@inproceedings{bafna-zabokrtsky-2022-subword,
title = "Subword-based Cross-lingual Transfer of Embeddings from {H}indi to {M}arathi and {N}epali",
author = "Bafna, Niyati and
{\v{Z}}abokrtsk{\'y}, Zden{\v{e}}k",
editor = "Nicolai, Garrett and
Chodroff, Eleanor",
booktitle = "Proceedings of the 19th SIGMORPHON Workshop on Computational Research in Phonetics, Phonology, and Morphology",
month = jul,
year = "2022",
address = "Seattle, Washington",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2022.sigmorphon-1.7",
doi = "10.18653/v1/2022.sigmorphon-1.7",
pages = "61--71",
abstract = "Word embeddings are growing to be a crucial resource in the field of NLP for any language. This work introduces a novel technique for static subword embeddings transfer for Indic languages from a relatively higher resource language to a genealogically related low resource language. We primarily work with HindiMarathi, simulating a low-resource scenario for Marathi, and confirm observed trends on Nepali. We demonstrate the consistent benefits of unsupervised morphemic segmentation on both source and target sides over the treatment performed by fastText. Our best-performing approach uses an EM-style approach to learning bilingual subword embeddings; we also show, for the first time, that a trivial {``}copyand-paste{''} embeddings transfer based on even perfect bilingual lexicons is inadequate in capturing language-specific relationships. We find that our approach substantially outperforms the fastText baselines for both Marathi and Nepali on the Word Similarity task as well as WordNetBased Synonymy Tests; on the former task, its performance for Marathi is close to that of pretrained fastText embeddings that use three orders of magnitude more Marathi data.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="bafna-zabokrtsky-2022-subword">
<titleInfo>
<title>Subword-based Cross-lingual Transfer of Embeddings from Hindi to Marathi and Nepali</title>
</titleInfo>
<name type="personal">
<namePart type="given">Niyati</namePart>
<namePart type="family">Bafna</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zdeněk</namePart>
<namePart type="family">Žabokrtský</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 19th SIGMORPHON Workshop on Computational Research in Phonetics, Phonology, and Morphology</title>
</titleInfo>
<name type="personal">
<namePart type="given">Garrett</namePart>
<namePart type="family">Nicolai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Eleanor</namePart>
<namePart type="family">Chodroff</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Seattle, Washington</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Word embeddings are growing to be a crucial resource in the field of NLP for any language. This work introduces a novel technique for static subword embeddings transfer for Indic languages from a relatively higher resource language to a genealogically related low resource language. We primarily work with HindiMarathi, simulating a low-resource scenario for Marathi, and confirm observed trends on Nepali. We demonstrate the consistent benefits of unsupervised morphemic segmentation on both source and target sides over the treatment performed by fastText. Our best-performing approach uses an EM-style approach to learning bilingual subword embeddings; we also show, for the first time, that a trivial “copyand-paste” embeddings transfer based on even perfect bilingual lexicons is inadequate in capturing language-specific relationships. We find that our approach substantially outperforms the fastText baselines for both Marathi and Nepali on the Word Similarity task as well as WordNetBased Synonymy Tests; on the former task, its performance for Marathi is close to that of pretrained fastText embeddings that use three orders of magnitude more Marathi data.</abstract>
<identifier type="citekey">bafna-zabokrtsky-2022-subword</identifier>
<identifier type="doi">10.18653/v1/2022.sigmorphon-1.7</identifier>
<location>
<url>https://aclanthology.org/2022.sigmorphon-1.7</url>
</location>
<part>
<date>2022-07</date>
<extent unit="page">
<start>61</start>
<end>71</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Subword-based Cross-lingual Transfer of Embeddings from Hindi to Marathi and Nepali
%A Bafna, Niyati
%A Žabokrtský, Zdeněk
%Y Nicolai, Garrett
%Y Chodroff, Eleanor
%S Proceedings of the 19th SIGMORPHON Workshop on Computational Research in Phonetics, Phonology, and Morphology
%D 2022
%8 July
%I Association for Computational Linguistics
%C Seattle, Washington
%F bafna-zabokrtsky-2022-subword
%X Word embeddings are growing to be a crucial resource in the field of NLP for any language. This work introduces a novel technique for static subword embeddings transfer for Indic languages from a relatively higher resource language to a genealogically related low resource language. We primarily work with HindiMarathi, simulating a low-resource scenario for Marathi, and confirm observed trends on Nepali. We demonstrate the consistent benefits of unsupervised morphemic segmentation on both source and target sides over the treatment performed by fastText. Our best-performing approach uses an EM-style approach to learning bilingual subword embeddings; we also show, for the first time, that a trivial “copyand-paste” embeddings transfer based on even perfect bilingual lexicons is inadequate in capturing language-specific relationships. We find that our approach substantially outperforms the fastText baselines for both Marathi and Nepali on the Word Similarity task as well as WordNetBased Synonymy Tests; on the former task, its performance for Marathi is close to that of pretrained fastText embeddings that use three orders of magnitude more Marathi data.
%R 10.18653/v1/2022.sigmorphon-1.7
%U https://aclanthology.org/2022.sigmorphon-1.7
%U https://doi.org/10.18653/v1/2022.sigmorphon-1.7
%P 61-71
Markdown (Informal)
[Subword-based Cross-lingual Transfer of Embeddings from Hindi to Marathi and Nepali](https://aclanthology.org/2022.sigmorphon-1.7) (Bafna & Žabokrtský, SIGMORPHON 2022)
ACL