Deep Learning-Based Abstractive Summarization for Brazilian Portuguese Texts

Detalhes bibliográficos
Autor(a) principal: Paiola, Pedro H. [UNESP]
Data de Publicação: 2022
Outros Autores: de Rosa, Gustavo H. [UNESP], Papa, João P. [UNESP]
Tipo de documento: Artigo de conferência
Idioma: eng
Título da fonte: Repositório Institucional da UNESP
Texto Completo: http://dx.doi.org/10.1007/978-3-031-21689-3_34
http://hdl.handle.net/11449/249521
Resumo: Automatic summarization captures the most relevant information and condenses it into an understandable text in natural language. Such a task can be classified as either extractive or abstractive summarization. Research on Brazilian Portuguese-based abstractive summarization is still scarce. This work explores abstractive summarization in Portuguese-based texts using a deep learning-based approach. The results are relatively satisfactory considering the ROUGE measurements and the quality of the generated summaries. Still, there are some problems regarding coherence, readability, and grammar. We strongly believe they are linked to the inherent complexity of generating an abstract and the degradation of text quality by the translation steps. These results should be seen as preliminary, serving as a basis for future research.
id UNSP_fc0e68d186d48f51d121c40bb530e2ae
oai_identifier_str oai:repositorio.unesp.br:11449/249521
network_acronym_str UNSP
network_name_str Repositório Institucional da UNESP
repository_id_str 2946
spelling Deep Learning-Based Abstractive Summarization for Brazilian Portuguese TextsAbstractive summarizationBrazilian PortugueseMachine learningNatural language processingSummarizationAutomatic summarization captures the most relevant information and condenses it into an understandable text in natural language. Such a task can be classified as either extractive or abstractive summarization. Research on Brazilian Portuguese-based abstractive summarization is still scarce. This work explores abstractive summarization in Portuguese-based texts using a deep learning-based approach. The results are relatively satisfactory considering the ROUGE measurements and the quality of the generated summaries. Still, there are some problems regarding coherence, readability, and grammar. We strongly believe they are linked to the inherent complexity of generating an abstract and the degradation of text quality by the translation steps. These results should be seen as preliminary, serving as a basis for future research.Department of Computing São Paulo State University, SPDepartment of Computing São Paulo State University, SPUniversidade Estadual Paulista (UNESP)Paiola, Pedro H. [UNESP]de Rosa, Gustavo H. [UNESP]Papa, João P. [UNESP]2023-07-29T16:01:59Z2023-07-29T16:01:59Z2022-01-01info:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/conferenceObject479-493http://dx.doi.org/10.1007/978-3-031-21689-3_34Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), v. 13654 LNAI, p. 479-493.1611-33490302-9743http://hdl.handle.net/11449/24952110.1007/978-3-031-21689-3_342-s2.0-85145252790Scopusreponame:Repositório Institucional da UNESPinstname:Universidade Estadual Paulista (UNESP)instacron:UNESPengLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)info:eu-repo/semantics/openAccess2024-04-23T16:11:19Zoai:repositorio.unesp.br:11449/249521Repositório InstitucionalPUBhttp://repositorio.unesp.br/oai/requestopendoar:29462024-08-05T16:22:25.405525Repositório Institucional da UNESP - Universidade Estadual Paulista (UNESP)false
dc.title.none.fl_str_mv Deep Learning-Based Abstractive Summarization for Brazilian Portuguese Texts
title Deep Learning-Based Abstractive Summarization for Brazilian Portuguese Texts
spellingShingle Deep Learning-Based Abstractive Summarization for Brazilian Portuguese Texts
Paiola, Pedro H. [UNESP]
Abstractive summarization
Brazilian Portuguese
Machine learning
Natural language processing
Summarization
title_short Deep Learning-Based Abstractive Summarization for Brazilian Portuguese Texts
title_full Deep Learning-Based Abstractive Summarization for Brazilian Portuguese Texts
title_fullStr Deep Learning-Based Abstractive Summarization for Brazilian Portuguese Texts
title_full_unstemmed Deep Learning-Based Abstractive Summarization for Brazilian Portuguese Texts
title_sort Deep Learning-Based Abstractive Summarization for Brazilian Portuguese Texts
author Paiola, Pedro H. [UNESP]
author_facet Paiola, Pedro H. [UNESP]
de Rosa, Gustavo H. [UNESP]
Papa, João P. [UNESP]
author_role author
author2 de Rosa, Gustavo H. [UNESP]
Papa, João P. [UNESP]
author2_role author
author
dc.contributor.none.fl_str_mv Universidade Estadual Paulista (UNESP)
dc.contributor.author.fl_str_mv Paiola, Pedro H. [UNESP]
de Rosa, Gustavo H. [UNESP]
Papa, João P. [UNESP]
dc.subject.por.fl_str_mv Abstractive summarization
Brazilian Portuguese
Machine learning
Natural language processing
Summarization
topic Abstractive summarization
Brazilian Portuguese
Machine learning
Natural language processing
Summarization
description Automatic summarization captures the most relevant information and condenses it into an understandable text in natural language. Such a task can be classified as either extractive or abstractive summarization. Research on Brazilian Portuguese-based abstractive summarization is still scarce. This work explores abstractive summarization in Portuguese-based texts using a deep learning-based approach. The results are relatively satisfactory considering the ROUGE measurements and the quality of the generated summaries. Still, there are some problems regarding coherence, readability, and grammar. We strongly believe they are linked to the inherent complexity of generating an abstract and the degradation of text quality by the translation steps. These results should be seen as preliminary, serving as a basis for future research.
publishDate 2022
dc.date.none.fl_str_mv 2022-01-01
2023-07-29T16:01:59Z
2023-07-29T16:01:59Z
dc.type.status.fl_str_mv info:eu-repo/semantics/publishedVersion
dc.type.driver.fl_str_mv info:eu-repo/semantics/conferenceObject
format conferenceObject
status_str publishedVersion
dc.identifier.uri.fl_str_mv http://dx.doi.org/10.1007/978-3-031-21689-3_34
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), v. 13654 LNAI, p. 479-493.
1611-3349
0302-9743
http://hdl.handle.net/11449/249521
10.1007/978-3-031-21689-3_34
2-s2.0-85145252790
url http://dx.doi.org/10.1007/978-3-031-21689-3_34
http://hdl.handle.net/11449/249521
identifier_str_mv Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), v. 13654 LNAI, p. 479-493.
1611-3349
0302-9743
10.1007/978-3-031-21689-3_34
2-s2.0-85145252790
dc.language.iso.fl_str_mv eng
language eng
dc.relation.none.fl_str_mv Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
dc.rights.driver.fl_str_mv info:eu-repo/semantics/openAccess
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv 479-493
dc.source.none.fl_str_mv Scopus
reponame:Repositório Institucional da UNESP
instname:Universidade Estadual Paulista (UNESP)
instacron:UNESP
instname_str Universidade Estadual Paulista (UNESP)
instacron_str UNESP
institution UNESP
reponame_str Repositório Institucional da UNESP
collection Repositório Institucional da UNESP
repository.name.fl_str_mv Repositório Institucional da UNESP - Universidade Estadual Paulista (UNESP)
repository.mail.fl_str_mv
_version_ 1808128639037865984

Registros relacionados