Multilingual Universal Sentence Encoder for Semantic Retrieval

Multilingual Universal Sentence Encoder for Semantic Retrieval

We introduce two pre-trained retrieval focused multilingual sentence encoding models, respectively based on the Transformer and CNN model architectures. The models embed text from 16 languages into a single semantic space using a multi-task trained dual-encoder that learns tied representations using...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Yang, Yinfei, Cer, Daniel, Ahmad, Amin, Guo, Mandy, Law, Jax, Constant, Noah, Abrego, Gustavo Hernandez, Yuan, Steve, Tar, Chris, Sung, Yun-Hsuan, Strope, Brian, Kurzweil, Ray
Format: Artikel
Sprache:eng
Schlagworte:
Computer Science - Computation and Language
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title
container_volume
creator Yang, Yinfei
Cer, Daniel
Ahmad, Amin
Guo, Mandy
Law, Jax
Constant, Noah
Abrego, Gustavo Hernandez
Yuan, Steve
Tar, Chris
Sung, Yun-Hsuan
Strope, Brian
Kurzweil, Ray
description We introduce two pre-trained retrieval focused multilingual sentence encoding models, respectively based on the Transformer and CNN model architectures. The models embed text from 16 languages into a single semantic space using a multi-task trained dual-encoder that learns tied representations using translation based bridge tasks (Chidambaram al., 2018). The models provide performance that is competitive with the state-of-the-art on: semantic retrieval (SR), translation pair bitext retrieval (BR) and retrieval question answering (ReQA). On English transfer learning tasks, our sentence-level embeddings approach, and in some cases exceed, the performance of monolingual, English only, sentence embedding models. Our models are made available for download on TensorFlow Hub.
doi_str_mv 10.48550/arxiv.1907.04307
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_1907_04307</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>1907_04307</sourcerecordid><originalsourceid>FETCH-LOGICAL-a1157-adb3501f8a12d305cd26642cc4ad7c903a52be23305aa63ea276bdce6e5b27f93</originalsourceid><addsrcrecordid>eNotj82KwjAUhbNxMagP4Gr6Aq35aRK7GRBRR1AEna7LbXI7BGKUWIu-_VSd1TmcDw58hEwYzfKZlHQK8e66jBVUZzQXVH-Qr93Nt8678HsDn5TBdRivfTtiaDEYTJbBnC3GpDnHfjxBaJ1JDthGhx34ERk04K84_s8hKVfLn8V3ut2vN4v5NgXGpE7B1kJS1syAcSuoNJYrlXNjcrDaFFSA5DVy0SMAJRC4VrU1qFDWXDeFGJLP9-_LoLpEd4L4qJ4m1ctE_AH_qkPv</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Multilingual Universal Sentence Encoder for Semantic Retrieval</title><source>arXiv.org</source><creator>Yang, Yinfei ; Cer, Daniel ; Ahmad, Amin ; Guo, Mandy ; Law, Jax ; Constant, Noah ; Abrego, Gustavo Hernandez ; Yuan, Steve ; Tar, Chris ; Sung, Yun-Hsuan ; Strope, Brian ; Kurzweil, Ray</creator><creatorcontrib>Yang, Yinfei ; Cer, Daniel ; Ahmad, Amin ; Guo, Mandy ; Law, Jax ; Constant, Noah ; Abrego, Gustavo Hernandez ; Yuan, Steve ; Tar, Chris ; Sung, Yun-Hsuan ; Strope, Brian ; Kurzweil, Ray</creatorcontrib><description>We introduce two pre-trained retrieval focused multilingual sentence encoding models, respectively based on the Transformer and CNN model architectures. The models embed text from 16 languages into a single semantic space using a multi-task trained dual-encoder that learns tied representations using translation based bridge tasks (Chidambaram al., 2018). The models provide performance that is competitive with the state-of-the-art on: semantic retrieval (SR), translation pair bitext retrieval (BR) and retrieval question answering (ReQA). On English transfer learning tasks, our sentence-level embeddings approach, and in some cases exceed, the performance of monolingual, English only, sentence embedding models. Our models are made available for download on TensorFlow Hub.</description><identifier>DOI: 10.48550/arxiv.1907.04307</identifier><language>eng</language><subject>Computer Science - Computation and Language</subject><creationdate>2019-07</creationdate><rights>http://arxiv.org/licenses/nonexclusive-distrib/1.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-a1157-adb3501f8a12d305cd26642cc4ad7c903a52be23305aa63ea276bdce6e5b27f93</citedby></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,776,881</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/1907.04307$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.1907.04307$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Yang, Yinfei</creatorcontrib><creatorcontrib>Cer, Daniel</creatorcontrib><creatorcontrib>Ahmad, Amin</creatorcontrib><creatorcontrib>Guo, Mandy</creatorcontrib><creatorcontrib>Law, Jax</creatorcontrib><creatorcontrib>Constant, Noah</creatorcontrib><creatorcontrib>Abrego, Gustavo Hernandez</creatorcontrib><creatorcontrib>Yuan, Steve</creatorcontrib><creatorcontrib>Tar, Chris</creatorcontrib><creatorcontrib>Sung, Yun-Hsuan</creatorcontrib><creatorcontrib>Strope, Brian</creatorcontrib><creatorcontrib>Kurzweil, Ray</creatorcontrib><title>Multilingual Universal Sentence Encoder for Semantic Retrieval</title><description>We introduce two pre-trained retrieval focused multilingual sentence encoding models, respectively based on the Transformer and CNN model architectures. The models embed text from 16 languages into a single semantic space using a multi-task trained dual-encoder that learns tied representations using translation based bridge tasks (Chidambaram al., 2018). The models provide performance that is competitive with the state-of-the-art on: semantic retrieval (SR), translation pair bitext retrieval (BR) and retrieval question answering (ReQA). On English transfer learning tasks, our sentence-level embeddings approach, and in some cases exceed, the performance of monolingual, English only, sentence embedding models. Our models are made available for download on TensorFlow Hub.</description><subject>Computer Science - Computation and Language</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2019</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotj82KwjAUhbNxMagP4Gr6Aq35aRK7GRBRR1AEna7LbXI7BGKUWIu-_VSd1TmcDw58hEwYzfKZlHQK8e66jBVUZzQXVH-Qr93Nt8678HsDn5TBdRivfTtiaDEYTJbBnC3GpDnHfjxBaJ1JDthGhx34ERk04K84_s8hKVfLn8V3ut2vN4v5NgXGpE7B1kJS1syAcSuoNJYrlXNjcrDaFFSA5DVy0SMAJRC4VrU1qFDWXDeFGJLP9-_LoLpEd4L4qJ4m1ctE_AH_qkPv</recordid><startdate>20190709</startdate><enddate>20190709</enddate><creator>Yang, Yinfei</creator><creator>Cer, Daniel</creator><creator>Ahmad, Amin</creator><creator>Guo, Mandy</creator><creator>Law, Jax</creator><creator>Constant, Noah</creator><creator>Abrego, Gustavo Hernandez</creator><creator>Yuan, Steve</creator><creator>Tar, Chris</creator><creator>Sung, Yun-Hsuan</creator><creator>Strope, Brian</creator><creator>Kurzweil, Ray</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20190709</creationdate><title>Multilingual Universal Sentence Encoder for Semantic Retrieval</title><author>Yang, Yinfei ; Cer, Daniel ; Ahmad, Amin ; Guo, Mandy ; Law, Jax ; Constant, Noah ; Abrego, Gustavo Hernandez ; Yuan, Steve ; Tar, Chris ; Sung, Yun-Hsuan ; Strope, Brian ; Kurzweil, Ray</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a1157-adb3501f8a12d305cd26642cc4ad7c903a52be23305aa63ea276bdce6e5b27f93</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2019</creationdate><topic>Computer Science - Computation and Language</topic><toplevel>online_resources</toplevel><creatorcontrib>Yang, Yinfei</creatorcontrib><creatorcontrib>Cer, Daniel</creatorcontrib><creatorcontrib>Ahmad, Amin</creatorcontrib><creatorcontrib>Guo, Mandy</creatorcontrib><creatorcontrib>Law, Jax</creatorcontrib><creatorcontrib>Constant, Noah</creatorcontrib><creatorcontrib>Abrego, Gustavo Hernandez</creatorcontrib><creatorcontrib>Yuan, Steve</creatorcontrib><creatorcontrib>Tar, Chris</creatorcontrib><creatorcontrib>Sung, Yun-Hsuan</creatorcontrib><creatorcontrib>Strope, Brian</creatorcontrib><creatorcontrib>Kurzweil, Ray</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Yang, Yinfei</au><au>Cer, Daniel</au><au>Ahmad, Amin</au><au>Guo, Mandy</au><au>Law, Jax</au><au>Constant, Noah</au><au>Abrego, Gustavo Hernandez</au><au>Yuan, Steve</au><au>Tar, Chris</au><au>Sung, Yun-Hsuan</au><au>Strope, Brian</au><au>Kurzweil, Ray</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Multilingual Universal Sentence Encoder for Semantic Retrieval</atitle><date>2019-07-09</date><risdate>2019</risdate><abstract>We introduce two pre-trained retrieval focused multilingual sentence encoding models, respectively based on the Transformer and CNN model architectures. The models embed text from 16 languages into a single semantic space using a multi-task trained dual-encoder that learns tied representations using translation based bridge tasks (Chidambaram al., 2018). The models provide performance that is competitive with the state-of-the-art on: semantic retrieval (SR), translation pair bitext retrieval (BR) and retrieval question answering (ReQA). On English transfer learning tasks, our sentence-level embeddings approach, and in some cases exceed, the performance of monolingual, English only, sentence embedding models. Our models are made available for download on TensorFlow Hub.</abstract><doi>10.48550/arxiv.1907.04307</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.1907.04307
ispartof
issn
language eng
recordid cdi_arxiv_primary_1907_04307
source arXiv.org
subjects Computer Science - Computation and Language
title Multilingual Universal Sentence Encoder for Semantic Retrieval
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-24T02%3A31%3A42IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Multilingual%20Universal%20Sentence%20Encoder%20for%20Semantic%20Retrieval&rft.au=Yang,%20Yinfei&rft.date=2019-07-09&rft_id=info:doi/10.48550/arxiv.1907.04307&rft_dat=%3Carxiv_GOX%3E1907_04307%3C/arxiv_GOX%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true