Enhancing Supervised Learning with Contrastive Markings in Neural Machine Translation Training

Supervised learning in Neural Machine Translation (NMT) typically follows a teacher forcing paradigm where reference tokens constitute the conditioning context in the model's prediction, instead of its own previous predictions. In order to alleviate this lack of exploration in the space of tran...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:arXiv.org 2023-07
Hauptverfasser: Berger, Nathaniel, Exel, Miriam, Huck, Matthias, Riezler, Stefan
Format: Artikel
Sprache:eng
Schlagworte:
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 Berger, Nathaniel
Exel, Miriam
Huck, Matthias
Riezler, Stefan
description Supervised learning in Neural Machine Translation (NMT) typically follows a teacher forcing paradigm where reference tokens constitute the conditioning context in the model's prediction, instead of its own previous predictions. In order to alleviate this lack of exploration in the space of translations, we present a simple extension of standard maximum likelihood estimation by a contrastive marking objective. The additional training signals are extracted automatically from reference translations by comparing the system hypothesis against the reference, and used for up/down-weighting correct/incorrect tokens. The proposed new training procedure requires one additional translation pass over the training set per epoch, and does not alter the standard inference setup. We show that training with contrastive markings yields improvements on top of supervised learning, and is especially useful when learning from postedits where contrastive markings indicate human error corrections to the original hypotheses. Code is publicly released.
format Article
fullrecord <record><control><sourceid>proquest</sourceid><recordid>TN_cdi_proquest_journals_2838879898</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2838879898</sourcerecordid><originalsourceid>FETCH-proquest_journals_28388798983</originalsourceid><addsrcrecordid>eNqNjMEKgkAURYcgSMp_GGgt2EzmuBajRbWpdTLYlM_kafNm7PdT6ANa3cs5lztjgZByE6mtEAsWEjVxHItdKpJEBuxWYK2xAnzyi--NHYDMnR-NtjixD7ia5x06q8nBYPhJ29coiAPys_FWtyOqakDDr1YjtdpBh1OH6WDF5g_dkgl_uWTrfXHND1Fvu7c35Mqm8xZHVQollUozlSn53-oLoYNE3A</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2838879898</pqid></control><display><type>article</type><title>Enhancing Supervised Learning with Contrastive Markings in Neural Machine Translation Training</title><source>Free E- Journals</source><creator>Berger, Nathaniel ; Exel, Miriam ; Huck, Matthias ; Riezler, Stefan</creator><creatorcontrib>Berger, Nathaniel ; Exel, Miriam ; Huck, Matthias ; Riezler, Stefan</creatorcontrib><description>Supervised learning in Neural Machine Translation (NMT) typically follows a teacher forcing paradigm where reference tokens constitute the conditioning context in the model's prediction, instead of its own previous predictions. In order to alleviate this lack of exploration in the space of translations, we present a simple extension of standard maximum likelihood estimation by a contrastive marking objective. The additional training signals are extracted automatically from reference translations by comparing the system hypothesis against the reference, and used for up/down-weighting correct/incorrect tokens. The proposed new training procedure requires one additional translation pass over the training set per epoch, and does not alter the standard inference setup. We show that training with contrastive markings yields improvements on top of supervised learning, and is especially useful when learning from postedits where contrastive markings indicate human error corrections to the original hypotheses. Code is publicly released.</description><identifier>EISSN: 2331-8422</identifier><language>eng</language><publisher>Ithaca: Cornell University Library, arXiv.org</publisher><subject>Error correction ; Human error ; Hypotheses ; Machine translation ; Maximum likelihood estimation ; Supervised learning ; Training</subject><ispartof>arXiv.org, 2023-07</ispartof><rights>2023. This work is published under http://creativecommons.org/licenses/by-nc-nd/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>776,780</link.rule.ids></links><search><creatorcontrib>Berger, Nathaniel</creatorcontrib><creatorcontrib>Exel, Miriam</creatorcontrib><creatorcontrib>Huck, Matthias</creatorcontrib><creatorcontrib>Riezler, Stefan</creatorcontrib><title>Enhancing Supervised Learning with Contrastive Markings in Neural Machine Translation Training</title><title>arXiv.org</title><description>Supervised learning in Neural Machine Translation (NMT) typically follows a teacher forcing paradigm where reference tokens constitute the conditioning context in the model's prediction, instead of its own previous predictions. In order to alleviate this lack of exploration in the space of translations, we present a simple extension of standard maximum likelihood estimation by a contrastive marking objective. The additional training signals are extracted automatically from reference translations by comparing the system hypothesis against the reference, and used for up/down-weighting correct/incorrect tokens. The proposed new training procedure requires one additional translation pass over the training set per epoch, and does not alter the standard inference setup. We show that training with contrastive markings yields improvements on top of supervised learning, and is especially useful when learning from postedits where contrastive markings indicate human error corrections to the original hypotheses. Code is publicly released.</description><subject>Error correction</subject><subject>Human error</subject><subject>Hypotheses</subject><subject>Machine translation</subject><subject>Maximum likelihood estimation</subject><subject>Supervised learning</subject><subject>Training</subject><issn>2331-8422</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2023</creationdate><recordtype>article</recordtype><sourceid>BENPR</sourceid><recordid>eNqNjMEKgkAURYcgSMp_GGgt2EzmuBajRbWpdTLYlM_kafNm7PdT6ANa3cs5lztjgZByE6mtEAsWEjVxHItdKpJEBuxWYK2xAnzyi--NHYDMnR-NtjixD7ia5x06q8nBYPhJ29coiAPys_FWtyOqakDDr1YjtdpBh1OH6WDF5g_dkgl_uWTrfXHND1Fvu7c35Mqm8xZHVQollUozlSn53-oLoYNE3A</recordid><startdate>20230717</startdate><enddate>20230717</enddate><creator>Berger, Nathaniel</creator><creator>Exel, Miriam</creator><creator>Huck, Matthias</creator><creator>Riezler, Stefan</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>20230717</creationdate><title>Enhancing Supervised Learning with Contrastive Markings in Neural Machine Translation Training</title><author>Berger, Nathaniel ; Exel, Miriam ; Huck, Matthias ; Riezler, Stefan</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-proquest_journals_28388798983</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2023</creationdate><topic>Error correction</topic><topic>Human error</topic><topic>Hypotheses</topic><topic>Machine translation</topic><topic>Maximum likelihood estimation</topic><topic>Supervised learning</topic><topic>Training</topic><toplevel>online_resources</toplevel><creatorcontrib>Berger, Nathaniel</creatorcontrib><creatorcontrib>Exel, Miriam</creatorcontrib><creatorcontrib>Huck, Matthias</creatorcontrib><creatorcontrib>Riezler, Stefan</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>Publicly Available Content Database</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>Berger, Nathaniel</au><au>Exel, Miriam</au><au>Huck, Matthias</au><au>Riezler, Stefan</au><format>book</format><genre>document</genre><ristype>GEN</ristype><atitle>Enhancing Supervised Learning with Contrastive Markings in Neural Machine Translation Training</atitle><jtitle>arXiv.org</jtitle><date>2023-07-17</date><risdate>2023</risdate><eissn>2331-8422</eissn><abstract>Supervised learning in Neural Machine Translation (NMT) typically follows a teacher forcing paradigm where reference tokens constitute the conditioning context in the model's prediction, instead of its own previous predictions. In order to alleviate this lack of exploration in the space of translations, we present a simple extension of standard maximum likelihood estimation by a contrastive marking objective. The additional training signals are extracted automatically from reference translations by comparing the system hypothesis against the reference, and used for up/down-weighting correct/incorrect tokens. The proposed new training procedure requires one additional translation pass over the training set per epoch, and does not alter the standard inference setup. We show that training with contrastive markings yields improvements on top of supervised learning, and is especially useful when learning from postedits where contrastive markings indicate human error corrections to the original hypotheses. Code is publicly released.</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, 2023-07
issn 2331-8422
language eng
recordid cdi_proquest_journals_2838879898
source Free E- Journals
subjects Error correction
Human error
Hypotheses
Machine translation
Maximum likelihood estimation
Supervised learning
Training
title Enhancing Supervised Learning with Contrastive Markings in Neural Machine Translation Training
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-30T06%3A06%3A44IST&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=Enhancing%20Supervised%20Learning%20with%20Contrastive%20Markings%20in%20Neural%20Machine%20Translation%20Training&rft.jtitle=arXiv.org&rft.au=Berger,%20Nathaniel&rft.date=2023-07-17&rft.eissn=2331-8422&rft_id=info:doi/&rft_dat=%3Cproquest%3E2838879898%3C/proquest%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2838879898&rft_id=info:pmid/&rfr_iscdi=true