@inproceedings{luo-etal-2022-choose,
title = "Choose Your {QA} Model Wisely: A Systematic Study of Generative and Extractive Readers for Question Answering",
author = "Luo, Man and
Hashimoto, Kazuma and
Yavuz, Semih and
Liu, Zhiwei and
Baral, Chitta and
Zhou, Yingbo",
editor = "Das, Rajarshi and
Lewis, Patrick and
Min, Sewon and
Thai, June and
Zaheer, Manzil",
booktitle = "Proceedings of the 1st Workshop on Semiparametric Methods in NLP: Decoupling Logic from Knowledge",
month = may,
year = "2022",
address = "Dublin, Ireland and Online",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2022.spanlp-1.2/",
doi = "10.18653/v1/2022.spanlp-1.2",
pages = "7--22",
abstract = "While both extractive and generative readers have been successfully applied to the Question Answering (QA) task, little attention has been paid toward the systematic comparison of them. Characterizing the strengths and weaknesses of the two readers is crucial not only for making a more informed reader selection in practice but also for developing a deeper understanding to foster further research on improving readers in a principled manner. Motivated by this goal, we make the first attempt to systematically study the comparison of extractive and generative readers for question answering. To be aligned with the state-of-the-art, we explore nine transformer-based large pre-trained language models (PrLMs) as backbone architectures. Furthermore, we organize our findings under two main categories: (1) keeping the architecture invariant, and (2) varying the underlying PrLMs. Among several interesting findings, it is important to highlight that (1) the generative readers perform better in long context QA, (2) the extractive readers perform better in short context while also showing better out-of-domain generalization, and (3) the encoder of encoder-decoder PrLMs (e.g., T5) turns out to be a strong extractive reader and outperforms the standard choice of encoder-only PrLMs (e.g., RoBERTa). We also study the effect of multi-task learning on the two types of readers varying the underlying PrLMs and perform qualitative and quantitative diagnosis to provide further insights into future directions in modeling better readers."
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="luo-etal-2022-choose">
<titleInfo>
<title>Choose Your QA Model Wisely: A Systematic Study of Generative and Extractive Readers for Question Answering</title>
</titleInfo>
<name type="personal">
<namePart type="given">Man</namePart>
<namePart type="family">Luo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kazuma</namePart>
<namePart type="family">Hashimoto</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Semih</namePart>
<namePart type="family">Yavuz</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhiwei</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chitta</namePart>
<namePart type="family">Baral</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yingbo</namePart>
<namePart type="family">Zhou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2022-05</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 1st Workshop on Semiparametric Methods in NLP: Decoupling Logic from Knowledge</title>
</titleInfo>
<name type="personal">
<namePart type="given">Rajarshi</namePart>
<namePart type="family">Das</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Patrick</namePart>
<namePart type="family">Lewis</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sewon</namePart>
<namePart type="family">Min</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">June</namePart>
<namePart type="family">Thai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Manzil</namePart>
<namePart type="family">Zaheer</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Dublin, Ireland and Online</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>While both extractive and generative readers have been successfully applied to the Question Answering (QA) task, little attention has been paid toward the systematic comparison of them. Characterizing the strengths and weaknesses of the two readers is crucial not only for making a more informed reader selection in practice but also for developing a deeper understanding to foster further research on improving readers in a principled manner. Motivated by this goal, we make the first attempt to systematically study the comparison of extractive and generative readers for question answering. To be aligned with the state-of-the-art, we explore nine transformer-based large pre-trained language models (PrLMs) as backbone architectures. Furthermore, we organize our findings under two main categories: (1) keeping the architecture invariant, and (2) varying the underlying PrLMs. Among several interesting findings, it is important to highlight that (1) the generative readers perform better in long context QA, (2) the extractive readers perform better in short context while also showing better out-of-domain generalization, and (3) the encoder of encoder-decoder PrLMs (e.g., T5) turns out to be a strong extractive reader and outperforms the standard choice of encoder-only PrLMs (e.g., RoBERTa). We also study the effect of multi-task learning on the two types of readers varying the underlying PrLMs and perform qualitative and quantitative diagnosis to provide further insights into future directions in modeling better readers.</abstract>
<identifier type="citekey">luo-etal-2022-choose</identifier>
<identifier type="doi">10.18653/v1/2022.spanlp-1.2</identifier>
<location>
<url>https://aclanthology.org/2022.spanlp-1.2/</url>
</location>
<part>
<date>2022-05</date>
<extent unit="page">
<start>7</start>
<end>22</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Choose Your QA Model Wisely: A Systematic Study of Generative and Extractive Readers for Question Answering
%A Luo, Man
%A Hashimoto, Kazuma
%A Yavuz, Semih
%A Liu, Zhiwei
%A Baral, Chitta
%A Zhou, Yingbo
%Y Das, Rajarshi
%Y Lewis, Patrick
%Y Min, Sewon
%Y Thai, June
%Y Zaheer, Manzil
%S Proceedings of the 1st Workshop on Semiparametric Methods in NLP: Decoupling Logic from Knowledge
%D 2022
%8 May
%I Association for Computational Linguistics
%C Dublin, Ireland and Online
%F luo-etal-2022-choose
%X While both extractive and generative readers have been successfully applied to the Question Answering (QA) task, little attention has been paid toward the systematic comparison of them. Characterizing the strengths and weaknesses of the two readers is crucial not only for making a more informed reader selection in practice but also for developing a deeper understanding to foster further research on improving readers in a principled manner. Motivated by this goal, we make the first attempt to systematically study the comparison of extractive and generative readers for question answering. To be aligned with the state-of-the-art, we explore nine transformer-based large pre-trained language models (PrLMs) as backbone architectures. Furthermore, we organize our findings under two main categories: (1) keeping the architecture invariant, and (2) varying the underlying PrLMs. Among several interesting findings, it is important to highlight that (1) the generative readers perform better in long context QA, (2) the extractive readers perform better in short context while also showing better out-of-domain generalization, and (3) the encoder of encoder-decoder PrLMs (e.g., T5) turns out to be a strong extractive reader and outperforms the standard choice of encoder-only PrLMs (e.g., RoBERTa). We also study the effect of multi-task learning on the two types of readers varying the underlying PrLMs and perform qualitative and quantitative diagnosis to provide further insights into future directions in modeling better readers.
%R 10.18653/v1/2022.spanlp-1.2
%U https://aclanthology.org/2022.spanlp-1.2/
%U https://doi.org/10.18653/v1/2022.spanlp-1.2
%P 7-22
Markdown (Informal)
[Choose Your QA Model Wisely: A Systematic Study of Generative and Extractive Readers for Question Answering](https://aclanthology.org/2022.spanlp-1.2/) (Luo et al., SpaNLP 2022)
ACL