<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="toraman-etal-2022-large">
    <titleInfo>
        <title>Large-Scale Hate Speech Detection with Cross-Domain Transfer</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Cagri</namePart>
        <namePart type="family">Toraman</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Furkan</namePart>
        <namePart type="family">Şahinuç</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Eyup</namePart>
        <namePart type="family">Yilmaz</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2022-06</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the Thirteenth Language Resources and Evaluation Conference</title>
        </titleInfo>
        <originInfo>
            <publisher>European Language Resources Association</publisher>
            <place>
                <placeTerm type="text">Marseille, France</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
    </relatedItem>
    <abstract>The performance of hate speech detection models relies on the datasets on which the models are trained. Existing datasets are mostly prepared with a limited number of instances or hate domains that define hate topics. This hinders large-scale analysis and transfer learning with respect to hate domains. In this study, we construct large-scale tweet datasets for hate speech detection in English and a low-resource language, Turkish, consisting of human-labeled 100k tweets per each. Our datasets are designed to have equal number of tweets distributed over five domains. The experimental results supported by statistical tests show that Transformer-based language models outperform conventional bag-of-words and neural models by at least 5% in English and 10% in Turkish for large-scale hate speech detection. The performance is also scalable to different training sizes, such that 98% of performance in English, and 97% in Turkish, are recovered when 20% of training instances are used. We further examine the generalization ability of cross-domain transfer among hate domains. We show that 96% of the performance of a target domain in average is recovered by other domains for English, and 92% for Turkish. Gender and religion are more successful to generalize to other domains, while sports fail most.</abstract>
    <identifier type="citekey">toraman-etal-2022-large</identifier>
    <location>
        <url>https://aclanthology.org/2022.lrec-1.238</url>
    </location>
    <part>
        <date>2022-06</date>
        <extent unit="page">
            <start>2215</start>
            <end>2225</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Large-Scale Hate Speech Detection with Cross-Domain Transfer
%A Toraman, Cagri
%A Şahinuç, Furkan
%A Yilmaz, Eyup
%S Proceedings of the Thirteenth Language Resources and Evaluation Conference
%D 2022
%8 June
%I European Language Resources Association
%C Marseille, France
%F toraman-etal-2022-large
%X The performance of hate speech detection models relies on the datasets on which the models are trained. Existing datasets are mostly prepared with a limited number of instances or hate domains that define hate topics. This hinders large-scale analysis and transfer learning with respect to hate domains. In this study, we construct large-scale tweet datasets for hate speech detection in English and a low-resource language, Turkish, consisting of human-labeled 100k tweets per each. Our datasets are designed to have equal number of tweets distributed over five domains. The experimental results supported by statistical tests show that Transformer-based language models outperform conventional bag-of-words and neural models by at least 5% in English and 10% in Turkish for large-scale hate speech detection. The performance is also scalable to different training sizes, such that 98% of performance in English, and 97% in Turkish, are recovered when 20% of training instances are used. We further examine the generalization ability of cross-domain transfer among hate domains. We show that 96% of the performance of a target domain in average is recovered by other domains for English, and 92% for Turkish. Gender and religion are more successful to generalize to other domains, while sports fail most.
%U https://aclanthology.org/2022.lrec-1.238
%P 2215-2225

Markdown (Informal)

[Large-Scale Hate Speech Detection with Cross-Domain Transfer](https://aclanthology.org/2022.lrec-1.238) (Toraman et al., LREC 2022)

• Large-Scale Hate Speech Detection with Cross-Domain Transfer (Toraman et al., LREC 2022)

ACL

• Cagri Toraman, Furkan Şahinuç, and Eyup Yilmaz. 2022. Large-Scale Hate Speech Detection with Cross-Domain Transfer. In Proceedings of the Thirteenth Language Resources and Evaluation Conference, pages 2215–2225, Marseille, France. European Language Resources Association.