<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="tran-kruschwitz-2021-ur">
    <titleInfo>
        <title>ur-iw-hnt at GermEval 2021: An Ensembling Strategy with Multiple BERT Models</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Hoai</namePart>
        <namePart type="given">Nam</namePart>
        <namePart type="family">Tran</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Udo</namePart>
        <namePart type="family">Kruschwitz</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2021-09</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the GermEval 2021 Shared Task on the Identification of Toxic, Engaging, and Fact-Claiming Comments</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Julian</namePart>
            <namePart type="family">Risch</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Anke</namePart>
            <namePart type="family">Stoll</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Lena</namePart>
            <namePart type="family">Wilms</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Michael</namePart>
            <namePart type="family">Wiegand</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Duesseldorf, Germany</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
    </relatedItem>
    <abstract>This paper describes our approach (ur-iw-hnt) for the Shared Task of GermEval2021 to identify toxic, engaging, and fact-claiming comments. We submitted three runs using an ensembling strategy by majority (hard) voting with multiple different BERT models of three different types: German-based, Twitter-based, and multilingual models. All ensemble models outperform single models, while BERTweet is the winner of all individual models in every subtask. Twitter-based models perform better than GermanBERT models, and multilingual models perform worse but by a small margin.</abstract>
    <identifier type="citekey">tran-kruschwitz-2021-ur</identifier>
    <location>
        <url>https://aclanthology.org/2021.germeval-1.12</url>
    </location>
    <part>
        <date>2021-09</date>
        <extent unit="page">
            <start>83</start>
            <end>87</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T ur-iw-hnt at GermEval 2021: An Ensembling Strategy with Multiple BERT Models
%A Tran, Hoai Nam
%A Kruschwitz, Udo
%Y Risch, Julian
%Y Stoll, Anke
%Y Wilms, Lena
%Y Wiegand, Michael
%S Proceedings of the GermEval 2021 Shared Task on the Identification of Toxic, Engaging, and Fact-Claiming Comments
%D 2021
%8 September
%I Association for Computational Linguistics
%C Duesseldorf, Germany
%F tran-kruschwitz-2021-ur
%X This paper describes our approach (ur-iw-hnt) for the Shared Task of GermEval2021 to identify toxic, engaging, and fact-claiming comments. We submitted three runs using an ensembling strategy by majority (hard) voting with multiple different BERT models of three different types: German-based, Twitter-based, and multilingual models. All ensemble models outperform single models, while BERTweet is the winner of all individual models in every subtask. Twitter-based models perform better than GermanBERT models, and multilingual models perform worse but by a small margin.
%U https://aclanthology.org/2021.germeval-1.12
%P 83-87

Markdown (Informal)

[ur-iw-hnt at GermEval 2021: An Ensembling Strategy with Multiple BERT Models](https://aclanthology.org/2021.germeval-1.12) (Tran & Kruschwitz, GermEval 2021)

• ur-iw-hnt at GermEval 2021: An Ensembling Strategy with Multiple BERT Models (Tran & Kruschwitz, GermEval 2021)

ACL

• Hoai Nam Tran and Udo Kruschwitz. 2021. ur-iw-hnt at GermEval 2021: An Ensembling Strategy with Multiple BERT Models. In Proceedings of the GermEval 2021 Shared Task on the Identification of Toxic, Engaging, and Fact-Claiming Comments, pages 83–87, Duesseldorf, Germany. Association for Computational Linguistics.