@inproceedings{fekete-bjerva-2023-gradual,
title = "Gradual Language Model Adaptation Using Fine-Grained Typology",
author = "Fekete, Marcell Richard and
Bjerva, Johannes",
editor = "Beinborn, Lisa and
Goswami, Koustava and
Murado{\u{g}}lu, Saliha and
Sorokin, Alexey and
Kumar, Ritesh and
Shcherbakov, Andreas and
Ponti, Edoardo M. and
Cotterell, Ryan and
Vylomova, Ekaterina",
booktitle = "Proceedings of the 5th Workshop on Research in Computational Linguistic Typology and Multilingual NLP",
month = may,
year = "2023",
address = "Dubrovnik, Croatia",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.sigtyp-1.19",
doi = "10.18653/v1/2023.sigtyp-1.19",
pages = "153--158",
abstract = {Transformer-based language models (LMs) offer superior performance in a wide range of NLP tasks compared to previous paradigms. However, the vast majority of the world{'}s languages do not have adequate training data available for monolingual LMs (Joshi et al., 2020). While the use of multilingual LMs might address this data imbalance, there is evidence that multilingual LMs struggle when it comes to model adaptation to to resource-poor languages (Wu and Dredze, 2020), or to languages which have typological characteristics unseen by the LM ({\"U}st{\"u}n et al., 2022). Other approaches aim to adapt monolingual LMs to resource-poor languages that are related to the model language. However, there are conflicting findings regarding whether language relatedness correlates with successful adaptation (de Vries et al., 2021), or not ({\'A}cs et al., 2021). With gradual LM adaptation, our approach presented in this extended abstract, we add to the research direction of monolingual LM adaptation. Instead of direct adaptation to a target language, we propose adaptation in stages, first adapting to one or more intermediate languages before the final adaptation step. Inspired by principles of curriculum learning (Bengio et al., 2009), we search for an ideal ordering of languages that can result in improved LM performance on the target language. We follow evidence that typological similarity might correlate with the success of cross-lingual transfer (Pires et al., 2019; {\"U}st{\"u}n et al., 2022; de Vries et al., 2021) as we believe the success of this transfer is essential for successful model adaptation. Thus we order languages based on their relative typological similarity between them. In our approach, we quantify typological similarity using structural vectors as derived from counts of dependency links (Bjerva et al., 2019), as such fine-grained measures can give a more accurate picture of the typological characteristics of languages (Ponti et al., 2019). We believe that gradual LM adaptation may lead to improved LM performance on a range of resource-poor languages and typologically diverse languages. Additionally, it enables future research to evaluate the correlation between the success of cross-lingual transfer and various typological similarity measures.},
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="fekete-bjerva-2023-gradual">
<titleInfo>
<title>Gradual Language Model Adaptation Using Fine-Grained Typology</title>
</titleInfo>
<name type="personal">
<namePart type="given">Marcell</namePart>
<namePart type="given">Richard</namePart>
<namePart type="family">Fekete</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Johannes</namePart>
<namePart type="family">Bjerva</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2023-05</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 5th Workshop on Research in Computational Linguistic Typology and Multilingual NLP</title>
</titleInfo>
<name type="personal">
<namePart type="given">Lisa</namePart>
<namePart type="family">Beinborn</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Koustava</namePart>
<namePart type="family">Goswami</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Saliha</namePart>
<namePart type="family">Muradoğlu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alexey</namePart>
<namePart type="family">Sorokin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ritesh</namePart>
<namePart type="family">Kumar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Andreas</namePart>
<namePart type="family">Shcherbakov</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Edoardo</namePart>
<namePart type="given">M</namePart>
<namePart type="family">Ponti</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ryan</namePart>
<namePart type="family">Cotterell</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ekaterina</namePart>
<namePart type="family">Vylomova</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Dubrovnik, Croatia</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Transformer-based language models (LMs) offer superior performance in a wide range of NLP tasks compared to previous paradigms. However, the vast majority of the world’s languages do not have adequate training data available for monolingual LMs (Joshi et al., 2020). While the use of multilingual LMs might address this data imbalance, there is evidence that multilingual LMs struggle when it comes to model adaptation to to resource-poor languages (Wu and Dredze, 2020), or to languages which have typological characteristics unseen by the LM (Üstün et al., 2022). Other approaches aim to adapt monolingual LMs to resource-poor languages that are related to the model language. However, there are conflicting findings regarding whether language relatedness correlates with successful adaptation (de Vries et al., 2021), or not (Ács et al., 2021). With gradual LM adaptation, our approach presented in this extended abstract, we add to the research direction of monolingual LM adaptation. Instead of direct adaptation to a target language, we propose adaptation in stages, first adapting to one or more intermediate languages before the final adaptation step. Inspired by principles of curriculum learning (Bengio et al., 2009), we search for an ideal ordering of languages that can result in improved LM performance on the target language. We follow evidence that typological similarity might correlate with the success of cross-lingual transfer (Pires et al., 2019; Üstün et al., 2022; de Vries et al., 2021) as we believe the success of this transfer is essential for successful model adaptation. Thus we order languages based on their relative typological similarity between them. In our approach, we quantify typological similarity using structural vectors as derived from counts of dependency links (Bjerva et al., 2019), as such fine-grained measures can give a more accurate picture of the typological characteristics of languages (Ponti et al., 2019). We believe that gradual LM adaptation may lead to improved LM performance on a range of resource-poor languages and typologically diverse languages. Additionally, it enables future research to evaluate the correlation between the success of cross-lingual transfer and various typological similarity measures.</abstract>
<identifier type="citekey">fekete-bjerva-2023-gradual</identifier>
<identifier type="doi">10.18653/v1/2023.sigtyp-1.19</identifier>
<location>
<url>https://aclanthology.org/2023.sigtyp-1.19</url>
</location>
<part>
<date>2023-05</date>
<extent unit="page">
<start>153</start>
<end>158</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Gradual Language Model Adaptation Using Fine-Grained Typology
%A Fekete, Marcell Richard
%A Bjerva, Johannes
%Y Beinborn, Lisa
%Y Goswami, Koustava
%Y Muradoğlu, Saliha
%Y Sorokin, Alexey
%Y Kumar, Ritesh
%Y Shcherbakov, Andreas
%Y Ponti, Edoardo M.
%Y Cotterell, Ryan
%Y Vylomova, Ekaterina
%S Proceedings of the 5th Workshop on Research in Computational Linguistic Typology and Multilingual NLP
%D 2023
%8 May
%I Association for Computational Linguistics
%C Dubrovnik, Croatia
%F fekete-bjerva-2023-gradual
%X Transformer-based language models (LMs) offer superior performance in a wide range of NLP tasks compared to previous paradigms. However, the vast majority of the world’s languages do not have adequate training data available for monolingual LMs (Joshi et al., 2020). While the use of multilingual LMs might address this data imbalance, there is evidence that multilingual LMs struggle when it comes to model adaptation to to resource-poor languages (Wu and Dredze, 2020), or to languages which have typological characteristics unseen by the LM (Üstün et al., 2022). Other approaches aim to adapt monolingual LMs to resource-poor languages that are related to the model language. However, there are conflicting findings regarding whether language relatedness correlates with successful adaptation (de Vries et al., 2021), or not (Ács et al., 2021). With gradual LM adaptation, our approach presented in this extended abstract, we add to the research direction of monolingual LM adaptation. Instead of direct adaptation to a target language, we propose adaptation in stages, first adapting to one or more intermediate languages before the final adaptation step. Inspired by principles of curriculum learning (Bengio et al., 2009), we search for an ideal ordering of languages that can result in improved LM performance on the target language. We follow evidence that typological similarity might correlate with the success of cross-lingual transfer (Pires et al., 2019; Üstün et al., 2022; de Vries et al., 2021) as we believe the success of this transfer is essential for successful model adaptation. Thus we order languages based on their relative typological similarity between them. In our approach, we quantify typological similarity using structural vectors as derived from counts of dependency links (Bjerva et al., 2019), as such fine-grained measures can give a more accurate picture of the typological characteristics of languages (Ponti et al., 2019). We believe that gradual LM adaptation may lead to improved LM performance on a range of resource-poor languages and typologically diverse languages. Additionally, it enables future research to evaluate the correlation between the success of cross-lingual transfer and various typological similarity measures.
%R 10.18653/v1/2023.sigtyp-1.19
%U https://aclanthology.org/2023.sigtyp-1.19
%U https://doi.org/10.18653/v1/2023.sigtyp-1.19
%P 153-158
Markdown (Informal)
[Gradual Language Model Adaptation Using Fine-Grained Typology](https://aclanthology.org/2023.sigtyp-1.19) (Fekete & Bjerva, SIGTYP 2023)
ACL