<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="van-der-werff-etal-2022-automatic">
    <titleInfo>
        <title>Automatic Discrimination of Human and Neural Machine Translation: A Study with Multiple Pre-Trained Models and Longer Context</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Tobias</namePart>
        <namePart type="family">van der Werff</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Rik</namePart>
        <namePart type="family">van Noord</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Antonio</namePart>
        <namePart type="family">Toral</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 23rd Annual Conference of the European Association for Machine Translation</title>
        </titleInfo>
        <originInfo>
            <publisher>European Association for Machine Translation</publisher>
            <place>
                <placeTerm type="text">Ghent, Belgium</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
    </relatedItem>
    <abstract>We address the task of automatically distinguishing between human-translated (HT) and machine translated (MT) texts. Following recent work, we fine-tune pre-trained language models (LMs) to perform this task. Our work differs in that we use state-of-the-art pre-trained LMs, as well as the test sets of the WMT news shared tasks as training data, to ensure the sentences were not seen during training of the MT system itself. Moreover, we analyse performance for a number of different experimental setups, such as adding translationese data, going beyond the sentence-level and normalizing punctuation. We show that (i) choosing a state-of-the-art LM can make quite a difference: our best baseline system (DeBERTa) outperforms both BERT and RoBERTa by over 3% accuracy, (ii) adding translationese data is only beneficial if there is not much data available, (iii) considerable improvements can be obtained by classifying at the document-level and (iv) normalizing punctuation and thus avoiding (some) shortcuts has no impact on model performance.</abstract>
    <identifier type="citekey">van-der-werff-etal-2022-automatic</identifier>
    <location>
        <url>https://aclanthology.org/2022.eamt-1.19</url>
    </location>
    <part>
        <date>2022-06</date>
        <extent unit="page">
            <start>161</start>
            <end>170</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Automatic Discrimination of Human and Neural Machine Translation: A Study with Multiple Pre-Trained Models and Longer Context
%A van der Werff, Tobias
%A van Noord, Rik
%A Toral, Antonio
%S Proceedings of the 23rd Annual Conference of the European Association for Machine Translation
%D 2022
%8 June
%I European Association for Machine Translation
%C Ghent, Belgium
%F van-der-werff-etal-2022-automatic
%X We address the task of automatically distinguishing between human-translated (HT) and machine translated (MT) texts. Following recent work, we fine-tune pre-trained language models (LMs) to perform this task. Our work differs in that we use state-of-the-art pre-trained LMs, as well as the test sets of the WMT news shared tasks as training data, to ensure the sentences were not seen during training of the MT system itself. Moreover, we analyse performance for a number of different experimental setups, such as adding translationese data, going beyond the sentence-level and normalizing punctuation. We show that (i) choosing a state-of-the-art LM can make quite a difference: our best baseline system (DeBERTa) outperforms both BERT and RoBERTa by over 3% accuracy, (ii) adding translationese data is only beneficial if there is not much data available, (iii) considerable improvements can be obtained by classifying at the document-level and (iv) normalizing punctuation and thus avoiding (some) shortcuts has no impact on model performance.
%U https://aclanthology.org/2022.eamt-1.19
%P 161-170

Markdown (Informal)

[Automatic Discrimination of Human and Neural Machine Translation: A Study with Multiple Pre-Trained Models and Longer Context](https://aclanthology.org/2022.eamt-1.19) (van der Werff et al., EAMT 2022)

• Automatic Discrimination of Human and Neural Machine Translation: A Study with Multiple Pre-Trained Models and Longer Context (van der Werff et al., EAMT 2022)

ACL

• Tobias van der Werff, Rik van Noord, and Antonio Toral. 2022. Automatic Discrimination of Human and Neural Machine Translation: A Study with Multiple Pre-Trained Models and Longer Context. In Proceedings of the 23rd Annual Conference of the European Association for Machine Translation, pages 161–170, Ghent, Belgium. European Association for Machine Translation.