<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="lu-etal-2023-toward">
    <titleInfo>
        <title>Toward Human-Like Evaluation for Natural Language Generation with Error Analysis</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Qingyu</namePart>
        <namePart type="family">Lu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Liang</namePart>
        <namePart type="family">Ding</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Liping</namePart>
        <namePart type="family">Xie</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Kanjian</namePart>
        <namePart type="family">Zhang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Derek</namePart>
        <namePart type="given">F</namePart>
        <namePart type="family">Wong</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Dacheng</namePart>
        <namePart type="family">Tao</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2023-07</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Anna</namePart>
            <namePart type="family">Rogers</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Jordan</namePart>
            <namePart type="family">Boyd-Graber</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Naoaki</namePart>
            <namePart type="family">Okazaki</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Toronto, Canada</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
    </relatedItem>
    <abstract>The pretrained language model (PLM) based metrics have been successfully used in evaluating language generation tasks. Recent studies of the human evaluation community show that considering both major errors (e.g. mistranslated tokens) and minor errors (e.g. imperfections in fluency) can produce high-quality judgments. This inspires us to approach the final goal of the automatic metrics (human-like evaluations) by fine-grained error analysis. In this paper, we argue that the ability to estimate sentence confidence is the tip of the iceberg for PLM-based metrics. And it can be used to refine the generated sentence toward higher confidence and more reference-grounded, where the costs of refining and approaching reference are used to determine the major and minor errors, respectively. To this end, we take BARTScore as the testbed and present an innovative solution to marry the unexploited sentence refining capacity of BARTScore and human-like error analysis, where the final score consists of both the evaluations of major and minor errors. Experiments show that our solution consistently and significantly improves BARTScore, and outperforms top-scoring metrics in 19/25 test settings. Analyses demonstrate our method robustly and efficiently approaches human-like evaluations, enjoying better interpretability. Our code and scripts will be publicly released in https://github.com/Coldmist-Lu/ErrorAnalysis_NLGEvaluation.</abstract>
    <identifier type="citekey">lu-etal-2023-toward</identifier>
    <identifier type="doi">10.18653/v1/2023.acl-long.324</identifier>
    <location>
        <url>https://aclanthology.org/2023.acl-long.324</url>
    </location>
    <part>
        <date>2023-07</date>
        <extent unit="page">
            <start>5892</start>
            <end>5907</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Toward Human-Like Evaluation for Natural Language Generation with Error Analysis
%A Lu, Qingyu
%A Ding, Liang
%A Xie, Liping
%A Zhang, Kanjian
%A Wong, Derek F.
%A Tao, Dacheng
%Y Rogers, Anna
%Y Boyd-Graber, Jordan
%Y Okazaki, Naoaki
%S Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2023
%8 July
%I Association for Computational Linguistics
%C Toronto, Canada
%F lu-etal-2023-toward
%X The pretrained language model (PLM) based metrics have been successfully used in evaluating language generation tasks. Recent studies of the human evaluation community show that considering both major errors (e.g. mistranslated tokens) and minor errors (e.g. imperfections in fluency) can produce high-quality judgments. This inspires us to approach the final goal of the automatic metrics (human-like evaluations) by fine-grained error analysis. In this paper, we argue that the ability to estimate sentence confidence is the tip of the iceberg for PLM-based metrics. And it can be used to refine the generated sentence toward higher confidence and more reference-grounded, where the costs of refining and approaching reference are used to determine the major and minor errors, respectively. To this end, we take BARTScore as the testbed and present an innovative solution to marry the unexploited sentence refining capacity of BARTScore and human-like error analysis, where the final score consists of both the evaluations of major and minor errors. Experiments show that our solution consistently and significantly improves BARTScore, and outperforms top-scoring metrics in 19/25 test settings. Analyses demonstrate our method robustly and efficiently approaches human-like evaluations, enjoying better interpretability. Our code and scripts will be publicly released in https://github.com/Coldmist-Lu/ErrorAnalysis_NLGEvaluation.
%R 10.18653/v1/2023.acl-long.324
%U https://aclanthology.org/2023.acl-long.324
%U https://doi.org/10.18653/v1/2023.acl-long.324
%P 5892-5907

Markdown (Informal)

[Toward Human-Like Evaluation for Natural Language Generation with Error Analysis](https://aclanthology.org/2023.acl-long.324) (Lu et al., ACL 2023)

• Toward Human-Like Evaluation for Natural Language Generation with Error Analysis (Lu et al., ACL 2023)

ACL

• Qingyu Lu, Liang Ding, Liping Xie, Kanjian Zhang, Derek F. Wong, and Dacheng Tao. 2023. Toward Human-Like Evaluation for Natural Language Generation with Error Analysis. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 5892–5907, Toronto, Canada. Association for Computational Linguistics.