@inproceedings{mellah-etal-2022-larsa22,
title = "{LARSA}22 at Qur{'}an {QA} 2022: Text-to-Text Transformer for Finding Answers to Questions from Qur{'}an",
author = "Mellah, Youssef and
Touahri, Ibtissam and
Kaddari, Zakaria and
Haja, Zakaria and
Berrich, Jamal and
Bouchentouf, Toumi",
editor = "Al-Khalifa, Hend and
Elsayed, Tamer and
Mubarak, Hamdy and
Al-Thubaity, Abdulmohsen and
Magdy, Walid and
Darwish, Kareem",
booktitle = "Proceedinsg of the 5th Workshop on Open-Source Arabic Corpora and Processing Tools with Shared Tasks on Qur'an QA and Fine-Grained Hate Speech Detection",
month = jun,
year = "2022",
address = "Marseille, France",
publisher = "European Language Resources Association",
url = "https://aclanthology.org/2022.osact-1.13/",
pages = "112--119",
abstract = "Question Answering (QA) is one of the main fo{\cyrs}uses of Natural Language Pro{\cyrs}essing (NLP) resear{\cyrs}h. However, Arabi{\cyrs} Question Answering is still not within rea{\cyrs}h. The {\cyrs}hallenges of the Arabi{\cyrs} language and the la{\cyrs}k of resour{\cyrs}es have made it diffi{\cyrs}ult to provide powerful Arabi{\cyrs} QA systems with high a{\cyrs}{\cyrs}ura{\cyrs}y. While low a{\cyrs}{\cyrs}ura{\cyrs}y may be a{\cyrs}{\cyrs}epted for general purpose systems, it is {\cyrs}riti{\cyrs}al in some fields su{\cyrs}h as religious affairs. Therefore, there is a need for spe{\cyrs}ialized a{\cyrs}{\cyrs}urate systems that target these {\cyrs}riti{\cyrs}al fields. In this paper, we propose a Transformer-based QA system using the mT5 Language Model (LM). We finetuned the model on the Qur{'}ani{\cyrs} Reading {\CYRS}omprehension Dataset (QR{\CYRS}D) whi{\cyrs}h was provided in the {\cyrs}ontext of the Qur{'}an QA 2022 shared task. The QR{\CYRS}D dataset {\cyrs}onsists of question-passage pairs as input, and the {\cyrs}orresponding adequate answers provided by expert annotators as output. Evaluation results on the same DataSet show that our best model {\cyrs}an a{\cyrs}hieve 0.98 (F1 S{\cyrs}ore) on the Dev Set and 0.40 on the Test Set. We dis{\cyrs}uss those results and {\cyrs}hallenges, then propose potential solutions for possible improvements. The sour{\cyrs}e {\cyrs}ode is available on our repository."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="mellah-etal-2022-larsa22">
<titleInfo>
<title>LARSA22 at Qur’an QA 2022: Text-to-Text Transformer for Finding Answers to Questions from Qur’an</title>
</titleInfo>
<name type="personal">
<namePart type="given">Youssef</namePart>
<namePart type="family">Mellah</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ibtissam</namePart>
<namePart type="family">Touahri</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zakaria</namePart>
<namePart type="family">Kaddari</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zakaria</namePart>
<namePart type="family">Haja</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jamal</namePart>
<namePart type="family">Berrich</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Toumi</namePart>
<namePart type="family">Bouchentouf</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2022-06</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedinsg of the 5th Workshop on Open-Source Arabic Corpora and Processing Tools with Shared Tasks on Qur’an QA and Fine-Grained Hate Speech Detection</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hend</namePart>
<namePart type="family">Al-Khalifa</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tamer</namePart>
<namePart type="family">Elsayed</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hamdy</namePart>
<namePart type="family">Mubarak</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Abdulmohsen</namePart>
<namePart type="family">Al-Thubaity</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Walid</namePart>
<namePart type="family">Magdy</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kareem</namePart>
<namePart type="family">Darwish</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>European Language Resources Association</publisher>
<place>
<placeTerm type="text">Marseille, France</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Question Answering (QA) is one of the main fo\cyrsuses of Natural Language Pro\cyrsessing (NLP) resear\cyrsh. However, Arabi\cyrs Question Answering is still not within rea\cyrsh. The \cyrshallenges of the Arabi\cyrs language and the la\cyrsk of resour\cyrses have made it diffi\cyrsult to provide powerful Arabi\cyrs QA systems with high a\cyrs\cyrsura\cyrsy. While low a\cyrs\cyrsura\cyrsy may be a\cyrs\cyrsepted for general purpose systems, it is \cyrsriti\cyrsal in some fields su\cyrsh as religious affairs. Therefore, there is a need for spe\cyrsialized a\cyrs\cyrsurate systems that target these \cyrsriti\cyrsal fields. In this paper, we propose a Transformer-based QA system using the mT5 Language Model (LM). We finetuned the model on the Qur’ani\cyrs Reading \CYRSomprehension Dataset (QR\CYRSD) whi\cyrsh was provided in the \cyrsontext of the Qur’an QA 2022 shared task. The QR\CYRSD dataset \cyrsonsists of question-passage pairs as input, and the \cyrsorresponding adequate answers provided by expert annotators as output. Evaluation results on the same DataSet show that our best model \cyrsan a\cyrshieve 0.98 (F1 S\cyrsore) on the Dev Set and 0.40 on the Test Set. We dis\cyrsuss those results and \cyrshallenges, then propose potential solutions for possible improvements. The sour\cyrse \cyrsode is available on our repository.</abstract>
<identifier type="citekey">mellah-etal-2022-larsa22</identifier>
<location>
<url>https://aclanthology.org/2022.osact-1.13/</url>
</location>
<part>
<date>2022-06</date>
<extent unit="page">
<start>112</start>
<end>119</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T LARSA22 at Qur’an QA 2022: Text-to-Text Transformer for Finding Answers to Questions from Qur’an
%A Mellah, Youssef
%A Touahri, Ibtissam
%A Kaddari, Zakaria
%A Haja, Zakaria
%A Berrich, Jamal
%A Bouchentouf, Toumi
%Y Al-Khalifa, Hend
%Y Elsayed, Tamer
%Y Mubarak, Hamdy
%Y Al-Thubaity, Abdulmohsen
%Y Magdy, Walid
%Y Darwish, Kareem
%S Proceedinsg of the 5th Workshop on Open-Source Arabic Corpora and Processing Tools with Shared Tasks on Qur’an QA and Fine-Grained Hate Speech Detection
%D 2022
%8 June
%I European Language Resources Association
%C Marseille, France
%F mellah-etal-2022-larsa22
%X Question Answering (QA) is one of the main fo\cyrsuses of Natural Language Pro\cyrsessing (NLP) resear\cyrsh. However, Arabi\cyrs Question Answering is still not within rea\cyrsh. The \cyrshallenges of the Arabi\cyrs language and the la\cyrsk of resour\cyrses have made it diffi\cyrsult to provide powerful Arabi\cyrs QA systems with high a\cyrs\cyrsura\cyrsy. While low a\cyrs\cyrsura\cyrsy may be a\cyrs\cyrsepted for general purpose systems, it is \cyrsriti\cyrsal in some fields su\cyrsh as religious affairs. Therefore, there is a need for spe\cyrsialized a\cyrs\cyrsurate systems that target these \cyrsriti\cyrsal fields. In this paper, we propose a Transformer-based QA system using the mT5 Language Model (LM). We finetuned the model on the Qur’ani\cyrs Reading \CYRSomprehension Dataset (QR\CYRSD) whi\cyrsh was provided in the \cyrsontext of the Qur’an QA 2022 shared task. The QR\CYRSD dataset \cyrsonsists of question-passage pairs as input, and the \cyrsorresponding adequate answers provided by expert annotators as output. Evaluation results on the same DataSet show that our best model \cyrsan a\cyrshieve 0.98 (F1 S\cyrsore) on the Dev Set and 0.40 on the Test Set. We dis\cyrsuss those results and \cyrshallenges, then propose potential solutions for possible improvements. The sour\cyrse \cyrsode is available on our repository.
%U https://aclanthology.org/2022.osact-1.13/
%P 112-119
Markdown (Informal)
[LARSA22 at Qur’an QA 2022: Text-to-Text Transformer for Finding Answers to Questions from Qur’an](https://aclanthology.org/2022.osact-1.13/) (Mellah et al., OSACT 2022)
ACL