meštrović</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nguyen</namePart>
<namePart type="given">Thu</namePart>
<namePart type="family">Huong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ales</namePart>
<namePart type="family">Fajmut</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Gregor</namePart>
<namePart type="family">Stiglic</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>The Proceedings of Workshop on Document Intelligence and Understanding</title>
</titleInfo>
<name type="personal">
<namePart type="given">Santosh</namePart>
<namePart type="family">T.Y.S.S.</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Amit</namePart>
<namePart type="family">Agarwal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Vivek</namePart>
<namePart type="family">Gupta</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yova</namePart>
<namePart type="family">Kementchedjhieva</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Aparna</namePart>
<namePart type="family">Garimella</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yufang</namePart>
<namePart type="family">Hou</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">979-8-89176-477-4</identifier>
</relatedItem>
<abstract>Introduction: Biomedical PDFs distribute answer evidence across prose, tables, figures, and cross-document relations, complicating the retrieval of answer-sufficient context. We examine whether graph and multimodal retrieval provide complementary answers and whether direct concatenation exploits this complementarity. Methods: We screen 13 local generators on 205 pharmacokinetics/pharmacodynamics multiple-choice questions and evaluate 11 retrieval conditions over ten runs for five selected models. We quantify questions answered correctly by either or both retrieval routes and compare direct fusion with each standalone route using paired question-level bootstrap intervals. A separate ten-run panel tests answer order, irrelevant context, paraphrasing, generated distractors, and option removal. Results: Multimodal retrieval alone answers 11.5–16.2% of questions correctly that graph retrieval misses, while graph retrieval alone answers 7.9–13.9% that multimodal retrieval misses. Counting a question as correct whenever either route succeeds yields an oracle upper bound of 69.5–77.7%, exceeding direct concatenation by 8.4–13.5 percentage points. Direct fusion improves on the stronger standalone route by at most 0.6 points; paired intervals include zero in all ten model–depth comparisons. In repeated evaluation, Gemma-4-31B with multimodal retrieval achieved the highest accuracy (66.6%, 95% CI 65.4–67.8). Discussion: The central finding is substantial answer-level complementarity that direct concatenation does not reliably exploit. The oracle bound measures retrospective selection, not an implemented routing method. These findings motivate evaluating source-aware routing or reranking to select between complementary retrieval routes.</abstract>
<identifier type="citekey">kocbek-etal-2026-graph</identifier>
<location>
<url>https://aclanthology.org/2026.docinsights-1.14/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>186</start>
<end>195</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Graph Retrieval in Multimodal RAG for Biomedical Documents: A Pharmacokinetics and Pharmacodynamics Case Study
%A Kocbek, Primoz
%A Meštrović, Katarina
%A Huong, Nguyen Thu
%A Fajmut, Ales
%A Stiglic, Gregor
%Y T.Y.S.S., Santosh
%Y Agarwal, Amit
%Y Gupta, Vivek
%Y Kementchedjhieva, Yova
%Y Garimella, Aparna
%Y Hou, Yufang
%S The Proceedings of Workshop on Document Intelligence and Understanding
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-477-4
%F kocbek-etal-2026-graph
%X Introduction: Biomedical PDFs distribute answer evidence across prose, tables, figures, and cross-document relations, complicating the retrieval of answer-sufficient context. We examine whether graph and multimodal retrieval provide complementary answers and whether direct concatenation exploits this complementarity. Methods: We screen 13 local generators on 205 pharmacokinetics/pharmacodynamics multiple-choice questions and evaluate 11 retrieval conditions over ten runs for five selected models. We quantify questions answered correctly by either or both retrieval routes and compare direct fusion with each standalone route using paired question-level bootstrap intervals. A separate ten-run panel tests answer order, irrelevant context, paraphrasing, generated distractors, and option removal. Results: Multimodal retrieval alone answers 11.5–16.2% of questions correctly that graph retrieval misses, while graph retrieval alone answers 7.9–13.9% that multimodal retrieval misses. Counting a question as correct whenever either route succeeds yields an oracle upper bound of 69.5–77.7%, exceeding direct concatenation by 8.4–13.5 percentage points. Direct fusion improves on the stronger standalone route by at most 0.6 points; paired intervals include zero in all ten model–depth comparisons. In repeated evaluation, Gemma-4-31B with multimodal retrieval achieved the highest accuracy (66.6%, 95% CI 65.4–67.8). Discussion: The central finding is substantial answer-level complementarity that direct concatenation does not reliably exploit. The oracle bound measures retrospective selection, not an implemented routing method. These findings motivate evaluating source-aware routing or reranking to select between complementary retrieval routes.
%U https://aclanthology.org/2026.docinsights-1.14/
%P 186-195
Markdown (Informal)
[Graph Retrieval in Multimodal RAG for Biomedical Documents: A Pharmacokinetics and Pharmacodynamics Case Study](https://aclanthology.org/2026.docinsights-1.14/) (Kocbek et al., DocInsights 2026)
ACL