Enhancing Korean Named Entity Recognition With Linguistic Tokenization Strategies

Tokenization is a significant primary step for the training of the Pre-trained Language Model (PLM), which alleviates the challenging Out-of-Vocabulary problem in the area of Natural Language Processing. As tokenization strategies can change linguistic understanding, it is essential to consider the...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Gyeongmin Kim, Junyoung Son, Jinsung Kim, Hyunhee Lee, Heuiseok Lim
Formato:	article
Lenguaje:	EN
Publicado:	IEEE 2021
Materias:	Named entity recognition Korean pre-trained language model natural language processing tokenization linguistic segmentation agglutinative language Electrical engineering. Electronics. Nuclear engineering TK1-9971
Acceso en línea:	https://doaj.org/article/d665db0beed8491ba11d763eda19afbd
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:d665db0beed8491ba11d763eda19afbd
record_format	dspace
spelling	oai:doaj.org-article:d665db0beed8491ba11d763eda19afbd2021-11-17T00:01:11ZEnhancing Korean Named Entity Recognition With Linguistic Tokenization Strategies2169-353610.1109/ACCESS.2021.3126882https://doaj.org/article/d665db0beed8491ba11d763eda19afbd2021-01-01T00:00:00Zhttps://ieeexplore.ieee.org/document/9610031/https://doaj.org/toc/2169-3536Tokenization is a significant primary step for the training of the Pre-trained Language Model (PLM), which alleviates the challenging Out-of-Vocabulary problem in the area of Natural Language Processing. As tokenization strategies can change linguistic understanding, it is essential to consider the composition of input features based on the characteristics of the language for model performance. This study answers the question of “Which tokenization strategy enhances the characteristics of the Korean language for the Named Entity Recognition (NER) task based on a language model?” focusing on tokenization, which significantly affects the quality of input features. We present two significant challenges for the NER task with the agglutinative characteristics in the Korean language. Next, we quantitatively and qualitatively analyze the coping process of each tokenization strategy for these challenges. By adopting various linguistic segmentation such as morpheme, syllable and subcharacter, we demonstrate the effectiveness and prove the performance between PLMs based on each tokenization strategy. We validate that the most consistent strategy for the challenges of the Korean language is a syllable based on Sentencepiece.Gyeongmin KimJunyoung SonJinsung KimHyunhee LeeHeuiseok LimIEEEarticleNamed entity recognitionKorean pre-trained language modelnatural language processingtokenizationlinguistic segmentationagglutinative languageElectrical engineering. Electronics. Nuclear engineeringTK1-9971ENIEEE Access, Vol 9, Pp 151814-151823 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Named entity recognition Korean pre-trained language model natural language processing tokenization linguistic segmentation agglutinative language Electrical engineering. Electronics. Nuclear engineering TK1-9971
spellingShingle	Named entity recognition Korean pre-trained language model natural language processing tokenization linguistic segmentation agglutinative language Electrical engineering. Electronics. Nuclear engineering TK1-9971 Gyeongmin Kim Junyoung Son Jinsung Kim Hyunhee Lee Heuiseok Lim Enhancing Korean Named Entity Recognition With Linguistic Tokenization Strategies
description	Tokenization is a significant primary step for the training of the Pre-trained Language Model (PLM), which alleviates the challenging Out-of-Vocabulary problem in the area of Natural Language Processing. As tokenization strategies can change linguistic understanding, it is essential to consider the composition of input features based on the characteristics of the language for model performance. This study answers the question of “Which tokenization strategy enhances the characteristics of the Korean language for the Named Entity Recognition (NER) task based on a language model?” focusing on tokenization, which significantly affects the quality of input features. We present two significant challenges for the NER task with the agglutinative characteristics in the Korean language. Next, we quantitatively and qualitatively analyze the coping process of each tokenization strategy for these challenges. By adopting various linguistic segmentation such as morpheme, syllable and subcharacter, we demonstrate the effectiveness and prove the performance between PLMs based on each tokenization strategy. We validate that the most consistent strategy for the challenges of the Korean language is a syllable based on Sentencepiece.
format	article
author	Gyeongmin Kim Junyoung Son Jinsung Kim Hyunhee Lee Heuiseok Lim
author_facet	Gyeongmin Kim Junyoung Son Jinsung Kim Hyunhee Lee Heuiseok Lim
author_sort	Gyeongmin Kim
title	Enhancing Korean Named Entity Recognition With Linguistic Tokenization Strategies
title_short	Enhancing Korean Named Entity Recognition With Linguistic Tokenization Strategies
title_full	Enhancing Korean Named Entity Recognition With Linguistic Tokenization Strategies
title_fullStr	Enhancing Korean Named Entity Recognition With Linguistic Tokenization Strategies
title_full_unstemmed	Enhancing Korean Named Entity Recognition With Linguistic Tokenization Strategies
title_sort	enhancing korean named entity recognition with linguistic tokenization strategies
publisher	IEEE
publishDate	2021
url	https://doaj.org/article/d665db0beed8491ba11d763eda19afbd
work_keys_str_mv	AT gyeongminkim enhancingkoreannamedentityrecognitionwithlinguistictokenizationstrategies AT junyoungson enhancingkoreannamedentityrecognitionwithlinguistictokenizationstrategies AT jinsungkim enhancingkoreannamedentityrecognitionwithlinguistictokenizationstrategies AT hyunheelee enhancingkoreannamedentityrecognitionwithlinguistictokenizationstrategies AT heuiseoklim enhancingkoreannamedentityrecognitionwithlinguistictokenizationstrategies
_version_	1718426077527801856

Enhancing Korean Named Entity Recognition With Linguistic Tokenization Strategies

Ejemplares similares