<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="bell-etal-2025-translate">
    <titleInfo>
        <title>Translate, Then Detect: Leveraging Machine Translation for Cross-Lingual Toxicity Classification</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Samuel</namePart>
        <namePart type="family">Bell</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Eduardo</namePart>
        <namePart type="family">Sánchez</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">David</namePart>
        <namePart type="family">Dale</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Pontus</namePart>
        <namePart type="family">Stenetorp</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Mikel</namePart>
        <namePart type="family">Artetxe</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Marta</namePart>
        <namePart type="given">R</namePart>
        <namePart type="family">Costa-Jussà</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2025-11</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the Tenth Conference on Machine Translation</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Barry</namePart>
            <namePart type="family">Haddow</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Tom</namePart>
            <namePart type="family">Kocmi</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Philipp</namePart>
            <namePart type="family">Koehn</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Christof</namePart>
            <namePart type="family">Monz</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Suzhou, China</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-341-8</identifier>
    </relatedItem>
    <abstract>Multilingual toxicity detection remains a significant challenge due to the scarcity of training data and resources for many languages. While prior work has leveraged the translate-test paradigm to support cross-lingual transfer across a range of classification tasks, the utility of translation in supporting toxicity detection at scale remains unclear.In this work, we conduct a comprehensive comparison of translation-based and language-specific/multilingual classification pipelines.We find that translation-based pipelines consistently outperform out-of-distribution classifiers in 81.3% of cases (13 of 16 languages), with translation benefits strongly correlated with both the resource level of the target language and the quality of the machine translation (MT) system.Our analysis reveals that traditional classifiers continue to outperform LLM-based judgment methods, with this advantage being particularly pronounced for low-resource languages, where translate-classify methods dominate translate-judge approaches in 6 out of 7 cases.We show that MT-specific fine-tuning on LLMs yields lower refusal rates compared to standard instruction-tuned models, but it can negatively impact toxicity detection accuracy for low-resource languages.These findings offer actionable guidance for practitioners developing scalable multilingual content moderation systems.</abstract>
    <identifier type="citekey">bell-etal-2025-translate</identifier>
    <location>
        <url>https://aclanthology.org/2025.wmt-1.15/</url>
    </location>
    <part>
        <date>2025-11</date>
        <extent unit="page">
            <start>253</start>
            <end>268</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Translate, Then Detect: Leveraging Machine Translation for Cross-Lingual Toxicity Classification
%A Bell, Samuel
%A Sánchez, Eduardo
%A Dale, David
%A Stenetorp, Pontus
%A Artetxe, Mikel
%A Costa-Jussà, Marta R.
%Y Haddow, Barry
%Y Kocmi, Tom
%Y Koehn, Philipp
%Y Monz, Christof
%S Proceedings of the Tenth Conference on Machine Translation
%D 2025
%8 November
%I Association for Computational Linguistics
%C Suzhou, China
%@ 979-8-89176-341-8
%F bell-etal-2025-translate
%X Multilingual toxicity detection remains a significant challenge due to the scarcity of training data and resources for many languages. While prior work has leveraged the translate-test paradigm to support cross-lingual transfer across a range of classification tasks, the utility of translation in supporting toxicity detection at scale remains unclear.In this work, we conduct a comprehensive comparison of translation-based and language-specific/multilingual classification pipelines.We find that translation-based pipelines consistently outperform out-of-distribution classifiers in 81.3% of cases (13 of 16 languages), with translation benefits strongly correlated with both the resource level of the target language and the quality of the machine translation (MT) system.Our analysis reveals that traditional classifiers continue to outperform LLM-based judgment methods, with this advantage being particularly pronounced for low-resource languages, where translate-classify methods dominate translate-judge approaches in 6 out of 7 cases.We show that MT-specific fine-tuning on LLMs yields lower refusal rates compared to standard instruction-tuned models, but it can negatively impact toxicity detection accuracy for low-resource languages.These findings offer actionable guidance for practitioners developing scalable multilingual content moderation systems.
%U https://aclanthology.org/2025.wmt-1.15/
%P 253-268

Markdown (Informal)

[Translate, Then Detect: Leveraging Machine Translation for Cross-Lingual Toxicity Classification](https://aclanthology.org/2025.wmt-1.15/) (Bell et al., WMT 2025)

• Translate, Then Detect: Leveraging Machine Translation for Cross-Lingual Toxicity Classification (Bell et al., WMT 2025)

ACL

• Samuel Bell, Eduardo Sánchez, David Dale, Pontus Stenetorp, Mikel Artetxe, and Marta R. Costa-Jussà. 2025. Translate, Then Detect: Leveraging Machine Translation for Cross-Lingual Toxicity Classification. In Proceedings of the Tenth Conference on Machine Translation, pages 253–268, Suzhou, China. Association for Computational Linguistics.