<?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-05214704</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="2026-05-03T11:19:39+02:00"/>
      </publicationStmt>
      <sourceDesc>
        <p part="N">HAL API Platform</p>
      </sourceDesc>
    </fileDesc>
  </teiHeader>
  <text>
    <body>
      <listBibl>
        <biblFull>
          <titleStmt>
            <title xml:lang="en">In the LLM era, Word Sense Induction remains unsolved</title>
            <author role="aut">
              <persName>
                <forename type="first">Anna</forename>
                <surname>Mosolova</surname>
              </persName>
              <email type="md5">4e87330a0dab966051ef53e30930f1aa</email>
              <email type="domain">gmail.com</email>
              <idno type="idhal" notation="string">anna-mosolova</idno>
              <idno type="idhal" notation="numeric">1163367</idno>
              <idno type="halauthorid" notation="string">3189451-1163367</idno>
              <affiliation ref="#struct-1004855"/>
            </author>
            <author role="aut">
              <persName>
                <forename type="first">Marie</forename>
                <surname>Candito</surname>
              </persName>
              <email type="md5">7ecb2efd6169b6577eee989107ccad3d</email>
              <email type="domain">gmail.com</email>
              <idno type="idhal" notation="string">marie-candito</idno>
              <idno type="idhal" notation="numeric">13596</idno>
              <idno type="halauthorid" notation="string">14130-13596</idno>
              <idno type="IDREF">https://www.idref.fr/153698616</idno>
              <idno type="ORCID">https://orcid.org/0000-0001-8306-4859</idno>
              <affiliation ref="#struct-1004855"/>
            </author>
            <author role="aut">
              <persName>
                <forename type="first">Carlos</forename>
                <surname>Ramisch</surname>
              </persName>
              <email type="md5">73383d7126291352e96e0e82b6f5dfaf</email>
              <email type="domain">lis-lab.fr</email>
              <idno type="idhal" notation="string">carlos-ramisch</idno>
              <idno type="idhal" notation="numeric">5103</idno>
              <idno type="halauthorid" notation="string">15660-5103</idno>
              <idno type="ORCID">https://orcid.org/0000-0001-7466-9039</idno>
              <idno type="IDREF">https://www.idref.fr/170720802</idno>
              <affiliation ref="#struct-530703"/>
            </author>
            <editor role="depositor">
              <persName>
                <forename>Marie</forename>
                <surname>Candito</surname>
              </persName>
              <email type="md5">7ecb2efd6169b6577eee989107ccad3d</email>
              <email type="domain">gmail.com</email>
            </editor>
            <funder ref="#projanr-57264"/>
          </titleStmt>
          <editionStmt>
            <edition n="v1" type="current">
              <date type="whenSubmitted">2025-08-19 15:15:06</date>
              <date type="whenModified">2026-04-22 09:38:23</date>
              <date type="whenReleased">2025-08-19 15:51:51</date>
              <date type="whenProduced">2025-07-27</date>
              <date type="whenEndEmbargoed">2025-08-19</date>
              <ref type="file" target="https://hal.science/hal-05214704v1/document">
                <date notBefore="2025-08-19"/>
              </ref>
              <ref type="file" subtype="greenPublisher" n="1" target="https://hal.science/hal-05214704v1/file/2025.findings-acl.882.pdf" id="file-5214704-4487847">
                <date notBefore="2025-08-19"/>
              </ref>
            </edition>
            <respStmt>
              <resp>contributor</resp>
              <name key="134260">
                <persName>
                  <forename>Marie</forename>
                  <surname>Candito</surname>
                </persName>
                <email type="md5">7ecb2efd6169b6577eee989107ccad3d</email>
                <email type="domain">gmail.com</email>
              </name>
            </respStmt>
          </editionStmt>
          <publicationStmt>
            <distributor>CCSD</distributor>
            <idno type="halId">hal-05214704</idno>
            <idno type="halUri">https://hal.science/hal-05214704</idno>
            <idno type="halBibtex">mosolova:hal-05214704</idno>
            <idno type="halRefHtml">&lt;i&gt;Findings of the Association for Computational Linguistics: ACL 2025&lt;/i&gt;, Jul 2025, Vienne, Austria. pp.17161-17178, &lt;a target="_blank" href="https://dx.doi.org/10.18653/v1/2025.findings-acl.882"&gt;&amp;#x27E8;10.18653/v1/2025.findings-acl.882&amp;#x27E9;&lt;/a&gt;</idno>
            <idno type="halRef">Findings of the Association for Computational Linguistics: ACL 2025, Jul 2025, Vienne, Austria. pp.17161-17178, &amp;#x27E8;10.18653/v1/2025.findings-acl.882&amp;#x27E9;</idno>
            <availability status="restricted">
              <licence target="https://creativecommons.org/licenses/by/4.0/">CC BY 4.0 - Attribution<ref corresp="#file-5214704-4487847"/></licence>
            </availability>
          </publicationStmt>
          <seriesStmt>
            <idno type="stamp" n="SHS">Sciences de l'Homme et de la Société</idno>
            <idno type="stamp" n="UNIV-TLN">Université de Toulon</idno>
            <idno type="stamp" n="CNRS">CNRS - Centre national de la recherche scientifique</idno>
            <idno type="stamp" n="UNIV-AMU">Aix Marseille Université</idno>
            <idno type="stamp" n="LLF">Laboratoire de Linguistique Formelle</idno>
            <idno type="stamp" n="AO-LINGUISTIQUE">Archives ouvertes de la Linguistique</idno>
            <idno type="stamp" n="CAMPUS-AAR">Campus-AAR</idno>
            <idno type="stamp" n="AAI">Collection LaCAS - Open Archives in Language and Cultural Area Studies</idno>
            <idno type="stamp" n="LIS-LAB">Laboratoire d'Informatique et Systèmes</idno>
            <idno type="stamp" n="UNIV-PARIS">Université Paris Cité</idno>
            <idno type="stamp" n="UNIVERSITE-PARIS" corresp="UNIV-PARIS">Université Paris Cité</idno>
            <idno type="stamp" n="UP-SOCIETES-HUMANITES">Université Paris Cité - Faculté Sociétés et Humanités</idno>
            <idno type="stamp" n="ANR">ANR</idno>
            <idno type="stamp" n="ANR-IA-21" corresp="ANR-IA">ANR-IA-21</idno>
            <idno type="stamp" n="ANR-IA">ANR-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">In the LLM era, Word Sense Induction remains unsolved</title>
                <author role="aut">
                  <persName>
                    <forename type="first">Anna</forename>
                    <surname>Mosolova</surname>
                  </persName>
                  <email type="md5">4e87330a0dab966051ef53e30930f1aa</email>
                  <email type="domain">gmail.com</email>
                  <idno type="idhal" notation="string">anna-mosolova</idno>
                  <idno type="idhal" notation="numeric">1163367</idno>
                  <idno type="halauthorid" notation="string">3189451-1163367</idno>
                  <affiliation ref="#struct-1004855"/>
                </author>
                <author role="aut">
                  <persName>
                    <forename type="first">Marie</forename>
                    <surname>Candito</surname>
                  </persName>
                  <email type="md5">7ecb2efd6169b6577eee989107ccad3d</email>
                  <email type="domain">gmail.com</email>
                  <idno type="idhal" notation="string">marie-candito</idno>
                  <idno type="idhal" notation="numeric">13596</idno>
                  <idno type="halauthorid" notation="string">14130-13596</idno>
                  <idno type="IDREF">https://www.idref.fr/153698616</idno>
                  <idno type="ORCID">https://orcid.org/0000-0001-8306-4859</idno>
                  <affiliation ref="#struct-1004855"/>
                </author>
                <author role="aut">
                  <persName>
                    <forename type="first">Carlos</forename>
                    <surname>Ramisch</surname>
                  </persName>
                  <email type="md5">73383d7126291352e96e0e82b6f5dfaf</email>
                  <email type="domain">lis-lab.fr</email>
                  <idno type="idhal" notation="string">carlos-ramisch</idno>
                  <idno type="idhal" notation="numeric">5103</idno>
                  <idno type="halauthorid" notation="string">15660-5103</idno>
                  <idno type="ORCID">https://orcid.org/0000-0001-7466-9039</idno>
                  <idno type="IDREF">https://www.idref.fr/170720802</idno>
                  <affiliation ref="#struct-530703"/>
                </author>
              </analytic>
              <monogr>
                <meeting>
                  <title>Findings of the Association for Computational Linguistics: ACL 2025</title>
                  <date type="start">2025-07-27</date>
                  <date type="end">2025-08-01</date>
                  <settlement>Vienne</settlement>
                  <country key="AT">Austria</country>
                </meeting>
                <imprint>
                  <publisher>Association for Computational Linguistics</publisher>
                  <biblScope unit="serie">Findings of the Association for Computational Linguistics: ACL 2025</biblScope>
                  <biblScope unit="pp">17161-17178</biblScope>
                </imprint>
              </monogr>
              <idno type="doi">10.18653/v1/2025.findings-acl.882</idno>
              <ref type="publisher">https://2025.aclweb.org/</ref>
            </biblStruct>
          </sourceDesc>
          <profileDesc>
            <langUsage>
              <language ident="en">English</language>
            </langUsage>
            <textClass>
              <keywords scheme="author">
                <term xml:lang="en">Clustering</term>
                <term xml:lang="en">Natural language processing</term>
                <term xml:lang="en">Unsupervised machine learning</term>
                <term xml:lang="en">Word sense induction</term>
              </keywords>
              <classCode scheme="halDomain" n="info.info-cl">Computer Science [cs]/Computation and Language [cs.CL]</classCode>
              <classCode scheme="halDomain" n="shs.langue">Humanities and Social Sciences/Linguistics</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>In the absence of sense-annotated data, word sense induction (WSI) is a compelling alternative to word sense disambiguation, particularly in low-resource or domain-specific settings. In this paper, we emphasize methodological problems in current WSI evaluation. We propose an evaluation on a SemCor-derived dataset, respecting the original corpus polysemy and frequency distributions. We assess pre-trained embeddings and clustering algorithms across parts of speech, and propose and evaluate an LLMbased WSI method for English. We evaluate data augmentation sources (LLM-generated, corpus and lexicon), and semi-supervised scenarios using Wiktionary for data augmentation, must-link constraints, number of clusters per lemma.We find that no unsupervised method (whether ours or previous) surpasses the strong "one cluster per lemma" heuristic (1cpl). We also show that (i) results and best systems may vary across POS, (ii) LLMs have troubles performing this task, (iii) data augmentation is beneficial and (iv) capitalizing on Wiktionary does help. It surpasses previous SOTA system on our test set by 3.3%. WSI is not solved, and calls for a better articulation of lexicons and LLMs' lexical semantics capabilities.</p>
            </abstract>
          </profileDesc>
        </biblFull>
      </listBibl>
    </body>
    <back>
      <listOrg type="structures">
        <org type="laboratory" xml:id="struct-1004855" status="VALID">
          <idno type="IdRef">026390574</idno>
          <idno type="RNSR">200112497J</idno>
          <idno type="ROR">https://ror.org/0223bz716</idno>
          <orgName>Laboratoire de Linguistique Formelle</orgName>
          <orgName type="acronym">LLF - UMR7110</orgName>
          <date type="start">2020-01-01</date>
          <desc>
            <address>
              <addrLine>8 Rue Albert Einstein, 75013 Paris</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">http://www.llf.cnrs.fr/</ref>
          </desc>
          <listRelation>
            <relation name="UMR7110" active="#struct-441569" type="direct"/>
            <relation name="UMR_7110" active="#struct-557826" type="direct"/>
          </listRelation>
        </org>
        <org type="researchteam" xml:id="struct-530703" status="VALID">
          <orgName>Traitement Automatique du Langage Ecrit et Parlé</orgName>
          <orgName type="acronym">TALEP</orgName>
          <desc>
            <address>
              <country key="FR"/>
            </address>
            <ref type="url">http://www.lis-lab.fr/talep</ref>
          </desc>
          <listRelation>
            <relation active="#struct-527033" type="direct"/>
            <relation name="UMR7020" active="#struct-198056" type="indirect"/>
            <relation name="UMR7020" active="#struct-303091" type="indirect"/>
            <relation name="UMR7020" active="#struct-441569" type="indirect"/>
          </listRelation>
        </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="institution" xml:id="struct-557826" status="VALID">
          <idno type="IdRef">236453505</idno>
          <idno type="ISNI">0000 0004 7885 7602</idno>
          <idno type="ROR">https://ror.org/05f82e368</idno>
          <orgName>Université Paris Cité</orgName>
          <orgName type="acronym">UPCité</orgName>
          <date type="start">2020-01-01</date>
          <desc>
            <address>
              <addrLine>85 boulevard Saint-Germain75006 Paris</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">https://u-paris.fr/</ref>
          </desc>
        </org>
        <org type="laboratory" xml:id="struct-527033" status="VALID">
          <idno type="IdRef">234768002</idno>
          <idno type="RNSR">201822714R</idno>
          <idno type="ROR">https://ror.org/0257sgk90</idno>
          <idno type="Wikidata">Q52605252</idno>
          <orgName>Laboratoire d'Informatique et des Systèmes (LIS) (Marseille, Toulon)</orgName>
          <orgName type="acronym">LIS</orgName>
          <date type="start">2018-01-01</date>
          <desc>
            <address>
              <addrLine>Aix Marseille Université – Campus de Saint Jérôme – Bat. Polytech, 52 Av. Escadrille Normandie Niemen, 13397 Marseille Cedex 20</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">http://www.lis-lab.fr</ref>
          </desc>
          <listRelation>
            <relation name="UMR7020" active="#struct-198056" type="direct"/>
            <relation name="UMR7020" active="#struct-303091" type="direct"/>
            <relation name="UMR7020" active="#struct-441569" type="direct"/>
          </listRelation>
        </org>
        <org type="institution" xml:id="struct-198056" status="VALID">
          <idno type="IdRef">15863621X</idno>
          <idno type="ISNI">0000 0001 2176 4817</idno>
          <idno type="ROR">https://ror.org/035xkbk20</idno>
          <orgName>Aix Marseille Université</orgName>
          <orgName type="acronym">AMU</orgName>
          <date type="start">2012-01-01</date>
          <desc>
            <address>
              <addrLine>Aix-Marseille UniversitéJardins du Pharo58 Boulevard Charles Livon13284 Marseille cedex 7</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">http://www.univ-amu.fr/</ref>
          </desc>
        </org>
        <org type="institution" xml:id="struct-303091" status="VALID">
          <idno type="IdRef">031122337</idno>
          <idno type="ROR">https://ror.org/02m9kbe37</idno>
          <orgName>Université de Toulon</orgName>
          <orgName type="acronym">UTLN</orgName>
          <desc>
            <address>
              <addrLine>Avenue de l'Université - BP 20132 - 83957 La Garde Cedex</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">http://www.univ-tln.fr/</ref>
          </desc>
        </org>
      </listOrg>
      <listOrg type="projects">
        <org type="anrProject" xml:id="projanr-57264" status="VALID">
          <idno type="anr">ANR-21-CE23-0033</idno>
          <orgName>SELEXINI</orgName>
          <desc>Induction de lexiques sémantiques pour l'interprétabilité et la diversité en traitement de textes</desc>
          <date type="start">2021</date>
        </org>
      </listOrg>
    </back>
  </text>
</TEI>