@inproceedings{yang-etal-2026-divide,
title = "Divide by Question, Conquer by Agent: {SPLIT}-{RAG} with Question-driven graph partitioning",
author = "Yang, Ruiyi and
Xue, Hao and
Razzak, Imran and
Pan, Shirui and
Hacid, Hakim and
Salim, Flora D.",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.2110/",
pages = "48009--48026",
ISBN = "979-8-89176-459-0",
abstract = "Retrieval-augmented generation over knowledge graphs (KGs) can improve the factual grounding of large language models, but efficient retrieval remains challenging when the underlying graph is large and multi-hop evidence is distributed across different regions. We present **SPLIT-RAG**, an efficiency-oriented multi-agent KG-RAG framework that organises a KG into reusable, question-aligned partitions using reasoning patterns observed in training questions. At inference time, SPLIT-RAG forms a query-specific retrieval plan, activates only a small set of relevant partition-specialised agents, and aggregates their retrieved evidence before final answer generation. To improve reliability when evidence from different agents is combined, the framework applies compatibility-based filtering over retrieved triples prior to synthesis. Experiments on WebQSP, CWQ, and MetaQA show that SPLIT-RAG achieves strong performance against recent KG+LLM baselines while reducing end-to-end latency under matched budgets. Component ablations and model-tier analyses further show that selective multi-agent retrieval is central to the performance gain, and that lightweight retrieval agents paired with a stronger head agent provide an effective accuracy{--}latency trade-off."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="yang-etal-2026-divide">
<titleInfo>
<title>Divide by Question, Conquer by Agent: SPLIT-RAG with Question-driven graph partitioning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ruiyi</namePart>
<namePart type="family">Yang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hao</namePart>
<namePart type="family">Xue</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Imran</namePart>
<namePart type="family">Razzak</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shirui</namePart>
<namePart type="family">Pan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hakim</namePart>
<namePart type="family">Hacid</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Flora</namePart>
<namePart type="given">D</namePart>
<namePart type="family">Salim</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>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</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">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>Retrieval-augmented generation over knowledge graphs (KGs) can improve the factual grounding of large language models, but efficient retrieval remains challenging when the underlying graph is large and multi-hop evidence is distributed across different regions. We present **SPLIT-RAG**, an efficiency-oriented multi-agent KG-RAG framework that organises a KG into reusable, question-aligned partitions using reasoning patterns observed in training questions. At inference time, SPLIT-RAG forms a query-specific retrieval plan, activates only a small set of relevant partition-specialised agents, and aggregates their retrieved evidence before final answer generation. To improve reliability when evidence from different agents is combined, the framework applies compatibility-based filtering over retrieved triples prior to synthesis. Experiments on WebQSP, CWQ, and MetaQA show that SPLIT-RAG achieves strong performance against recent KG+LLM baselines while reducing end-to-end latency under matched budgets. Component ablations and model-tier analyses further show that selective multi-agent retrieval is central to the performance gain, and that lightweight retrieval agents paired with a stronger head agent provide an effective accuracy–latency trade-off.</abstract>
<identifier type="citekey">yang-etal-2026-divide</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.2110/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>48009</start>
<end>48026</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Divide by Question, Conquer by Agent: SPLIT-RAG with Question-driven graph partitioning
%A Yang, Ruiyi
%A Xue, Hao
%A Razzak, Imran
%A Pan, Shirui
%A Hacid, Hakim
%A Salim, Flora D.
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F yang-etal-2026-divide
%X Retrieval-augmented generation over knowledge graphs (KGs) can improve the factual grounding of large language models, but efficient retrieval remains challenging when the underlying graph is large and multi-hop evidence is distributed across different regions. We present **SPLIT-RAG**, an efficiency-oriented multi-agent KG-RAG framework that organises a KG into reusable, question-aligned partitions using reasoning patterns observed in training questions. At inference time, SPLIT-RAG forms a query-specific retrieval plan, activates only a small set of relevant partition-specialised agents, and aggregates their retrieved evidence before final answer generation. To improve reliability when evidence from different agents is combined, the framework applies compatibility-based filtering over retrieved triples prior to synthesis. Experiments on WebQSP, CWQ, and MetaQA show that SPLIT-RAG achieves strong performance against recent KG+LLM baselines while reducing end-to-end latency under matched budgets. Component ablations and model-tier analyses further show that selective multi-agent retrieval is central to the performance gain, and that lightweight retrieval agents paired with a stronger head agent provide an effective accuracy–latency trade-off.
%U https://aclanthology.org/2026.emnlp-main.2110/
%P 48009-48026
Markdown (Informal)
[Divide by Question, Conquer by Agent: SPLIT-RAG with Question-driven graph partitioning](https://aclanthology.org/2026.emnlp-main.2110/) (Yang et al., EMNLP 2026)
ACL