Specializing Multi-domain NMT via Penalizing Low Mutual Information

Specializing Multi-domain NMT via Penalizing Low Mutual Information

Multi-domain Neural Machine Translation (NMT) trains a single model with multiple domains. It is appealing because of its efficacy in handling multiple domains within one model. An ideal multi-domain NMT should learn distinctive domain characteristics simultaneously, however, grasping the domain pec...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:arXiv.org 2022-10
Hauptverfasser: Lee, Jiyoung, Kim, Hantae, Cho, Hyunchang, Choi, Edward, Park, Cheonbok
Format: Artikel
Sprache:eng
Schlagworte:
Machine translation
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title arXiv.org
container_volume
creator Lee, Jiyoung
Kim, Hantae
Cho, Hyunchang
Choi, Edward
Park, Cheonbok
description Multi-domain Neural Machine Translation (NMT) trains a single model with multiple domains. It is appealing because of its efficacy in handling multiple domains within one model. An ideal multi-domain NMT should learn distinctive domain characteristics simultaneously, however, grasping the domain peculiarity is a non-trivial task. In this paper, we investigate domain-specific information through the lens of mutual information (MI) and propose a new objective that penalizes low MI to become higher. Our method achieved the state-of-the-art performance among the current competitive multi-domain NMT models. Also, we empirically show our objective promotes low MI to be higher resulting in domain-specialized multi-domain NMT.
format Article
fullrecord <record><control><sourceid>proquest</sourceid><recordid>TN_cdi_proquest_journals_2728313322</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2728313322</sourcerecordid><originalsourceid>FETCH-proquest_journals_27283133223</originalsourceid><addsrcrecordid>eNqNjEELgjAYQEcQJOV_GHQW9Ps0vUtRkBHkXUbN-GRu5raCfn0e6t7pHd7jzVgAiElUpAALFlrbxXEMmxyyDANWXgZ5JaHoTfrOK68cRTfTC9L8VNX8SYKfpf75o3lNjfNC8YNuzdgLR0av2LwVysrwyyVb77Z1uY-G0Ty8tK7pjB-niW0ghwITRAD8r_oA34I57A</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2728313322</pqid></control><display><type>article</type><title>Specializing Multi-domain NMT via Penalizing Low Mutual Information</title><source>Freely Accessible Journals</source><creator>Lee, Jiyoung ; Kim, Hantae ; Cho, Hyunchang ; Choi, Edward ; Park, Cheonbok</creator><creatorcontrib>Lee, Jiyoung ; Kim, Hantae ; Cho, Hyunchang ; Choi, Edward ; Park, Cheonbok</creatorcontrib><description>Multi-domain Neural Machine Translation (NMT) trains a single model with multiple domains. It is appealing because of its efficacy in handling multiple domains within one model. An ideal multi-domain NMT should learn distinctive domain characteristics simultaneously, however, grasping the domain peculiarity is a non-trivial task. In this paper, we investigate domain-specific information through the lens of mutual information (MI) and propose a new objective that penalizes low MI to become higher. Our method achieved the state-of-the-art performance among the current competitive multi-domain NMT models. Also, we empirically show our objective promotes low MI to be higher resulting in domain-specialized multi-domain NMT.</description><identifier>EISSN: 2331-8422</identifier><language>eng</language><publisher>Ithaca: Cornell University Library, arXiv.org</publisher><subject>Machine translation</subject><ispartof>arXiv.org, 2022-10</ispartof><rights>2022. This work is published under http://creativecommons.org/licenses/by/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>781,785</link.rule.ids></links><search><creatorcontrib>Lee, Jiyoung</creatorcontrib><creatorcontrib>Kim, Hantae</creatorcontrib><creatorcontrib>Cho, Hyunchang</creatorcontrib><creatorcontrib>Choi, Edward</creatorcontrib><creatorcontrib>Park, Cheonbok</creatorcontrib><title>Specializing Multi-domain NMT via Penalizing Low Mutual Information</title><title>arXiv.org</title><description>Multi-domain Neural Machine Translation (NMT) trains a single model with multiple domains. It is appealing because of its efficacy in handling multiple domains within one model. An ideal multi-domain NMT should learn distinctive domain characteristics simultaneously, however, grasping the domain peculiarity is a non-trivial task. In this paper, we investigate domain-specific information through the lens of mutual information (MI) and propose a new objective that penalizes low MI to become higher. Our method achieved the state-of-the-art performance among the current competitive multi-domain NMT models. Also, we empirically show our objective promotes low MI to be higher resulting in domain-specialized multi-domain NMT.</description><subject>Machine translation</subject><issn>2331-8422</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2022</creationdate><recordtype>article</recordtype><sourceid>ABUWG</sourceid><sourceid>AFKRA</sourceid><sourceid>AZQEC</sourceid><sourceid>BENPR</sourceid><sourceid>CCPQU</sourceid><sourceid>DWQXO</sourceid><recordid>eNqNjEELgjAYQEcQJOV_GHQW9Ps0vUtRkBHkXUbN-GRu5raCfn0e6t7pHd7jzVgAiElUpAALFlrbxXEMmxyyDANWXgZ5JaHoTfrOK68cRTfTC9L8VNX8SYKfpf75o3lNjfNC8YNuzdgLR0av2LwVysrwyyVb77Z1uY-G0Ty8tK7pjB-niW0ghwITRAD8r_oA34I57A</recordid><startdate>20221024</startdate><enddate>20221024</enddate><creator>Lee, Jiyoung</creator><creator>Kim, Hantae</creator><creator>Cho, Hyunchang</creator><creator>Choi, Edward</creator><creator>Park, Cheonbok</creator><general>Cornell University Library, arXiv.org</general><scope>8FE</scope><scope>8FG</scope><scope>ABJCF</scope><scope>ABUWG</scope><scope>AFKRA</scope><scope>AZQEC</scope><scope>BENPR</scope><scope>BGLVJ</scope><scope>CCPQU</scope><scope>DWQXO</scope><scope>HCIFZ</scope><scope>L6V</scope><scope>M7S</scope><scope>PIMPY</scope><scope>PQEST</scope><scope>PQQKQ</scope><scope>PQUKI</scope><scope>PRINS</scope><scope>PTHSS</scope></search><sort><creationdate>20221024</creationdate><title>Specializing Multi-domain NMT via Penalizing Low Mutual Information</title><author>Lee, Jiyoung ; Kim, Hantae ; Cho, Hyunchang ; Choi, Edward ; Park, Cheonbok</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-proquest_journals_27283133223</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2022</creationdate><topic>Machine translation</topic><toplevel>online_resources</toplevel><creatorcontrib>Lee, Jiyoung</creatorcontrib><creatorcontrib>Kim, Hantae</creatorcontrib><creatorcontrib>Cho, Hyunchang</creatorcontrib><creatorcontrib>Choi, Edward</creatorcontrib><creatorcontrib>Park, Cheonbok</creatorcontrib><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>Materials Science &amp; Engineering Collection</collection><collection>ProQuest Central (Alumni Edition)</collection><collection>ProQuest Central UK/Ireland</collection><collection>ProQuest Central Essentials</collection><collection>Proquest Central</collection><collection>Technology Collection</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central Korea</collection><collection>SciTech Premium Collection</collection><collection>ProQuest Engineering Collection</collection><collection>Engineering Database</collection><collection>Access via ProQuest (Open Access)</collection><collection>ProQuest One Academic Eastern Edition (DO NOT USE)</collection><collection>ProQuest One Academic</collection><collection>ProQuest One Academic UKI Edition</collection><collection>ProQuest Central China</collection><collection>Engineering Collection</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Lee, Jiyoung</au><au>Kim, Hantae</au><au>Cho, Hyunchang</au><au>Choi, Edward</au><au>Park, Cheonbok</au><format>book</format><genre>document</genre><ristype>GEN</ristype><atitle>Specializing Multi-domain NMT via Penalizing Low Mutual Information</atitle><jtitle>arXiv.org</jtitle><date>2022-10-24</date><risdate>2022</risdate><eissn>2331-8422</eissn><abstract>Multi-domain Neural Machine Translation (NMT) trains a single model with multiple domains. It is appealing because of its efficacy in handling multiple domains within one model. An ideal multi-domain NMT should learn distinctive domain characteristics simultaneously, however, grasping the domain peculiarity is a non-trivial task. In this paper, we investigate domain-specific information through the lens of mutual information (MI) and propose a new objective that penalizes low MI to become higher. Our method achieved the state-of-the-art performance among the current competitive multi-domain NMT models. Also, we empirically show our objective promotes low MI to be higher resulting in domain-specialized multi-domain NMT.</abstract><cop>Ithaca</cop><pub>Cornell University Library, arXiv.org</pub><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier EISSN: 2331-8422
ispartof arXiv.org, 2022-10
issn 2331-8422
language eng
recordid cdi_proquest_journals_2728313322
source Freely Accessible Journals
subjects Machine translation
title Specializing Multi-domain NMT via Penalizing Low Mutual Information
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-15T00%3A49%3A24IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=document&rft.atitle=Specializing%20Multi-domain%20NMT%20via%20Penalizing%20Low%20Mutual%20Information&rft.jtitle=arXiv.org&rft.au=Lee,%20Jiyoung&rft.date=2022-10-24&rft.eissn=2331-8422&rft_id=info:doi/&rft_dat=%3Cproquest%3E2728313322%3C/proquest%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2728313322&rft_id=info:pmid/&rfr_iscdi=true