@inproceedings{bhojani-etal-2026-biorag,
title = "{B}io{RAG}: A Systematic Ablation Study of Retrieval Strategies for Biomedical Question Answering",
author = "Bhojani, Krushil and
Waghmare, Mayank and
Nandyala, Hima Bindu",
editor = "Demner-Fushman, Dina and
Ananiadou, Sophia and
Roberts, Kirk and
Tsujii, Junichi",
booktitle = "{B}io{NLP} 2026",
month = jul,
year = "2026",
address = "San Diego, California",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.bionlp-1.10/",
pages = "104--114",
ISBN = "979-8-89176-434-7",
abstract = "Retrieval strategy selection is a critical but understudied design decision in biomedical RAG systems. Existing evaluations rely on lexical metrics that miss answer grounding, or require proprietary infrastructure that limits reproducibility. We present BioRAG, a head-to-head ablation of seven retrieval strategies on BioASQ-13b, evaluated using four RAGAs metrics with a locally deployed judge at zero monetary cost. Hybrid BM25 plus dense retrieval with Reciprocal Rank Fusion achieves faithfulness of 0.534 and context recall of 0.507, improvements of 50{\%} and 85{\%} over naive dense retrieval, confirmed across three random seed re-samples. HyDE improves faithfulness by 14{\%} but reduces context precision by 52{\%}, a tradeoff not previously documented on BioASQ. No single strategy dominates all four metrics, indicating that strategy selection must be application-driven. Sensitivity analysis across k in {\{}3,5,10{\}} confirms ranking stability. A domain mismatch diagnostic confirms 2{\%} corpus coverage failure. The full pipeline runs on consumer hardware without paid APIs, directly addressing BioNLP 2026{'}s emphasis on reproducibility and evaluation frameworks for health-related applications."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="bhojani-etal-2026-biorag">
<titleInfo>
<title>BioRAG: A Systematic Ablation Study of Retrieval Strategies for Biomedical Question Answering</title>
</titleInfo>
<name type="personal">
<namePart type="given">Krushil</namePart>
<namePart type="family">Bhojani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mayank</namePart>
<namePart type="family">Waghmare</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hima</namePart>
<namePart type="given">Bindu</namePart>
<namePart type="family">Nandyala</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>BioNLP 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Dina</namePart>
<namePart type="family">Demner-Fushman</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sophia</namePart>
<namePart type="family">Ananiadou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kirk</namePart>
<namePart type="family">Roberts</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Junichi</namePart>
<namePart type="family">Tsujii</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">San Diego, California</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-434-7</identifier>
</relatedItem>
<abstract>Retrieval strategy selection is a critical but understudied design decision in biomedical RAG systems. Existing evaluations rely on lexical metrics that miss answer grounding, or require proprietary infrastructure that limits reproducibility. We present BioRAG, a head-to-head ablation of seven retrieval strategies on BioASQ-13b, evaluated using four RAGAs metrics with a locally deployed judge at zero monetary cost. Hybrid BM25 plus dense retrieval with Reciprocal Rank Fusion achieves faithfulness of 0.534 and context recall of 0.507, improvements of 50% and 85% over naive dense retrieval, confirmed across three random seed re-samples. HyDE improves faithfulness by 14% but reduces context precision by 52%, a tradeoff not previously documented on BioASQ. No single strategy dominates all four metrics, indicating that strategy selection must be application-driven. Sensitivity analysis across k in {3,5,10} confirms ranking stability. A domain mismatch diagnostic confirms 2% corpus coverage failure. The full pipeline runs on consumer hardware without paid APIs, directly addressing BioNLP 2026’s emphasis on reproducibility and evaluation frameworks for health-related applications.</abstract>
<identifier type="citekey">bhojani-etal-2026-biorag</identifier>
<location>
<url>https://aclanthology.org/2026.bionlp-1.10/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>104</start>
<end>114</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T BioRAG: A Systematic Ablation Study of Retrieval Strategies for Biomedical Question Answering
%A Bhojani, Krushil
%A Waghmare, Mayank
%A Nandyala, Hima Bindu
%Y Demner-Fushman, Dina
%Y Ananiadou, Sophia
%Y Roberts, Kirk
%Y Tsujii, Junichi
%S BioNLP 2026
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California
%@ 979-8-89176-434-7
%F bhojani-etal-2026-biorag
%X Retrieval strategy selection is a critical but understudied design decision in biomedical RAG systems. Existing evaluations rely on lexical metrics that miss answer grounding, or require proprietary infrastructure that limits reproducibility. We present BioRAG, a head-to-head ablation of seven retrieval strategies on BioASQ-13b, evaluated using four RAGAs metrics with a locally deployed judge at zero monetary cost. Hybrid BM25 plus dense retrieval with Reciprocal Rank Fusion achieves faithfulness of 0.534 and context recall of 0.507, improvements of 50% and 85% over naive dense retrieval, confirmed across three random seed re-samples. HyDE improves faithfulness by 14% but reduces context precision by 52%, a tradeoff not previously documented on BioASQ. No single strategy dominates all four metrics, indicating that strategy selection must be application-driven. Sensitivity analysis across k in {3,5,10} confirms ranking stability. A domain mismatch diagnostic confirms 2% corpus coverage failure. The full pipeline runs on consumer hardware without paid APIs, directly addressing BioNLP 2026’s emphasis on reproducibility and evaluation frameworks for health-related applications.
%U https://aclanthology.org/2026.bionlp-1.10/
%P 104-114Markdown (Informal)
[BioRAG: A Systematic Ablation Study of Retrieval Strategies for Biomedical Question Answering](https://aclanthology.org/2026.bionlp-1.10/) (Bhojani et al., BioNLP 2026)
ACL