@inproceedings{hung-etal-2022-zusammenqa,
title = "{Z}usammen{QA}: Data Augmentation with Specialized Models for Cross-lingual Open-retrieval Question Answering System",
author = "Hung, Chia-Chien and
Green, Tommaso and
Litschko, Robert and
Tsereteli, Tornike and
Takeshita, Sotaro and
Bombieri, Marco and
Glava{\v{s}}, Goran and
Ponzetto, Simone Paolo",
editor = "Asai, Akari and
Choi, Eunsol and
Clark, Jonathan H. and
Hu, Junjie and
Lee, Chia-Hsuan and
Kasai, Jungo and
Longpre, Shayne and
Yamada, Ikuya and
Zhang, Rui",
booktitle = "Proceedings of the Workshop on Multilingual Information Access (MIA)",
month = jul,
year = "2022",
address = "Seattle, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2022.mia-1.8",
doi = "10.18653/v1/2022.mia-1.8",
pages = "77--90",
abstract = "This paper introduces our proposed system for the MIA Shared Task on Cross-lingual Openretrieval Question Answering (COQA). In this challenging scenario, given an input question the system has to gather evidence documents from a multilingual pool and generate from them an answer in the language of the question. We devised several approaches combining different model variants for three main components: Data Augmentation, Passage Retrieval, and Answer Generation. For passage retrieval, we evaluated the monolingual BM25 ranker against the ensemble of re-rankers based on multilingual pretrained language models (PLMs) and also variants of the shared task baseline, re-training it from scratch using a recently introduced contrastive loss that maintains a strong gradient signal throughout training by means of mixed negative samples. For answer generation, we focused on languageand domain-specialization by means of continued language model (LM) pretraining of existing multilingual encoders. Additionally, for both passage retrieval and answer generation, we augmented the training data provided by the task organizers with automatically generated question-answer pairs created from Wikipedia passages to mitigate the issue of data scarcity, particularly for the low-resource languages for which no training data were provided. Our results show that language- and domain-specialization as well as data augmentation help, especially for low-resource languages.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="hung-etal-2022-zusammenqa">
<titleInfo>
<title>ZusammenQA: Data Augmentation with Specialized Models for Cross-lingual Open-retrieval Question Answering System</title>
</titleInfo>
<name type="personal">
<namePart type="given">Chia-Chien</namePart>
<namePart type="family">Hung</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tommaso</namePart>
<namePart type="family">Green</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Robert</namePart>
<namePart type="family">Litschko</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tornike</namePart>
<namePart type="family">Tsereteli</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sotaro</namePart>
<namePart type="family">Takeshita</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Marco</namePart>
<namePart type="family">Bombieri</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Goran</namePart>
<namePart type="family">Glavaš</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Simone</namePart>
<namePart type="given">Paolo</namePart>
<namePart type="family">Ponzetto</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2022-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Workshop on Multilingual Information Access (MIA)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Akari</namePart>
<namePart type="family">Asai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Eunsol</namePart>
<namePart type="family">Choi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jonathan</namePart>
<namePart type="given">H</namePart>
<namePart type="family">Clark</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Junjie</namePart>
<namePart type="family">Hu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chia-Hsuan</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jungo</namePart>
<namePart type="family">Kasai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shayne</namePart>
<namePart type="family">Longpre</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ikuya</namePart>
<namePart type="family">Yamada</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Rui</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Seattle, USA</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>This paper introduces our proposed system for the MIA Shared Task on Cross-lingual Openretrieval Question Answering (COQA). In this challenging scenario, given an input question the system has to gather evidence documents from a multilingual pool and generate from them an answer in the language of the question. We devised several approaches combining different model variants for three main components: Data Augmentation, Passage Retrieval, and Answer Generation. For passage retrieval, we evaluated the monolingual BM25 ranker against the ensemble of re-rankers based on multilingual pretrained language models (PLMs) and also variants of the shared task baseline, re-training it from scratch using a recently introduced contrastive loss that maintains a strong gradient signal throughout training by means of mixed negative samples. For answer generation, we focused on languageand domain-specialization by means of continued language model (LM) pretraining of existing multilingual encoders. Additionally, for both passage retrieval and answer generation, we augmented the training data provided by the task organizers with automatically generated question-answer pairs created from Wikipedia passages to mitigate the issue of data scarcity, particularly for the low-resource languages for which no training data were provided. Our results show that language- and domain-specialization as well as data augmentation help, especially for low-resource languages.</abstract>
<identifier type="citekey">hung-etal-2022-zusammenqa</identifier>
<identifier type="doi">10.18653/v1/2022.mia-1.8</identifier>
<location>
<url>https://aclanthology.org/2022.mia-1.8</url>
</location>
<part>
<date>2022-07</date>
<extent unit="page">
<start>77</start>
<end>90</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T ZusammenQA: Data Augmentation with Specialized Models for Cross-lingual Open-retrieval Question Answering System
%A Hung, Chia-Chien
%A Green, Tommaso
%A Litschko, Robert
%A Tsereteli, Tornike
%A Takeshita, Sotaro
%A Bombieri, Marco
%A Glavaš, Goran
%A Ponzetto, Simone Paolo
%Y Asai, Akari
%Y Choi, Eunsol
%Y Clark, Jonathan H.
%Y Hu, Junjie
%Y Lee, Chia-Hsuan
%Y Kasai, Jungo
%Y Longpre, Shayne
%Y Yamada, Ikuya
%Y Zhang, Rui
%S Proceedings of the Workshop on Multilingual Information Access (MIA)
%D 2022
%8 July
%I Association for Computational Linguistics
%C Seattle, USA
%F hung-etal-2022-zusammenqa
%X This paper introduces our proposed system for the MIA Shared Task on Cross-lingual Openretrieval Question Answering (COQA). In this challenging scenario, given an input question the system has to gather evidence documents from a multilingual pool and generate from them an answer in the language of the question. We devised several approaches combining different model variants for three main components: Data Augmentation, Passage Retrieval, and Answer Generation. For passage retrieval, we evaluated the monolingual BM25 ranker against the ensemble of re-rankers based on multilingual pretrained language models (PLMs) and also variants of the shared task baseline, re-training it from scratch using a recently introduced contrastive loss that maintains a strong gradient signal throughout training by means of mixed negative samples. For answer generation, we focused on languageand domain-specialization by means of continued language model (LM) pretraining of existing multilingual encoders. Additionally, for both passage retrieval and answer generation, we augmented the training data provided by the task organizers with automatically generated question-answer pairs created from Wikipedia passages to mitigate the issue of data scarcity, particularly for the low-resource languages for which no training data were provided. Our results show that language- and domain-specialization as well as data augmentation help, especially for low-resource languages.
%R 10.18653/v1/2022.mia-1.8
%U https://aclanthology.org/2022.mia-1.8
%U https://doi.org/10.18653/v1/2022.mia-1.8
%P 77-90
Markdown (Informal)
[ZusammenQA: Data Augmentation with Specialized Models for Cross-lingual Open-retrieval Question Answering System](https://aclanthology.org/2022.mia-1.8) (Hung et al., MIA 2022)
ACL
- Chia-Chien Hung, Tommaso Green, Robert Litschko, Tornike Tsereteli, Sotaro Takeshita, Marco Bombieri, Goran Glavaš, and Simone Paolo Ponzetto. 2022. ZusammenQA: Data Augmentation with Specialized Models for Cross-lingual Open-retrieval Question Answering System. In Proceedings of the Workshop on Multilingual Information Access (MIA), pages 77–90, Seattle, USA. Association for Computational Linguistics.