<?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-03423971</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="2026-05-17T03:25:55+02:00"/>
      </publicationStmt>
      <sourceDesc>
        <p part="N">HAL API Platform</p>
      </sourceDesc>
    </fileDesc>
  </teiHeader>
  <text>
    <body>
      <listBibl>
        <biblFull>
          <titleStmt>
            <title xml:lang="en">Automatic Text Evaluation through the Lens of Wasserstein Barycenters</title>
            <author role="aut">
              <persName>
                <forename type="first">Pierre</forename>
                <surname>Colombo</surname>
              </persName>
              <email type="md5">e2d5dac3d92757ed48e37e126a3464a3</email>
              <email type="domain">gmail.com</email>
              <idno type="idhal" notation="string">pierre-colombo</idno>
              <idno type="idhal" notation="numeric">743646</idno>
              <idno type="halauthorid" notation="string">1007669-743646</idno>
              <idno type="GOOGLE SCHOLAR">https://scholar.google.com/citations?user=yPoMt8gAAAAJ&amp;hl=fr</idno>
              <idno type="IDREF">https://www.idref.fr/258975059</idno>
              <affiliation ref="#struct-569943"/>
              <affiliation ref="#struct-554452"/>
              <affiliation ref="#struct-554512"/>
            </author>
            <author role="aut">
              <persName>
                <forename type="first">Guillaume</forename>
                <surname>Staerman</surname>
              </persName>
              <email type="md5">ba00e776b3beff8866402cbd0b3329dd</email>
              <email type="domain">telecom-paris.fr</email>
              <idno type="idhal" notation="string">guillaume-staerman</idno>
              <idno type="idhal" notation="numeric">743604</idno>
              <idno type="halauthorid" notation="string">47196-743604</idno>
              <idno type="GOOGLE SCHOLAR">https://scholar.google.com/citations?user=Zb2ax0wAAAAJ&amp;hl=fr</idno>
              <affiliation ref="#struct-554452"/>
              <affiliation ref="#struct-554512"/>
            </author>
            <author role="aut">
              <persName>
                <forename type="first">Pablo</forename>
                <surname>Piantanida</surname>
              </persName>
              <email type="md5">4d99eb094c1c3424e1f0788dcb8f365d</email>
              <email type="domain">centralesupelec.fr</email>
              <idno type="idhal" notation="string">pablo-piantanida</idno>
              <idno type="idhal" notation="numeric">736967</idno>
              <idno type="halauthorid" notation="string">28498-736967</idno>
              <idno type="ORCID">https://orcid.org/0000-0002-8717-2117</idno>
              <idno type="IDREF">https://www.idref.fr/116306300</idno>
              <affiliation ref="#struct-1289"/>
            </author>
            <author role="aut">
              <persName>
                <forename type="first">Chloé</forename>
                <surname>Clavel</surname>
              </persName>
              <email type="md5">6ba9d6af7b9330a658fc5c43473ce641</email>
              <email type="domain">telecom-paristech.fr</email>
              <idno type="idhal" notation="string">chloe-clavel</idno>
              <idno type="idhal" notation="numeric">745667</idno>
              <idno type="halauthorid" notation="string">10977-745667</idno>
              <idno type="ORCID">https://orcid.org/0000-0003-4850-3398</idno>
              <idno type="GOOGLE SCHOLAR">https://scholar.google.fr/citations?user=TAZbfksAAAAJ&amp;hl=en</idno>
              <idno type="IDREF">https://www.idref.fr/116841281</idno>
              <affiliation ref="#struct-554452"/>
              <affiliation ref="#struct-554512"/>
            </author>
            <editor role="depositor">
              <persName>
                <forename>Pierre</forename>
                <surname>COLOMBO</surname>
              </persName>
              <email type="md5">e2d5dac3d92757ed48e37e126a3464a3</email>
              <email type="domain">gmail.com</email>
            </editor>
          </titleStmt>
          <editionStmt>
            <edition n="v1" type="current">
              <date type="whenSubmitted">2024-05-14 14:56:19</date>
              <date type="whenModified">2026-02-03 03:30:57</date>
              <date type="whenReleased">2024-05-16 08:27:01</date>
              <date type="whenProduced">2021-11-07</date>
              <date type="whenEndEmbargoed">2024-05-14</date>
              <ref type="file" target="https://hal.science/hal-03423971v1/document">
                <date notBefore="2024-05-14"/>
              </ref>
              <ref type="file" subtype="greenPublisher" n="1" target="https://hal.science/hal-03423971v1/file/2021.emnlp-main.817.pdf" id="file-4574844-3970133">
                <date notBefore="2024-05-14"/>
              </ref>
              <ref type="externalLink" target="http://arxiv.org/pdf/2108.12463"/>
            </edition>
            <respStmt>
              <resp>contributor</resp>
              <name key="987967">
                <persName>
                  <forename>Pierre</forename>
                  <surname>COLOMBO</surname>
                </persName>
                <email type="md5">e2d5dac3d92757ed48e37e126a3464a3</email>
                <email type="domain">gmail.com</email>
              </name>
            </respStmt>
          </editionStmt>
          <publicationStmt>
            <distributor>CCSD</distributor>
            <idno type="halId">hal-03423971</idno>
            <idno type="halUri">https://hal.science/hal-03423971</idno>
            <idno type="halBibtex">colombo:hal-03423971</idno>
            <idno type="halRefHtml">&lt;i&gt;EMNLP 2021&lt;/i&gt;, Nov 2021, Punta Cana, Dominican Republic. pp.10450-10466, &lt;a target="_blank" href="https://dx.doi.org/10.18653/v1/2021.emnlp-main.817"&gt;&amp;#x27E8;10.18653/v1/2021.emnlp-main.817&amp;#x27E9;&lt;/a&gt;</idno>
            <idno type="halRef">EMNLP 2021, Nov 2021, Punta Cana, Dominican Republic. pp.10450-10466, &amp;#x27E8;10.18653/v1/2021.emnlp-main.817&amp;#x27E9;</idno>
            <availability status="restricted">
              <licence target="https://creativecommons.org/licenses/by/4.0/">CC BY 4.0 - Attribution<ref corresp="#file-4574844-3970133"/></licence>
            </availability>
          </publicationStmt>
          <seriesStmt>
            <idno type="stamp" n="CNRS">CNRS - Centre national de la recherche scientifique</idno>
            <idno type="stamp" n="UNIV-PSUD">Université Paris Sud - Paris XI</idno>
            <idno type="stamp" n="ENST">Ecole Nationale Supérieure des Télécommunications</idno>
            <idno type="stamp" n="SUP_LSS">Laboratoire des Signaux et Systèmes</idno>
            <idno type="stamp" n="TELECOM-PARISTECH" corresp="PARISTECH">Télécom Paris</idno>
            <idno type="stamp" n="SUP_TELECOMS" corresp="SUP_LSS">Pôle Télécoms et Réseaux</idno>
            <idno type="stamp" n="CENTRALESUPELEC">Ecole CentraleSupélec</idno>
            <idno type="stamp" n="UNIV-PARIS-SACLAY">Université Paris-Saclay</idno>
            <idno type="stamp" n="UNIV-PSUD-SACLAY" corresp="UNIV-PARIS-SACLAY">Université Paris Sud pour Paris Saclay</idno>
            <idno type="stamp" n="LTCI" corresp="TELECOM-PARISTECH">Laboratoire Traitement et Communication de l'Information</idno>
            <idno type="stamp" n="IDS" corresp="TELECOM-PARISTECH">Département Image, Données, Signal</idno>
            <idno type="stamp" n="S2A" corresp="TELECOM-PARISTECH">Equipe Signal, Statistique et Apprentissage</idno>
            <idno type="stamp" n="IP_PARIS">Institut Polytechnique de Paris</idno>
            <idno type="stamp" n="INSTITUTS-TELECOM">composantes instituts telecom </idno>
            <idno type="stamp" n="UNIVERSITE-PARIS-SACLAY" corresp="UNIV-PARIS-SACLAY">Université Paris-Saclay</idno>
            <idno type="stamp" n="GS-ENGINEERING">Graduate School Sciences de l'Ingénierie et des Systèmes</idno>
            <idno type="stamp" n="GS-COMPUTER-SCIENCE">Graduate School Computer Science</idno>
            <idno type="stamp" n="HUB-IA">Hub Intelligence Artificielle de CentraleSupélec</idno>
            <idno type="stamp" n="INSTITUT-MINES-TELECOM">Institut Mines Telecom</idno>
            <idno type="stamp" n="DEPARTEMENT-DE-MATHEMATIQUES">Collection du Département de Mathématiques</idno>
            <idno type="stamp" n="IP-PARIS-DEPARTEMENT-MATHEMATIQUES">Département de Mathèmatiques</idno>
            <idno type="stamp" n="IP-PARIS-MATHEMATIQUES">IP Paris Département de Mathématiques</idno>
            <idno type="stamp" n="IP-PARIS-INFORMATION-COMMUNICATION-ELECTRONIQUE">IP Paris Département d'Information, de Communication et d'Électronique</idno>
            <idno type="stamp" n="IP-PARIS-INFORMATIQUE-DONNEES-ET-IA">IP Paris Département d'Informatique, de Données et d'IA</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">Automatic Text Evaluation through the Lens of Wasserstein Barycenters</title>
                <author role="aut">
                  <persName>
                    <forename type="first">Pierre</forename>
                    <surname>Colombo</surname>
                  </persName>
                  <email type="md5">e2d5dac3d92757ed48e37e126a3464a3</email>
                  <email type="domain">gmail.com</email>
                  <idno type="idhal" notation="string">pierre-colombo</idno>
                  <idno type="idhal" notation="numeric">743646</idno>
                  <idno type="halauthorid" notation="string">1007669-743646</idno>
                  <idno type="GOOGLE SCHOLAR">https://scholar.google.com/citations?user=yPoMt8gAAAAJ&amp;hl=fr</idno>
                  <idno type="IDREF">https://www.idref.fr/258975059</idno>
                  <affiliation ref="#struct-569943"/>
                  <affiliation ref="#struct-554452"/>
                  <affiliation ref="#struct-554512"/>
                </author>
                <author role="aut">
                  <persName>
                    <forename type="first">Guillaume</forename>
                    <surname>Staerman</surname>
                  </persName>
                  <email type="md5">ba00e776b3beff8866402cbd0b3329dd</email>
                  <email type="domain">telecom-paris.fr</email>
                  <idno type="idhal" notation="string">guillaume-staerman</idno>
                  <idno type="idhal" notation="numeric">743604</idno>
                  <idno type="halauthorid" notation="string">47196-743604</idno>
                  <idno type="GOOGLE SCHOLAR">https://scholar.google.com/citations?user=Zb2ax0wAAAAJ&amp;hl=fr</idno>
                  <affiliation ref="#struct-554452"/>
                  <affiliation ref="#struct-554512"/>
                </author>
                <author role="aut">
                  <persName>
                    <forename type="first">Pablo</forename>
                    <surname>Piantanida</surname>
                  </persName>
                  <email type="md5">4d99eb094c1c3424e1f0788dcb8f365d</email>
                  <email type="domain">centralesupelec.fr</email>
                  <idno type="idhal" notation="string">pablo-piantanida</idno>
                  <idno type="idhal" notation="numeric">736967</idno>
                  <idno type="halauthorid" notation="string">28498-736967</idno>
                  <idno type="ORCID">https://orcid.org/0000-0002-8717-2117</idno>
                  <idno type="IDREF">https://www.idref.fr/116306300</idno>
                  <affiliation ref="#struct-1289"/>
                </author>
                <author role="aut">
                  <persName>
                    <forename type="first">Chloé</forename>
                    <surname>Clavel</surname>
                  </persName>
                  <email type="md5">6ba9d6af7b9330a658fc5c43473ce641</email>
                  <email type="domain">telecom-paristech.fr</email>
                  <idno type="idhal" notation="string">chloe-clavel</idno>
                  <idno type="idhal" notation="numeric">745667</idno>
                  <idno type="halauthorid" notation="string">10977-745667</idno>
                  <idno type="ORCID">https://orcid.org/0000-0003-4850-3398</idno>
                  <idno type="GOOGLE SCHOLAR">https://scholar.google.fr/citations?user=TAZbfksAAAAJ&amp;hl=en</idno>
                  <idno type="IDREF">https://www.idref.fr/116841281</idno>
                  <affiliation ref="#struct-554452"/>
                  <affiliation ref="#struct-554512"/>
                </author>
              </analytic>
              <monogr>
                <title level="m">Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing</title>
                <meeting>
                  <title>EMNLP 2021</title>
                  <date type="start">2021-11-07</date>
                  <date type="end">2021-11-11</date>
                  <settlement>Punta Cana</settlement>
                  <country key="DO">Dominican Republic</country>
                </meeting>
                <imprint>
                  <biblScope unit="pp">10450–10466</biblScope>
                  <date type="datePub">2021</date>
                </imprint>
              </monogr>
              <idno type="arxiv">2108.12463</idno>
              <idno type="doi">10.18653/v1/2021.emnlp-main.817</idno>
            </biblStruct>
          </sourceDesc>
          <profileDesc>
            <langUsage>
              <language ident="en">English</language>
            </langUsage>
            <textClass>
              <classCode scheme="halDomain" n="info.info-tt">Computer Science [cs]/Document and Text Processing</classCode>
              <classCode scheme="halDomain" n="stat">Statistics [stat]</classCode>
              <classCode scheme="halDomain" n="stat.ml">Statistics [stat]/Machine Learning [stat.ML]</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>A new metric BaryScore to evaluate text generation based on deep contextualized embeddings (e.g., BERT, Roberta, ELMo) is introduced. This metric is motivated by a new framework relying on optimal transport tools, i.e., Wasserstein distance and barycenter. By modelling the layer output of deep contextualized embeddings as a probability distribution rather than by a vector embedding; this framework provides a natural way to aggregate the different outputs through the Wasserstein space topology. In addition, it provides theoretical grounds to our metric and offers an alternative to available solutions (e.g., MoverScore and BertScore). Numerical evaluation is performed on four different tasks: machine translation, summarization, data2text generation and image captioning. Our results show that BaryScore outperforms other BERT based metrics and exhibits more consistent behaviour in particular for text summarization.</p>
            </abstract>
          </profileDesc>
        </biblFull>
      </listBibl>
    </body>
    <back>
      <listOrg type="structures">
        <org type="researchteam" xml:id="struct-569943" status="VALID">
          <orgName>IBM France Lab [Biot]</orgName>
          <date type="start">2019-09-09</date>
          <desc>
            <address>
              <addrLine>965 Avenue Roumanille, 06410 Biot</addrLine>
              <country key="FR"/>
            </address>
          </desc>
          <listRelation>
            <relation active="#struct-38357" type="direct"/>
            <relation active="#struct-300665" type="indirect"/>
          </listRelation>
        </org>
        <org type="researchteam" xml:id="struct-554452" status="VALID">
          <orgName>Signal, Statistique et Apprentissage</orgName>
          <orgName type="acronym">S2A</orgName>
          <date type="start">2019-02-22</date>
          <desc>
            <address>
              <addrLine>Télécom Paris 19 Place Marguerite Perey 91120 Palaiseau</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">https://www.telecom-paris.fr/fr/recherche/laboratoires/laboratoire-traitement-et-communication-de-linformation-ltci/les-equipes-de-recherche/signal-statistique-et-apprentissage-s2a</ref>
          </desc>
          <listRelation>
            <relation active="#struct-484335" type="direct"/>
            <relation active="#struct-302102" type="indirect"/>
            <relation active="#struct-1048346" type="indirect"/>
            <relation active="#struct-563936" type="indirect"/>
          </listRelation>
        </org>
        <org type="department" xml:id="struct-554512" status="VALID">
          <orgName>Département Images, Données, Signal</orgName>
          <orgName type="acronym">IDS</orgName>
          <date type="start">2017-01-01</date>
          <desc>
            <address>
              <addrLine>46, rue Barrault 75013 Paris ; 15 Place Marguerite Perey 91120 Palaiseau (depuis oct 2019)</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">https://www.telecom-paris.fr/fr/recherche/laboratoires/laboratoire-traitement-et-communication-de-linformation-ltci/les-departements-denseignement-et-recherche</ref>
          </desc>
          <listRelation>
            <relation active="#struct-300362" type="direct"/>
          </listRelation>
        </org>
        <org type="laboratory" xml:id="struct-1289" status="OLD">
          <idno type="IdRef">133818799</idno>
          <idno type="RNSR">199812842X</idno>
          <orgName>Laboratoire des signaux et systèmes</orgName>
          <orgName type="acronym">L2S</orgName>
          <date type="end">2019-12-31</date>
          <desc>
            <address>
              <addrLine>Plateau de Moulon 3 rue Joliot Curie 91192 GIF SUR YVETTE CEDEX</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">http://www.l2s.centralesupelec.fr/</ref>
          </desc>
          <listRelation>
            <relation active="#struct-92966" type="direct"/>
            <relation active="#struct-411575" type="direct"/>
            <relation name="UMR8506" active="#struct-441569" type="direct"/>
          </listRelation>
        </org>
        <org type="laboratory" xml:id="struct-38357" status="VALID">
          <orgName>IBM - Paris [Bois-Colombes]</orgName>
          <desc>
            <address>
              <addrLine>17 avenue de l'Europe 92275 Bois-Colombes Cedex</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">http://www.ibm.com/fr/fr/</ref>
          </desc>
          <listRelation>
            <relation active="#struct-300665" type="direct"/>
          </listRelation>
        </org>
        <org type="institution" xml:id="struct-300665" status="VALID">
          <orgName>IBM</orgName>
          <desc>
            <address>
              <country key="US"/>
            </address>
          </desc>
        </org>
        <org type="laboratory" xml:id="struct-484335" status="VALID">
          <idno type="IdRef">162384270</idno>
          <idno type="ISNI">0000 0000 9194 9502</idno>
          <idno type="RNSR">200319327Z</idno>
          <idno type="ROR">https://ror.org/057er4c39</idno>
          <orgName>Laboratoire Traitement et Communication de l'Information</orgName>
          <orgName type="acronym">LTCI</orgName>
          <date type="start">2017-01-01</date>
          <desc>
            <address>
              <addrLine>Télécom Paris 19 Place Marguerite Perey 91120 PALAISEAU</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">https://www.telecom-paris.fr/fr/recherche/laboratoires/laboratoire-traitement-et-communication-de-linformation-ltci</ref>
          </desc>
          <listRelation>
            <relation active="#struct-302102" type="direct"/>
            <relation active="#struct-1048346" type="direct"/>
            <relation active="#struct-563936" type="indirect"/>
          </listRelation>
        </org>
        <org type="regroupinstitution" xml:id="struct-302102" status="VALID">
          <idno type="IdRef">192427156</idno>
          <idno type="ISNI">000000012202567X</idno>
          <idno type="ROR">https://ror.org/025vp2923</idno>
          <idno type="Wikidata">Q27962533</idno>
          <orgName>Institut Mines-Télécom [Paris]</orgName>
          <orgName type="acronym">IMT</orgName>
          <date type="start">2012-03-01</date>
          <desc>
            <address>
              <addrLine>19 Place Marguerite Perey, 91120 Palaiseau</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">https://www.imt.fr/</ref>
          </desc>
        </org>
        <org type="institution" xml:id="struct-1048346" status="VALID">
          <idno type="IdRef">026375273</idno>
          <idno type="ISNI">0000 0001 2108 2779</idno>
          <idno type="ROR">https://ror.org/01naq7912</idno>
          <orgName>Télécom Paris</orgName>
          <date type="start">2019-06-12</date>
          <desc>
            <address>
              <addrLine>19 Place Marguerite Perey 91120 Palaiseau</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">https://www.telecom-paris.fr</ref>
          </desc>
          <listRelation>
            <relation active="#struct-302102" type="direct"/>
            <relation active="#struct-563936" type="direct"/>
          </listRelation>
        </org>
        <org type="regroupinstitution" xml:id="struct-563936" status="VALID">
          <idno type="IdRef">238327159</idno>
          <idno type="ISNI">0000000502717600</idno>
          <idno type="ROR">https://ror.org/042tfbd02</idno>
          <idno type="Wikidata">Q48759778</idno>
          <orgName>Institut Polytechnique de Paris</orgName>
          <orgName type="acronym">IP Paris</orgName>
          <date type="start">2019-06-02</date>
          <desc>
            <address>
              <addrLine>Route de Saclay, 91120 Palaiseau Cedex, France</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">https://www.ip-paris.fr</ref>
          </desc>
        </org>
        <org type="institution" xml:id="struct-300362" status="VALID">
          <idno type="ROR">https://ror.org/01naq7912</idno>
          <orgName>Télécom ParisTech</orgName>
          <date type="start">2008-01-01</date>
          <date type="end">2019-06-11</date>
          <desc>
            <address>
              <addrLine>46 rue Barrault 75634 Paris Cedex 13</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">https://www.telecom-paris.fr</ref>
          </desc>
        </org>
        <org type="institution" xml:id="struct-92966" status="OLD">
          <idno type="ROR">https://ror.org/028rypz17</idno>
          <orgName>Université Paris-Sud - Paris 11</orgName>
          <orgName type="acronym">UP11</orgName>
          <date type="end">2019-12-31</date>
          <desc>
            <address>
              <addrLine>Bâtiment 300 - 91405 Orsay cedex</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">http://www.u-psud.fr/</ref>
          </desc>
        </org>
        <org type="institution" xml:id="struct-411575" status="VALID">
          <idno type="IdRef">184443237</idno>
          <idno type="ROR">https://ror.org/019tcpt25</idno>
          <orgName>CentraleSupélec</orgName>
          <desc>
            <address>
              <addrLine>3, rue Joliot Curie,Plateau de Moulon,91192 GIF-SUR-YVETTE Cedex</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">http://www.centralesupelec.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>
      </listOrg>
    </back>
  </text>
</TEI>