<?xml version="1.0" encoding="utf-8"?>
<TEI xmlns="http://www.tei-c.org/ns/1.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:hal="http://hal.archives-ouvertes.fr/" xmlns:gml="http://www.opengis.net/gml/3.3/" xmlns:gmlce="http://www.opengis.net/gml/3.3/ce" version="1.1" xsi:schemaLocation="http://www.tei-c.org/ns/1.0 http://api.archives-ouvertes.fr/documents/aofr-sword.xsd">
  <teiHeader>
    <fileDesc>
      <titleStmt>
        <title>HAL TEI export of hal-05335708</title>
      </titleStmt>
      <publicationStmt>
        <distributor>CCSD</distributor>
        <availability status="restricted">
          <licence target="https://creativecommons.org/publicdomain/zero/1.0/">CC0 1.0 - Universal</licence>
        </availability>
        <date when="2026-05-22T19:15:29+02:00"/>
      </publicationStmt>
      <sourceDesc>
        <p part="N">HAL API Platform</p>
      </sourceDesc>
    </fileDesc>
  </teiHeader>
  <text>
    <body>
      <listBibl>
        <biblFull>
          <titleStmt>
            <title xml:lang="en">Beyond BLEU: Ethical Risks of Misleading Evaluation in Domain-Specific QA with LLMs</title>
            <author role="aut">
              <persName>
                <forename type="first">Ayoub</forename>
                <surname>Nainia</surname>
              </persName>
              <email type="md5">a473aa72120d90a22915322348223291</email>
              <email type="domain">sorbonne-universite.fr</email>
              <idno type="idhal" notation="string">ayoub-nainia</idno>
              <idno type="idhal" notation="numeric">1587246</idno>
              <idno type="halauthorid" notation="string">3119909-1587246</idno>
              <idno type="GOOGLE SCHOLAR">https://scholar.google.fr/citations?user=_lR9ijMAAAAJ</idno>
              <affiliation ref="#struct-542193"/>
            </author>
            <author role="aut">
              <persName>
                <forename type="first">Régine</forename>
                <surname>Vignes-Lebbe</surname>
              </persName>
              <email type="md5">9016567b0f92b2da3dd56f775ce12ba5</email>
              <email type="domain">mnhn.fr</email>
              <idno type="idhal" notation="string">regine-vignes-lebbe</idno>
              <idno type="idhal" notation="numeric">176959</idno>
              <idno type="halauthorid" notation="string">932140-176959</idno>
              <idno type="ORCID">https://orcid.org/0000-0002-6912-6248</idno>
              <idno type="IDREF">https://www.idref.fr/081808097</idno>
              <affiliation ref="#struct-542193"/>
            </author>
            <author role="aut">
              <persName>
                <forename type="first">Hajar</forename>
                <surname>Mousannif</surname>
              </persName>
              <email type="md5">eb0f1db02cf155a567550b86e8423699</email>
              <email type="domain">uca.ac.ma</email>
              <idno type="idhal" notation="numeric">1372522</idno>
              <idno type="halauthorid" notation="string">1080381-1372522</idno>
              <affiliation ref="#struct-302207"/>
            </author>
            <author role="aut">
              <persName>
                <forename type="first">Jihad</forename>
                <surname>Zahir</surname>
              </persName>
              <email type="md5">a6793196c962dc117eb1ae32dcb6e399</email>
              <email type="domain">uca.ac.ma</email>
              <idno type="idhal" notation="numeric">1372523</idno>
              <idno type="halauthorid" notation="string">2888746-1372523</idno>
              <idno type="ORCID">https://orcid.org/0000-0002-5344-8277</idno>
              <idno type="IDREF">https://www.idref.fr/284465690</idno>
              <affiliation ref="#struct-302207"/>
              <affiliation ref="#struct-541946"/>
            </author>
            <editor role="depositor">
              <persName>
                <forename>Ayoub</forename>
                <surname>Nainia</surname>
              </persName>
              <email type="md5">a473aa72120d90a22915322348223291</email>
              <email type="domain">sorbonne-universite.fr</email>
            </editor>
            <funder ref="#projanr-57906"/>
            <funder ref="#projeurop-2797888"/>
            <funder>This work is co-funded by the European Union’s Horizon Europe research and innovation program Cofund SOUND.AI under the Marie Sklodowska Curie Grant Agreement No 101081674. This research is also part of the e-COL+ project (ANR-21-ESRE-0053).</funder>
          </titleStmt>
          <editionStmt>
            <edition n="v1" type="current">
              <date type="whenSubmitted">2025-10-28 18:10:59</date>
              <date type="whenModified">2026-02-07 05:32:42</date>
              <date type="whenReleased">2025-10-29 08:30:07</date>
              <date type="whenProduced">2025-09-08</date>
              <date type="whenEndEmbargoed">2025-10-28</date>
              <ref type="file" target="https://hal.science/hal-05335708v1/document">
                <date notBefore="2025-10-28"/>
              </ref>
              <ref type="file" subtype="author" n="1" target="https://hal.science/hal-05335708v1/file/BeyoundBleu-AyoubNainia-19.pdf" id="file-5335708-4568484">
                <date notBefore="2025-10-28"/>
              </ref>
            </edition>
            <respStmt>
              <resp>contributor</resp>
              <name key="1528429">
                <persName>
                  <forename>Ayoub</forename>
                  <surname>Nainia</surname>
                </persName>
                <email type="md5">a473aa72120d90a22915322348223291</email>
                <email type="domain">sorbonne-universite.fr</email>
              </name>
            </respStmt>
          </editionStmt>
          <publicationStmt>
            <distributor>CCSD</distributor>
            <idno type="halId">hal-05335708</idno>
            <idno type="halUri">https://hal.science/hal-05335708</idno>
            <idno type="halBibtex">nainia:hal-05335708</idno>
            <idno type="halRefHtml">&lt;i&gt;Proceedings of the First Workshop on Comparative Performance Evaluation: From Rules to Language Models associated with RANLP 2025&lt;/i&gt;, Lancaster University; UK University of Alicante; Institute of Information and Communication Technologies, BAS; Bulgarian Association for Computational Linguistics, Sep 2025, Varna, Bulgaria. pp.77-86, &lt;a target="_blank" href="https://dx.doi.org/10.26615/978-954-452-102-8_009"&gt;&amp;#x27E8;10.26615/978-954-452-102-8_009&amp;#x27E9;&lt;/a&gt;</idno>
            <idno type="halRef">Proceedings of the First Workshop on Comparative Performance Evaluation: From Rules to Language Models associated with RANLP 2025, Lancaster University; UK University of Alicante; Institute of Information and Communication Technologies, BAS; Bulgarian Association for Computational Linguistics, Sep 2025, Varna, Bulgaria. pp.77-86, &amp;#x27E8;10.26615/978-954-452-102-8_009&amp;#x27E9;</idno>
            <availability status="restricted">
              <licence target="https://creativecommons.org/licenses/by/4.0/">CC BY 4.0 - Attribution<ref corresp="#file-5335708-4568484"/></licence>
            </availability>
          </publicationStmt>
          <seriesStmt>
            <idno type="stamp" n="IRD">IRD - Institut de recherche pour le développement</idno>
            <idno type="stamp" n="MNHN">Muséum National d'Histoire Naturelle</idno>
            <idno type="stamp" n="EPHE">École Pratique des Hautes Études</idno>
            <idno type="stamp" n="UNIV-AG">Université des Antilles</idno>
            <idno type="stamp" n="AFRIQ">HAL-Francophonie, Afrique et océan indien</idno>
            <idno type="stamp" n="CNRS">CNRS - Centre national de la recherche scientifique</idno>
            <idno type="stamp" n="OPENAIRE">OpenAIRE</idno>
            <idno type="stamp" n="ISYEB" corresp="MNHN">Institut de Systématique, Evolution, Biodiversité</idno>
            <idno type="stamp" n="PSL">Université Paris sciences et lettres</idno>
            <idno type="stamp" n="SORBONNE-UNIVERSITE">Sorbonne Université</idno>
            <idno type="stamp" n="SORBONNE-UNIV" corresp="SORBONNE-UNIVERSITE">Sorbonne Université 01/01/2018</idno>
            <idno type="stamp" n="SU-SCIENCES" corresp="SORBONNE-UNIVERSITE">Faculté des Sciences de Sorbonne Université</idno>
            <idno type="stamp" n="TEST-DEV">TEST-DEV</idno>
            <idno type="stamp" n="UMI-209" corresp="SORBONNE-UNIVERSITE">Unité de modélisation mathématique et informatique des systèmes complexes</idno>
            <idno type="stamp" n="EPHE-PSL" corresp="PSL">École Pratique des Hautes Études - PSL</idno>
            <idno type="stamp" n="SU-TI">Sorbonne Université - Texte Intégral</idno>
            <idno type="stamp" n="ANR">ANR</idno>
            <idno type="stamp" n="ALLIANCE-SU"> Alliance Sorbonne Université</idno>
            <idno type="stamp" n="INEE-CNRS">Institut écologie et environnement du CNRS</idno>
            <idno type="stamp" n="SUPRA_INGENIERIE">Ingénierie</idno>
            <idno type="stamp" n="SUPRA_SCIENCES_NATURELLES">MNHN</idno>
          </seriesStmt>
          <notesStmt>
            <note type="audience" n="2">International</note>
            <note type="invited" n="0">No</note>
            <note type="popular" n="0">No</note>
            <note type="peer" n="1">Yes</note>
            <note type="proceedings" n="1">Yes</note>
          </notesStmt>
          <sourceDesc>
            <biblStruct>
              <analytic>
                <title xml:lang="en">Beyond BLEU: Ethical Risks of Misleading Evaluation in Domain-Specific QA with LLMs</title>
                <author role="aut">
                  <persName>
                    <forename type="first">Ayoub</forename>
                    <surname>Nainia</surname>
                  </persName>
                  <email type="md5">a473aa72120d90a22915322348223291</email>
                  <email type="domain">sorbonne-universite.fr</email>
                  <idno type="idhal" notation="string">ayoub-nainia</idno>
                  <idno type="idhal" notation="numeric">1587246</idno>
                  <idno type="halauthorid" notation="string">3119909-1587246</idno>
                  <idno type="GOOGLE SCHOLAR">https://scholar.google.fr/citations?user=_lR9ijMAAAAJ</idno>
                  <affiliation ref="#struct-542193"/>
                </author>
                <author role="aut">
                  <persName>
                    <forename type="first">Régine</forename>
                    <surname>Vignes-Lebbe</surname>
                  </persName>
                  <email type="md5">9016567b0f92b2da3dd56f775ce12ba5</email>
                  <email type="domain">mnhn.fr</email>
                  <idno type="idhal" notation="string">regine-vignes-lebbe</idno>
                  <idno type="idhal" notation="numeric">176959</idno>
                  <idno type="halauthorid" notation="string">932140-176959</idno>
                  <idno type="ORCID">https://orcid.org/0000-0002-6912-6248</idno>
                  <idno type="IDREF">https://www.idref.fr/081808097</idno>
                  <affiliation ref="#struct-542193"/>
                </author>
                <author role="aut">
                  <persName>
                    <forename type="first">Hajar</forename>
                    <surname>Mousannif</surname>
                  </persName>
                  <email type="md5">eb0f1db02cf155a567550b86e8423699</email>
                  <email type="domain">uca.ac.ma</email>
                  <idno type="idhal" notation="numeric">1372522</idno>
                  <idno type="halauthorid" notation="string">1080381-1372522</idno>
                  <affiliation ref="#struct-302207"/>
                </author>
                <author role="aut">
                  <persName>
                    <forename type="first">Jihad</forename>
                    <surname>Zahir</surname>
                  </persName>
                  <email type="md5">a6793196c962dc117eb1ae32dcb6e399</email>
                  <email type="domain">uca.ac.ma</email>
                  <idno type="idhal" notation="numeric">1372523</idno>
                  <idno type="halauthorid" notation="string">2888746-1372523</idno>
                  <idno type="ORCID">https://orcid.org/0000-0002-5344-8277</idno>
                  <idno type="IDREF">https://www.idref.fr/284465690</idno>
                  <affiliation ref="#struct-302207"/>
                  <affiliation ref="#struct-541946"/>
                </author>
              </analytic>
              <monogr>
                <idno type="isbn">978-954-452-102-8</idno>
                <title level="m">Proceedings of the First Workshop on Comparative Performance Evaluation: From Rules to Language Models associated with RANLP 2025</title>
                <meeting>
                  <title>Proceedings of the First Workshop on Comparative Performance Evaluation: From Rules to Language Models associated with RANLP 2025</title>
                  <date type="start">2025-09-08</date>
                  <date type="end">2025-09-10</date>
                  <settlement>Varna</settlement>
                  <country key="BG">Bulgaria</country>
                </meeting>
                <respStmt>
                  <resp>conferenceOrganizer</resp>
                  <name>Lancaster University</name>
                  <name>UK University of Alicante</name>
                  <name>Institute of Information and Communication Technologies, BAS</name>
                  <name>Bulgarian Association for Computational Linguistics</name>
                </respStmt>
                <editor>ACL Anthology</editor>
                <imprint>
                  <biblScope unit="pp">77-86</biblScope>
                </imprint>
              </monogr>
              <idno type="doi">10.26615/978-954-452-102-8_009</idno>
              <ref target="https://acl-bg.org/proceedings/2025/R2LM%202025/pdf/2025.r2lm-1.9.pdf" type="seeAlso"/>
              <ref type="publisher">https://ranlp.org/ranlp2025/</ref>
            </biblStruct>
          </sourceDesc>
          <profileDesc>
            <langUsage>
              <language ident="en">English</language>
            </langUsage>
            <textClass>
              <keywords scheme="author">
                <term xml:lang="en">AI Evaluation</term>
                <term xml:lang="en">Evaluation LLM</term>
                <term xml:lang="en">Question Answering</term>
                <term xml:lang="en">Large Language Model</term>
              </keywords>
              <classCode scheme="halDomain" n="info.info-ai">Computer Science [cs]/Artificial Intelligence [cs.AI]</classCode>
              <classCode scheme="halTypology" n="COMM">Conference papers</classCode>
              <classCode scheme="halOldTypology" n="COMM">Conference papers</classCode>
              <classCode scheme="halTreeTypology" n="COMM">Conference papers</classCode>
            </textClass>
            <abstract xml:lang="en">
              <p>Large Language Models (LLMs) are increasingly used in scientific question answering (QA), including high-stakes fields such as biodiversity informatics. However, standard evaluation metrics such as BLEU, ROUGE, Exact Match (EM), and BERTScore remain poorly aligned with the factual and domain-specific requirements of these tasks. In this work, we investigate the gap between automatic metrics and expert judgment in botanical QA by comparing metric scores with human ratings across five dimensions: accuracy, completeness, relevance, fluency, and terminology usage. Our results show that standard metrics often misrepresent response quality, particularly in the presence of paraphrasing, omission, or domain-specific language. Through both quantitative analysis and qualitative examples, we show that high-scoring responses may still exhibit critical factual errors or omissions. These findings highlight the need for domain-aware evaluation frameworks that incorporate expert feedback and raise important ethical concerns about the deployment of LLMs in scientific contexts.</p>
            </abstract>
          </profileDesc>
        </biblFull>
      </listBibl>
    </body>
    <back>
      <listOrg type="structures">
        <org type="laboratory" xml:id="struct-542193" status="VALID">
          <idno type="IdRef">191205702</idno>
          <idno type="ISNI">0000000403707618</idno>
          <idno type="RNSR">200511872M</idno>
          <idno type="ROR">https://ror.org/01dadvw90</idno>
          <orgName>Institut de Systématique, Evolution, Biodiversité</orgName>
          <orgName type="acronym">ISYEB</orgName>
          <date type="start">2017-10-09</date>
          <desc>
            <address>
              <addrLine>57 rue Cuvier - CP 50 - 75005 Paris</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">http://isyeb.mnhn.fr/</ref>
          </desc>
          <listRelation>
            <relation active="#struct-7512" type="direct"/>
            <relation active="#struct-110691" type="direct"/>
            <relation active="#struct-564132" type="indirect"/>
            <relation active="#struct-413221" type="direct"/>
            <relation name="UMR7205" active="#struct-441569" type="direct"/>
            <relation active="#struct-524399" type="direct"/>
          </listRelation>
        </org>
        <org type="regroupinstitution" xml:id="struct-302207" status="VALID">
          <idno type="ROR">https://ror.org/04xf6nm78</idno>
          <orgName>Université Cadi Ayyad = Cadi Ayyad University [Marrakech]</orgName>
          <orgName type="acronym">UCA</orgName>
          <desc>
            <address>
              <addrLine>Av Abdelkrim Khattabi, B.P. 511 - 40000 - Marrakech</addrLine>
              <country key="MA"/>
            </address>
            <ref type="url">https://www.uca.ma/</ref>
          </desc>
        </org>
        <org type="laboratory" xml:id="struct-541946" status="VALID">
          <idno type="IdRef">242450806</idno>
          <idno type="RNSR">201421254W</idno>
          <orgName>Unité de modélisation mathématique et informatique des systèmes complexes [Bondy]</orgName>
          <orgName type="acronym">UMMISCO</orgName>
          <desc>
            <address>
              <addrLine>IRD France Nord - 32 avenue Henri Varagnat - 93143 Bondy cedex</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">http://www.ummisco.ird.fr</ref>
          </desc>
          <listRelation>
            <relation active="#struct-48348" type="direct"/>
            <relation active="#struct-217395" type="direct"/>
            <relation active="#struct-301775" type="direct"/>
            <relation active="#struct-302207" type="direct"/>
            <relation name="UMI209" active="#struct-413221" type="direct"/>
            <relation active="#struct-451861" type="direct"/>
            <relation active="#struct-1175215" type="direct"/>
          </listRelation>
        </org>
        <org type="institution" xml:id="struct-7512" status="VALID">
          <idno type="IdRef">026394944</idno>
          <idno type="ROR">https://ror.org/03wkt5x30</idno>
          <orgName>Muséum national d'Histoire naturelle</orgName>
          <orgName type="acronym">MNHN</orgName>
          <desc>
            <address>
              <addrLine>57, rue Cuvier - 75231 Paris Cedex 05</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">http://www.mnhn.fr</ref>
          </desc>
        </org>
        <org type="institution" xml:id="struct-110691" status="VALID">
          <idno type="IdRef">026375478</idno>
          <idno type="ISNI">0000000121955365</idno>
          <idno type="ROR">https://ror.org/046b3cj80</idno>
          <orgName>École Pratique des Hautes Études</orgName>
          <orgName type="acronym">EPHE</orgName>
          <date type="start">1868-01-01</date>
          <desc>
            <address>
              <addrLine>4-14 Rue Ferrus, 75014 Paris</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">http://www.ephe.fr</ref>
          </desc>
          <listRelation>
            <relation active="#struct-564132" type="direct"/>
          </listRelation>
        </org>
        <org type="regroupinstitution" xml:id="struct-564132" status="VALID">
          <idno type="IdRef">241597595</idno>
          <idno type="ISNI">0000 0004 1784 3645</idno>
          <idno type="ROR">https://ror.org/013cjyk83</idno>
          <orgName>Université Paris Sciences et Lettres</orgName>
          <orgName type="acronym">PSL</orgName>
          <desc>
            <address>
              <addrLine>60 rue Mazarine 75006 Paris</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">https://www.psl.eu/</ref>
          </desc>
        </org>
        <org type="regroupinstitution" xml:id="struct-413221" status="VALID">
          <idno type="IdRef">221333754</idno>
          <idno type="ROR">https://ror.org/02en5vm52</idno>
          <orgName>Sorbonne Université</orgName>
          <orgName type="acronym">SU</orgName>
          <date type="start">2018-01-01</date>
          <desc>
            <address>
              <addrLine>21 rue de l’École de médecine - 75006 Paris</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">http://www.sorbonne-universite.fr/</ref>
          </desc>
        </org>
        <org type="regroupinstitution" xml:id="struct-441569" status="VALID">
          <idno type="IdRef">02636817X</idno>
          <idno type="ISNI">0000000122597504</idno>
          <idno type="ROR">https://ror.org/02feahw73</idno>
          <orgName>Centre National de la Recherche Scientifique</orgName>
          <orgName type="acronym">CNRS</orgName>
          <date type="start">1939-10-19</date>
          <desc>
            <address>
              <country key="FR"/>
            </address>
            <ref type="url">https://www.cnrs.fr/</ref>
          </desc>
        </org>
        <org type="regroupinstitution" xml:id="struct-524399" status="VALID">
          <idno type="IdRef">187841578</idno>
          <idno type="ROR">https://ror.org/02ryfmr77</idno>
          <orgName>Université des Antilles</orgName>
          <orgName type="acronym">UA</orgName>
          <date type="start">2015-01-01</date>
          <desc>
            <address>
              <addrLine>Université des Antilles | Fouillole - BP 250 - 97157 Pointe-à-Pitre</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">http://www.univ-antilles.fr/</ref>
          </desc>
        </org>
        <org type="institution" xml:id="struct-48348" status="VALID">
          <idno type="IdRef">03484404X</idno>
          <idno type="ROR">https://ror.org/01jp0tk64</idno>
          <orgName>Université Gaston Berger de Saint-Louis Sénégal</orgName>
          <orgName type="acronym">UGB</orgName>
          <date type="start">1990-01-01</date>
          <desc>
            <address>
              <addrLine>Route de Ngallèle, B.P. 234 Saint-Louis</addrLine>
              <country key="SN"/>
            </address>
            <ref type="url">http://www.ugb.sn</ref>
          </desc>
        </org>
        <org type="regroupinstitution" xml:id="struct-217395" status="VALID">
          <idno type="ISNI">0000000121738504</idno>
          <idno type="ROR">https://ror.org/022zbs961</idno>
          <orgName>Université de Yaoundé I</orgName>
          <orgName type="acronym">UY1</orgName>
          <date type="start">1993-01-29</date>
          <desc>
            <address>
              <addrLine>B.P. 337 Yaoundé</addrLine>
              <country key="CM"/>
            </address>
            <ref type="url">https://uy1.uninet.cm/</ref>
          </desc>
        </org>
        <org type="institution" xml:id="struct-301775" status="VALID">
          <orgName>Institut de la francophonie pour l'informatique</orgName>
          <desc>
            <address>
              <country key="FR"/>
            </address>
          </desc>
        </org>
        <org type="institution" xml:id="struct-451861" status="VALID">
          <orgName>Institut de Recherche pour le Développement</orgName>
          <orgName type="acronym">IRD [Ile-de-France]</orgName>
          <desc>
            <address>
              <addrLine>32, avenue Henri Varagnat93143 Bondy cedex</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">https://ile-de-france.ird.fr/</ref>
          </desc>
        </org>
        <org type="institution" xml:id="struct-1175215" status="VALID">
          <idno type="ROR">https://ror.org/04je6yw13</idno>
          <orgName>Université Cheikh Anta Diop de Dakar [Sénégal]</orgName>
          <orgName type="acronym">UCAD</orgName>
          <date type="start">1987-01-01</date>
          <desc>
            <address>
              <addrLine>Université Cheikh Anta Diop de Dakar - Rectorat, BP 5005 Dakar-fann</addrLine>
              <country key="SN"/>
            </address>
            <ref type="url">https://www.ucad.sn/</ref>
          </desc>
        </org>
      </listOrg>
      <listOrg type="projects">
        <org type="anrProject" xml:id="projanr-57906" status="VALID">
          <idno type="anr">ANR-21-ESRE-0053</idno>
          <orgName>E-COL+</orgName>
          <desc>e-COL+, the enhancement of natural history data in France</desc>
          <date type="start">2021</date>
        </org>
        <org type="europeanProject" xml:id="projeurop-2797888" status="VALID">
          <idno type="number">101081674</idno>
          <idno type="program">HORIZON-MSCA-2021-COFUND-01</idno>
          <idno type="call">HORIZON-MSCA-2021-COFUND-01</idno>
          <orgName>SOUND.AI</orgName>
          <desc>Sorbonne University for a New Deal on Artificial Intelligence</desc>
          <date type="start">2023-01-01</date>
          <date type="end">2027-12-31</date>
        </org>
      </listOrg>
    </back>
  </text>
</TEI>