@article{siriwardhana-etal-2023-improving,
title = "Improving the Domain Adaptation of Retrieval Augmented Generation ({RAG}) Models for Open Domain Question Answering",
author = "Siriwardhana, Shamane and
Weerasekera, Rivindu and
Wen, Elliott and
Kaluarachchi, Tharindu and
Rana, Rajib and
Nanayakkara, Suranga",
journal = "Transactions of the Association for Computational Linguistics",
volume = "11",
year = "2023",
address = "Cambridge, MA",
publisher = "MIT Press",
url = "https://aclanthology.org/2023.tacl-1.1",
doi = "10.1162/tacl_a_00530",
pages = "1--17",
abstract = "Retrieval Augment Generation (RAG) is a recent advancement in Open-Domain Question Answering (ODQA). RAG has only been trained and explored with a Wikipedia-based external knowledge base and is not optimized for use in other specialized domains such as healthcare and news. In this paper, we evaluate the impact of joint training of the retriever and generator components of RAG for the task of domain adaptation in ODQA. We propose RAG-end2end, an extension to RAG that can adapt to a domain-specific knowledge base by updating all components of the external knowledge base during training. In addition, we introduce an auxiliary training signal to inject more domain-specific knowledge. This auxiliary signal forces RAG-end2end to reconstruct a given sentence by accessing the relevant information from the external knowledge base. Our novel contribution is that, unlike RAG, RAG-end2end does joint training of the retriever and generator for the end QA task and domain adaptation. We evaluate our approach with datasets from three domains: COVID-19, News, and Conversations, and achieve significant performance improvements compared to the original RAG model. Our work has been open-sourced through the HuggingFace Transformers library, attesting to our work{'}s credibility and technical consistency.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="siriwardhana-etal-2023-improving">
<titleInfo>
<title>Improving the Domain Adaptation of Retrieval Augmented Generation (RAG) Models for Open Domain Question Answering</title>
</titleInfo>
<name type="personal">
<namePart type="given">Shamane</namePart>
<namePart type="family">Siriwardhana</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Rivindu</namePart>
<namePart type="family">Weerasekera</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Elliott</namePart>
<namePart type="family">Wen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tharindu</namePart>
<namePart type="family">Kaluarachchi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Rajib</namePart>
<namePart type="family">Rana</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Suranga</namePart>
<namePart type="family">Nanayakkara</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2023</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<genre authority="bibutilsgt">journal article</genre>
<relatedItem type="host">
<titleInfo>
<title>Transactions of the Association for Computational Linguistics</title>
</titleInfo>
<originInfo>
<issuance>continuing</issuance>
<publisher>MIT Press</publisher>
<place>
<placeTerm type="text">Cambridge, MA</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">periodical</genre>
<genre authority="bibutilsgt">academic journal</genre>
</relatedItem>
<abstract>Retrieval Augment Generation (RAG) is a recent advancement in Open-Domain Question Answering (ODQA). RAG has only been trained and explored with a Wikipedia-based external knowledge base and is not optimized for use in other specialized domains such as healthcare and news. In this paper, we evaluate the impact of joint training of the retriever and generator components of RAG for the task of domain adaptation in ODQA. We propose RAG-end2end, an extension to RAG that can adapt to a domain-specific knowledge base by updating all components of the external knowledge base during training. In addition, we introduce an auxiliary training signal to inject more domain-specific knowledge. This auxiliary signal forces RAG-end2end to reconstruct a given sentence by accessing the relevant information from the external knowledge base. Our novel contribution is that, unlike RAG, RAG-end2end does joint training of the retriever and generator for the end QA task and domain adaptation. We evaluate our approach with datasets from three domains: COVID-19, News, and Conversations, and achieve significant performance improvements compared to the original RAG model. Our work has been open-sourced through the HuggingFace Transformers library, attesting to our work’s credibility and technical consistency.</abstract>
<identifier type="citekey">siriwardhana-etal-2023-improving</identifier>
<identifier type="doi">10.1162/tacl_a_00530</identifier>
<location>
<url>https://aclanthology.org/2023.tacl-1.1</url>
</location>
<part>
<date>2023</date>
<detail type="volume"><number>11</number></detail>
<extent unit="page">
<start>1</start>
<end>17</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Journal Article
%T Improving the Domain Adaptation of Retrieval Augmented Generation (RAG) Models for Open Domain Question Answering
%A Siriwardhana, Shamane
%A Weerasekera, Rivindu
%A Wen, Elliott
%A Kaluarachchi, Tharindu
%A Rana, Rajib
%A Nanayakkara, Suranga
%J Transactions of the Association for Computational Linguistics
%D 2023
%V 11
%I MIT Press
%C Cambridge, MA
%F siriwardhana-etal-2023-improving
%X Retrieval Augment Generation (RAG) is a recent advancement in Open-Domain Question Answering (ODQA). RAG has only been trained and explored with a Wikipedia-based external knowledge base and is not optimized for use in other specialized domains such as healthcare and news. In this paper, we evaluate the impact of joint training of the retriever and generator components of RAG for the task of domain adaptation in ODQA. We propose RAG-end2end, an extension to RAG that can adapt to a domain-specific knowledge base by updating all components of the external knowledge base during training. In addition, we introduce an auxiliary training signal to inject more domain-specific knowledge. This auxiliary signal forces RAG-end2end to reconstruct a given sentence by accessing the relevant information from the external knowledge base. Our novel contribution is that, unlike RAG, RAG-end2end does joint training of the retriever and generator for the end QA task and domain adaptation. We evaluate our approach with datasets from three domains: COVID-19, News, and Conversations, and achieve significant performance improvements compared to the original RAG model. Our work has been open-sourced through the HuggingFace Transformers library, attesting to our work’s credibility and technical consistency.
%R 10.1162/tacl_a_00530
%U https://aclanthology.org/2023.tacl-1.1
%U https://doi.org/10.1162/tacl_a_00530
%P 1-17
Markdown (Informal)
[Improving the Domain Adaptation of Retrieval Augmented Generation (RAG) Models for Open Domain Question Answering](https://aclanthology.org/2023.tacl-1.1) (Siriwardhana et al., TACL 2023)
ACL