<?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-04591889</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="2026-05-02T15:51:57+02:00"/>
      </publicationStmt>
      <sourceDesc>
        <p part="N">HAL API Platform</p>
      </sourceDesc>
    </fileDesc>
  </teiHeader>
  <text>
    <body>
      <listBibl>
        <biblFull>
          <titleStmt>
            <title xml:lang="en">Exploring Inline Lexicon Injection for Cross-Domain Transfer in Neural Machine Translation</title>
            <author role="aut">
              <persName>
                <forename type="first">Jesujoba</forename>
                <forename type="middle">O</forename>
                <surname>Alabi</surname>
              </persName>
              <email type="md5">eaa5e9aa18933b3de486a71f0a9a5cc8</email>
              <email type="domain">lsv.uni-saarland.de</email>
              <idno type="idhal" notation="numeric">1180811</idno>
              <idno type="halauthorid" notation="string">2288298-1180811</idno>
              <affiliation ref="#struct-92733"/>
              <affiliation ref="#struct-1144334"/>
            </author>
            <author role="aut">
              <persName>
                <forename type="first">Rachel</forename>
                <surname>Bawden</surname>
              </persName>
              <email type="md5">a08b8a5ca6961e673dd659fdfe33c881</email>
              <email type="domain">inria.fr</email>
              <idno type="idhal" notation="string">rachel-bawden</idno>
              <idno type="idhal" notation="numeric">9441</idno>
              <idno type="halauthorid" notation="string">28540-9441</idno>
              <idno type="ORCID">https://orcid.org/0000-0001-9553-1768</idno>
              <idno type="IDREF">https://www.idref.fr/233174591</idno>
              <idno type="ARXIV">https://arxiv.org/a/bawden_r_1</idno>
              <affiliation ref="#struct-482775"/>
            </author>
            <editor role="depositor">
              <persName>
                <forename>Rachel</forename>
                <surname>Bawden</surname>
              </persName>
              <email type="md5">a08b8a5ca6961e673dd659fdfe33c881</email>
              <email type="domain">inria.fr</email>
            </editor>
            <funder ref="#projanr-50388"/>
          </titleStmt>
          <editionStmt>
            <edition n="v1" type="current">
              <date type="whenSubmitted">2024-05-29 10:16:16</date>
              <date type="whenModified">2025-09-12 17:04:02</date>
              <date type="whenReleased">2024-05-30 08:14:42</date>
              <date type="whenProduced">2024-06-27</date>
              <date type="whenEndEmbargoed">2024-05-29</date>
              <ref type="file" target="https://hal.science/hal-04591889v1/document">
                <date notBefore="2024-05-29"/>
              </ref>
              <ref type="file" subtype="author" n="1" target="https://hal.science/hal-04591889v1/file/DadaMT.pdf" id="file-4591889-3984266">
                <date notBefore="2024-05-29"/>
              </ref>
            </edition>
            <respStmt>
              <resp>contributor</resp>
              <name key="197025">
                <persName>
                  <forename>Rachel</forename>
                  <surname>Bawden</surname>
                </persName>
                <email type="md5">a08b8a5ca6961e673dd659fdfe33c881</email>
                <email type="domain">inria.fr</email>
              </name>
            </respStmt>
          </editionStmt>
          <publicationStmt>
            <distributor>CCSD</distributor>
            <idno type="halId">hal-04591889</idno>
            <idno type="halUri">https://hal.science/hal-04591889</idno>
            <idno type="halBibtex">alabi:hal-04591889</idno>
            <idno type="halRefHtml">&lt;i&gt;KEMT 2024 - First International Workshop on Knowledge-Enhanced Machine Translation&lt;/i&gt;, Jun 2024, Sheffield, United Kingdom</idno>
            <idno type="halRef">KEMT 2024 - First International Workshop on Knowledge-Enhanced Machine Translation, Jun 2024, Sheffield, United Kingdom</idno>
            <availability status="restricted">
              <licence target="https://about.hal.science/hal-authorisation-v1/">HAL Authorization<ref corresp="#file-4591889-3984266"/></licence>
            </availability>
          </publicationStmt>
          <seriesStmt>
            <idno type="stamp" n="INRIA">INRIA - Institut National de Recherche en Informatique et en Automatique</idno>
            <idno type="stamp" n="INRIA-ROCQ">INRIA Paris - Rocquencourt</idno>
            <idno type="stamp" n="TESTALAIN1">TESTALAIN1</idno>
            <idno type="stamp" n="INRIA2">INRIA 2</idno>
            <idno type="stamp" n="PNRIA">Programme National de Recherche en IA</idno>
            <idno type="stamp" n="ANR">ANR</idno>
            <idno type="stamp" n="PRAIRIE-IA">PaRis AI Research InstitutE</idno>
            <idno type="stamp" n="INRIA-ALLEMAGNE">INRIA-ALLEMAGNE</idno>
            <idno type="stamp" n="ANR-IA-19" corresp="ANR-IA">ANR-IA-19</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">Exploring Inline Lexicon Injection for Cross-Domain Transfer in Neural Machine Translation</title>
                <author role="aut">
                  <persName>
                    <forename type="first">Jesujoba</forename>
                    <forename type="middle">O</forename>
                    <surname>Alabi</surname>
                  </persName>
                  <email type="md5">eaa5e9aa18933b3de486a71f0a9a5cc8</email>
                  <email type="domain">lsv.uni-saarland.de</email>
                  <idno type="idhal" notation="numeric">1180811</idno>
                  <idno type="halauthorid" notation="string">2288298-1180811</idno>
                  <affiliation ref="#struct-92733"/>
                  <affiliation ref="#struct-1144334"/>
                </author>
                <author role="aut">
                  <persName>
                    <forename type="first">Rachel</forename>
                    <surname>Bawden</surname>
                  </persName>
                  <email type="md5">a08b8a5ca6961e673dd659fdfe33c881</email>
                  <email type="domain">inria.fr</email>
                  <idno type="idhal" notation="string">rachel-bawden</idno>
                  <idno type="idhal" notation="numeric">9441</idno>
                  <idno type="halauthorid" notation="string">28540-9441</idno>
                  <idno type="ORCID">https://orcid.org/0000-0001-9553-1768</idno>
                  <idno type="IDREF">https://www.idref.fr/233174591</idno>
                  <idno type="ARXIV">https://arxiv.org/a/bawden_r_1</idno>
                  <affiliation ref="#struct-482775"/>
                </author>
              </analytic>
              <monogr>
                <meeting>
                  <title>KEMT 2024 - First International Workshop on Knowledge-Enhanced Machine Translation</title>
                  <date type="start">2024-06-27</date>
                  <settlement>Sheffield</settlement>
                  <country key="GB">United Kingdom</country>
                </meeting>
                <imprint>
                  <biblScope unit="serie">Proceedings of the First International Workshop on Knowledge-Enhanced Machine Translation</biblScope>
                </imprint>
              </monogr>
            </biblStruct>
          </sourceDesc>
          <profileDesc>
            <langUsage>
              <language ident="en">English</language>
            </langUsage>
            <textClass>
              <keywords scheme="author">
                <term xml:lang="en">Machine translation</term>
                <term xml:lang="en">Large language models</term>
                <term xml:lang="en">Lexicon</term>
                <term xml:lang="en">Evaluation</term>
              </keywords>
              <classCode scheme="halDomain" n="info.info-cl">Computer Science [cs]/Computation and Language [cs.CL]</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>Domain transfer remains a challenge in machine translation (MT), particularly concerning rare or unseen words. Amongst the strategies proposed to address the issue, one of the simplest and most promising in terms of generalisation capacity is coupling the MT system with external resources such as bilingual lexicons and appending inline annotations within source sentences. This method has been shown to work well for controlled language settings, but its usability for general language (and ambiguous) MT is less certain. In this article we explore this question further, testing the strategy in a multi-domain transfer setting for German-to-English MT, using the mT5 language model fine-tuned on parallel data. We analyse the MT outputs and design evaluation strategies to understand the behaviour of such models. Our analysis using distractor annotations suggests that although improvements are not systematic according to automatic metrics, the model does learn to select appropriate translation candidates and ignore irrelevant ones, thereby exhibiting more than a systematic copying behaviour. However, we also find that the method is less successful in a higher-resource setting with a larger lexicon, suggesting that it is not a magic solution, especially when the baseline model is already exposed to a wide range of vocabulary.</p>
            </abstract>
          </profileDesc>
        </biblFull>
      </listBibl>
    </body>
    <back>
      <listOrg type="structures">
        <org type="institution" xml:id="struct-92733" status="VALID">
          <idno type="ROR">https://ror.org/01jdpyv68</idno>
          <orgName>Universität des Saarlandes [Saarbrücken] = Saarland University [Saarbrücken]</orgName>
          <desc>
            <address>
              <addrLine>Universität des Saarlandes Campus D-66123 Saarbrücken</addrLine>
              <country key="DE"/>
            </address>
            <ref type="url">http://www.uni-saarland.de</ref>
          </desc>
        </org>
        <org type="institution" xml:id="struct-1144334" status="VALID">
          <orgName>Spoken Language Systems</orgName>
          <orgName type="acronym">LSV</orgName>
          <desc>
            <address>
              <addrLine>Universität des SaarlandesSaarland Informatics CampusSprach- und Signalverarbeitung (LSV)Gebäude C7 166123 Saarbrücken</addrLine>
              <country key="DE"/>
            </address>
          </desc>
        </org>
        <org type="researchteam" xml:id="struct-482775" status="VALID">
          <idno type="RNSR">201722248N</idno>
          <idno type="ROR">https://ror.org/05rnwy495</idno>
          <orgName>Automatic Language Modelling and ANAlysis &amp; Computational Humanities</orgName>
          <orgName type="acronym">ALMAnaCH</orgName>
          <date type="start">2017-01-01</date>
          <date type="end">2029-04-30</date>
          <desc>
            <address>
              <addrLine>48 Rue Barrault, 75013 Paris</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">https://www.inria.fr/equipes/almanach</ref>
          </desc>
          <listRelation>
            <relation active="#struct-454310" type="direct"/>
            <relation active="#struct-300009" type="indirect"/>
          </listRelation>
        </org>
        <org type="laboratory" xml:id="struct-454310" status="VALID">
          <idno type="IdRef">241614864</idno>
          <idno type="RNSR">196718247G</idno>
          <idno type="ROR">https://ror.org/05eyd5d35</idno>
          <orgName>Centre Inria de Paris</orgName>
          <date type="start">2016-03-10</date>
          <desc>
            <address>
              <addrLine>48 Rue Barrault, 75013 Paris</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">http://www.inria.fr/centre/paris</ref>
          </desc>
          <listRelation>
            <relation active="#struct-300009" type="direct"/>
          </listRelation>
        </org>
        <org type="institution" xml:id="struct-300009" status="VALID">
          <idno type="ROR">https://ror.org/02kvxyf05</idno>
          <orgName>Institut National de Recherche en Informatique et en Automatique</orgName>
          <orgName type="acronym">Inria</orgName>
          <desc>
            <address>
              <addrLine>Domaine de VoluceauRocquencourt - BP 10578153 Le Chesnay Cedex</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">http://www.inria.fr/en/</ref>
          </desc>
        </org>
      </listOrg>
      <listOrg type="projects">
        <org type="anrProject" xml:id="projanr-50388" status="VALID">
          <idno type="anr">ANR-19-P3IA-0001</idno>
          <orgName>PRAIRIE</orgName>
          <desc>PaRis Artificial Intelligence Research InstitutE</desc>
          <date type="start">2019</date>
        </org>
      </listOrg>
    </back>
  </text>
</TEI>