Gender bias in multilingual neural machine translation: The architecture matters
Multilingual Neural Machine Translation architectures mainly differ in the amount of sharing modules and parameters among languages. In this paper, and from an algorithmic perspective, we explore if the chosen architecture, when trained with the same data, influences the gender bias accuracy. Experi...
| Autores: | , , , , , |
|---|---|
| Tipo de recurso: | informe técnico |
| Fecha de publicación: | 2020 |
| País: | España |
| Institución: | Universitat Politècnica de Catalunya (UPC) |
| Repositorio: | UPCommons. Portal del coneixement obert de la UPC |
| Idioma: | inglés |
| OAI Identifier: | oai:upcommons.upc.edu:2117/348079 |
| Acceso en línea: | https://hdl.handle.net/2117/348079 |
| Access Level: | acceso abierto |
| Palabra clave: | Sexism in language Machine translating Natural language processing (Computer science) Sexisme en el llenguatge Traducció automàtica Tractament del llenguatge natural (Informàtica) Àrees temàtiques de la UPC::Enginyeria de la telecomunicació::Processament del senyal::Processament de la parla i del senyal acústic |
| id |
ES_c756d9f77d93b2bac567d27647ef67ff |
|---|---|
| oai_identifier_str |
oai:upcommons.upc.edu:2117/348079 |
| network_acronym_str |
ES |
| network_name_str |
España |
| repository_id_str |
|
| spelling |
Gender bias in multilingual neural machine translation: The architecture mattersRuiz Costa-Jussà, Marta|||0000-0002-5703-520XEscolano Peinado, Carlos|||0000-0001-6657-673XBasta, Christine Raouf SaadFerrando Monsonís, Javier|||0000-0002-2637-0961Batlle, RoserKharitonova, KseniaSexism in languageMachine translatingNatural language processing (Computer science)Sexisme en el llenguatgeTraducció automàticaTractament del llenguatge natural (Informàtica)Àrees temàtiques de la UPC::Enginyeria de la telecomunicació::Processament del senyal::Processament de la parla i del senyal acústicMultilingual Neural Machine Translation architectures mainly differ in the amount of sharing modules and parameters among languages. In this paper, and from an algorithmic perspective, we explore if the chosen architecture, when trained with the same data, influences the gender bias accuracy. Experiments in four language pairs show that Language-Specific encoders-decoders exhibit less bias than the Shared encoder-decoder architecture. Further interpretability analysis of source embeddings and the attention shows that, in the LanguageSpecific case, the embeddings encode more gender information, and its attention is more diverted. Both behaviors help in mitigating gender bias.This work is supported by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No. 947657).20202020-12-2420212021-06-29reporthttp://purl.org/coar/resource_type/c_93fcAOhttp://purl.org/coar/version/c_b1a7d7d4d402bcceinfo:eu-repo/semantics/reportapplication/pdfhttps://hdl.handle.net/2117/348079reponame:UPCommons. Portal del coneixement obert de la UPCinstname:Universitat Politècnica de Catalunya (UPC)InglésengEuropean Commission http://doi.org/10.13039/100010661 Horizon 2020 Framework Programme 947657 Lifelong UNiversal lAnguage Representationopen accesshttp://purl.org/coar/access_right/c_abf2info:eu-repo/semantics/openAccessoai:upcommons.upc.edu:2117/3480792026-05-27T15:37:01Z |
| dc.title.none.fl_str_mv |
Gender bias in multilingual neural machine translation: The architecture matters |
| title |
Gender bias in multilingual neural machine translation: The architecture matters |
| spellingShingle |
Gender bias in multilingual neural machine translation: The architecture matters Ruiz Costa-Jussà, Marta|||0000-0002-5703-520X Sexism in language Machine translating Natural language processing (Computer science) Sexisme en el llenguatge Traducció automàtica Tractament del llenguatge natural (Informàtica) Àrees temàtiques de la UPC::Enginyeria de la telecomunicació::Processament del senyal::Processament de la parla i del senyal acústic |
| title_short |
Gender bias in multilingual neural machine translation: The architecture matters |
| title_full |
Gender bias in multilingual neural machine translation: The architecture matters |
| title_fullStr |
Gender bias in multilingual neural machine translation: The architecture matters |
| title_full_unstemmed |
Gender bias in multilingual neural machine translation: The architecture matters |
| title_sort |
Gender bias in multilingual neural machine translation: The architecture matters |
| dc.creator.none.fl_str_mv |
Ruiz Costa-Jussà, Marta|||0000-0002-5703-520X Escolano Peinado, Carlos|||0000-0001-6657-673X Basta, Christine Raouf Saad Ferrando Monsonís, Javier|||0000-0002-2637-0961 Batlle, Roser Kharitonova, Ksenia |
| author |
Ruiz Costa-Jussà, Marta|||0000-0002-5703-520X |
| author_facet |
Ruiz Costa-Jussà, Marta|||0000-0002-5703-520X Escolano Peinado, Carlos|||0000-0001-6657-673X Basta, Christine Raouf Saad Ferrando Monsonís, Javier|||0000-0002-2637-0961 Batlle, Roser Kharitonova, Ksenia |
| author_role |
author |
| author2 |
Escolano Peinado, Carlos|||0000-0001-6657-673X Basta, Christine Raouf Saad Ferrando Monsonís, Javier|||0000-0002-2637-0961 Batlle, Roser Kharitonova, Ksenia |
| author2_role |
author author author author author |
| dc.subject.none.fl_str_mv |
Sexism in language Machine translating Natural language processing (Computer science) Sexisme en el llenguatge Traducció automàtica Tractament del llenguatge natural (Informàtica) Àrees temàtiques de la UPC::Enginyeria de la telecomunicació::Processament del senyal::Processament de la parla i del senyal acústic |
| topic |
Sexism in language Machine translating Natural language processing (Computer science) Sexisme en el llenguatge Traducció automàtica Tractament del llenguatge natural (Informàtica) Àrees temàtiques de la UPC::Enginyeria de la telecomunicació::Processament del senyal::Processament de la parla i del senyal acústic |
| description |
Multilingual Neural Machine Translation architectures mainly differ in the amount of sharing modules and parameters among languages. In this paper, and from an algorithmic perspective, we explore if the chosen architecture, when trained with the same data, influences the gender bias accuracy. Experiments in four language pairs show that Language-Specific encoders-decoders exhibit less bias than the Shared encoder-decoder architecture. Further interpretability analysis of source embeddings and the attention shows that, in the LanguageSpecific case, the embeddings encode more gender information, and its attention is more diverted. Both behaviors help in mitigating gender bias. |
| publishDate |
2020 |
| dc.date.none.fl_str_mv |
2020 2020-12-24 2021 2021-06-29 |
| dc.type.none.fl_str_mv |
report http://purl.org/coar/resource_type/c_93fc AO http://purl.org/coar/version/c_b1a7d7d4d402bcce |
| dc.type.openaire.fl_str_mv |
info:eu-repo/semantics/report |
| format |
report |
| dc.identifier.none.fl_str_mv |
https://hdl.handle.net/2117/348079 |
| url |
https://hdl.handle.net/2117/348079 |
| dc.language.none.fl_str_mv |
Inglés eng |
| language_invalid_str_mv |
Inglés |
| language |
eng |
| dc.relation.none.fl_str_mv |
European Commission http://doi.org/10.13039/100010661 Horizon 2020 Framework Programme 947657 Lifelong UNiversal lAnguage Representation |
| dc.rights.none.fl_str_mv |
open access http://purl.org/coar/access_right/c_abf2 |
| dc.rights.openaire.fl_str_mv |
info:eu-repo/semantics/openAccess |
| rights_invalid_str_mv |
open access http://purl.org/coar/access_right/c_abf2 |
| eu_rights_str_mv |
openAccess |
| dc.format.none.fl_str_mv |
application/pdf |
| dc.source.none.fl_str_mv |
reponame:UPCommons. Portal del coneixement obert de la UPC instname:Universitat Politècnica de Catalunya (UPC) |
| instname_str |
Universitat Politècnica de Catalunya (UPC) |
| reponame_str |
UPCommons. Portal del coneixement obert de la UPC |
| collection |
UPCommons. Portal del coneixement obert de la UPC |
| repository.name.fl_str_mv |
|
| repository.mail.fl_str_mv |
|
| _version_ |
1869419148740657152 |
| score |
15.301603 |