@inproceedings{stahlberg-kumar-2024-synthetic,
title = "Synthetic Data Generation for Low-resource Grammatical Error Correction with Tagged Corruption Models",
author = "Stahlberg, Felix and
Kumar, Shankar",
editor = {Kochmar, Ekaterina and
Bexte, Marie and
Burstein, Jill and
Horbach, Andrea and
Laarmann-Quante, Ronja and
Tack, Ana{\"\i}s and
Yaneva, Victoria and
Yuan, Zheng},
booktitle = "Proceedings of the 19th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2024)",
month = jun,
year = "2024",
address = "Mexico City, Mexico",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.bea-1.2",
pages = "11--16",
abstract = "Tagged corruption models provide precise control over the introduction of grammatical errors into clean text. This capability has made them a powerful tool for generating pre-training data for grammatical error correction (GEC) in English. In this work, we demonstrate their application to four languages with substantially fewer GEC resources than English: German, Romanian, Russian, and Spanish. We release a new tagged-corruption dataset consisting of 2.5M examples per language that was generated by a fine-tuned PaLM 2 foundation model. Pre-training on tagged corruptions yields consistent gains across all four languages, especially for small model sizes and languages with limited human-labelled data.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="stahlberg-kumar-2024-synthetic">
<titleInfo>
<title>Synthetic Data Generation for Low-resource Grammatical Error Correction with Tagged Corruption Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Felix</namePart>
<namePart type="family">Stahlberg</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shankar</namePart>
<namePart type="family">Kumar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2024-06</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 19th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2024)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ekaterina</namePart>
<namePart type="family">Kochmar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Marie</namePart>
<namePart type="family">Bexte</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jill</namePart>
<namePart type="family">Burstein</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Andrea</namePart>
<namePart type="family">Horbach</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ronja</namePart>
<namePart type="family">Laarmann-Quante</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Anaïs</namePart>
<namePart type="family">Tack</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Victoria</namePart>
<namePart type="family">Yaneva</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zheng</namePart>
<namePart type="family">Yuan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Mexico City, Mexico</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Tagged corruption models provide precise control over the introduction of grammatical errors into clean text. This capability has made them a powerful tool for generating pre-training data for grammatical error correction (GEC) in English. In this work, we demonstrate their application to four languages with substantially fewer GEC resources than English: German, Romanian, Russian, and Spanish. We release a new tagged-corruption dataset consisting of 2.5M examples per language that was generated by a fine-tuned PaLM 2 foundation model. Pre-training on tagged corruptions yields consistent gains across all four languages, especially for small model sizes and languages with limited human-labelled data.</abstract>
<identifier type="citekey">stahlberg-kumar-2024-synthetic</identifier>
<location>
<url>https://aclanthology.org/2024.bea-1.2</url>
</location>
<part>
<date>2024-06</date>
<extent unit="page">
<start>11</start>
<end>16</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Synthetic Data Generation for Low-resource Grammatical Error Correction with Tagged Corruption Models
%A Stahlberg, Felix
%A Kumar, Shankar
%Y Kochmar, Ekaterina
%Y Bexte, Marie
%Y Burstein, Jill
%Y Horbach, Andrea
%Y Laarmann-Quante, Ronja
%Y Tack, Anaïs
%Y Yaneva, Victoria
%Y Yuan, Zheng
%S Proceedings of the 19th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2024)
%D 2024
%8 June
%I Association for Computational Linguistics
%C Mexico City, Mexico
%F stahlberg-kumar-2024-synthetic
%X Tagged corruption models provide precise control over the introduction of grammatical errors into clean text. This capability has made them a powerful tool for generating pre-training data for grammatical error correction (GEC) in English. In this work, we demonstrate their application to four languages with substantially fewer GEC resources than English: German, Romanian, Russian, and Spanish. We release a new tagged-corruption dataset consisting of 2.5M examples per language that was generated by a fine-tuned PaLM 2 foundation model. Pre-training on tagged corruptions yields consistent gains across all four languages, especially for small model sizes and languages with limited human-labelled data.
%U https://aclanthology.org/2024.bea-1.2
%P 11-16
Markdown (Informal)
[Synthetic Data Generation for Low-resource Grammatical Error Correction with Tagged Corruption Models](https://aclanthology.org/2024.bea-1.2) (Stahlberg & Kumar, BEA 2024)
ACL