@inproceedings{li-etal-2026-construction,
title = "Construction of a {J}apanese {RAG} Benchmark Using Synthetic Documents on Non-existent Entities and Events",
author = "Li, Shengzhe and
Ohagi, Masaya and
Tsukagoshi, Hayato and
Fukuchi, Akihiko and
Shibata, Tomohide and
Kawahara, Daisuke",
editor = "Piperidis, Stelios and
Bel, N{\'u}ria and
van den Heuvel, Henk and
Ide, Nancy and
Krek, Simon and
Toral, Antonio",
booktitle = "Proceedings of the Fifteenth Language Resources and Evaluation Conference",
month = may,
year = "2026",
address = "Palma de Mallorca, Spain",
publisher = "ELRA Language Resource Association",
url = "https://aclanthology.org/2026.lrec-1.589/",
doi = "10.63317/443h4s9sm3gy",
pages = "7435--7445",
abstract = "Retrieval-augmented generation (RAG) is a technique in which a large language model (LLM) generates answers based on relevant documents retrieved from an external document collection. Existing RAG evaluation benchmarks often use public data, such as Wikipedia and news articles, as the external document collection. However, these data are highly likely to be already included in the LLM{'}s pre-training corpus, which may prevent an accurate evaluation of the model{'}s ability to generate answers based on the retrieved documents. In this study, we construct a Japanese RAG benchmark by having an LLM synthesize documents about non-existent entities and events and use this collection of synthetic documents as the search target. Since these synthetic documents are not included in the LLM{'}s training data, the ability to generate answers based on retrieved documents can be evaluated more accurately. In addition to the synthetic documents, the benchmark is composed of questions and correct answers, which are created using a combination of LLMs and human effort. We then evaluated and analyzed the RAG performance of existing LLMs using the constructed benchmark."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="li-etal-2026-construction">
<titleInfo>
<title>Construction of a Japanese RAG Benchmark Using Synthetic Documents on Non-existent Entities and Events</title>
</titleInfo>
<name type="personal">
<namePart type="given">Shengzhe</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Masaya</namePart>
<namePart type="family">Ohagi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hayato</namePart>
<namePart type="family">Tsukagoshi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Akihiko</namePart>
<namePart type="family">Fukuchi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tomohide</namePart>
<namePart type="family">Shibata</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Daisuke</namePart>
<namePart type="family">Kawahara</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-05</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Fifteenth Language Resources and Evaluation Conference</title>
</titleInfo>
<name type="personal">
<namePart type="given">Stelios</namePart>
<namePart type="family">Piperidis</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Núria</namePart>
<namePart type="family">Bel</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Henk</namePart>
<namePart type="family">van den Heuvel</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nancy</namePart>
<namePart type="family">Ide</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Simon</namePart>
<namePart type="family">Krek</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Antonio</namePart>
<namePart type="family">Toral</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>ELRA Language Resource Association</publisher>
<place>
<placeTerm type="text">Palma de Mallorca, Spain</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Retrieval-augmented generation (RAG) is a technique in which a large language model (LLM) generates answers based on relevant documents retrieved from an external document collection. Existing RAG evaluation benchmarks often use public data, such as Wikipedia and news articles, as the external document collection. However, these data are highly likely to be already included in the LLM’s pre-training corpus, which may prevent an accurate evaluation of the model’s ability to generate answers based on the retrieved documents. In this study, we construct a Japanese RAG benchmark by having an LLM synthesize documents about non-existent entities and events and use this collection of synthetic documents as the search target. Since these synthetic documents are not included in the LLM’s training data, the ability to generate answers based on retrieved documents can be evaluated more accurately. In addition to the synthetic documents, the benchmark is composed of questions and correct answers, which are created using a combination of LLMs and human effort. We then evaluated and analyzed the RAG performance of existing LLMs using the constructed benchmark.</abstract>
<identifier type="citekey">li-etal-2026-construction</identifier>
<identifier type="doi">10.63317/443h4s9sm3gy</identifier>
<location>
<url>https://aclanthology.org/2026.lrec-1.589/</url>
</location>
<part>
<date>2026-05</date>
<extent unit="page">
<start>7435</start>
<end>7445</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Construction of a Japanese RAG Benchmark Using Synthetic Documents on Non-existent Entities and Events
%A Li, Shengzhe
%A Ohagi, Masaya
%A Tsukagoshi, Hayato
%A Fukuchi, Akihiko
%A Shibata, Tomohide
%A Kawahara, Daisuke
%Y Piperidis, Stelios
%Y Bel, Núria
%Y van den Heuvel, Henk
%Y Ide, Nancy
%Y Krek, Simon
%Y Toral, Antonio
%S Proceedings of the Fifteenth Language Resources and Evaluation Conference
%D 2026
%8 May
%I ELRA Language Resource Association
%C Palma de Mallorca, Spain
%F li-etal-2026-construction
%X Retrieval-augmented generation (RAG) is a technique in which a large language model (LLM) generates answers based on relevant documents retrieved from an external document collection. Existing RAG evaluation benchmarks often use public data, such as Wikipedia and news articles, as the external document collection. However, these data are highly likely to be already included in the LLM’s pre-training corpus, which may prevent an accurate evaluation of the model’s ability to generate answers based on the retrieved documents. In this study, we construct a Japanese RAG benchmark by having an LLM synthesize documents about non-existent entities and events and use this collection of synthetic documents as the search target. Since these synthetic documents are not included in the LLM’s training data, the ability to generate answers based on retrieved documents can be evaluated more accurately. In addition to the synthetic documents, the benchmark is composed of questions and correct answers, which are created using a combination of LLMs and human effort. We then evaluated and analyzed the RAG performance of existing LLMs using the constructed benchmark.
%R 10.63317/443h4s9sm3gy
%U https://aclanthology.org/2026.lrec-1.589/
%U https://doi.org/10.63317/443h4s9sm3gy
%P 7435-7445
Markdown (Informal)
[Construction of a Japanese RAG Benchmark Using Synthetic Documents on Non-existent Entities and Events](https://aclanthology.org/2026.lrec-1.589/) (Li et al., LREC 2026)
ACL