Distance Based Korean WordNet(alias. KorLex) Embedding Model

The objective of this study was to create graph embedding vectors using Korean WordNet (KorLex) and apply them to neural network word-embedding models. Semantic knowledge, especially lexical semantic knowledge in a language, can be represented by word-embedding vectors or graph structures of lexical...

Full description

Saved in:
Bibliographic Details
Main Authors: SeongReol Park, JoongMin Shin, Sanghyun Cho, Hyuk-Chul Kwon, Jung-Hun Lee
Format: Article
Language:English
Published: Taylor & Francis Group 2024-12-01
Series:Applied Artificial Intelligence
Online Access:https://www.tandfonline.com/doi/10.1080/08839514.2024.2398920
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850064198788186112
author SeongReol Park
JoongMin Shin
Sanghyun Cho
Hyuk-Chul Kwon
Jung-Hun Lee
author_facet SeongReol Park
JoongMin Shin
Sanghyun Cho
Hyuk-Chul Kwon
Jung-Hun Lee
author_sort SeongReol Park
collection DOAJ
description The objective of this study was to create graph embedding vectors using Korean WordNet (KorLex) and apply them to neural network word-embedding models. Semantic knowledge, especially lexical semantic knowledge in a language, can be represented by word-embedding vectors or graph structures of lexical databases, such as WordNet. Both representations capture common semantics; however, some semantic knowledge is only captured in a specific way or not at all. In a previous study, Path2vec mapped WordNet graphs to graph-embedding vectors using similarity scores between two words. In this study, we propose two main approaches. First, we mapped the knowledge in the Korean lexical database KorLex onto graph-embedding vectors. We then applied these embedding vectors to deep neural network word embeddings to capture additional semantic knowledge in the Korean language. On a custom test set, the proposed approach improved performance by capturing additional semantic knowledge in similarity and analogy analyses. We plan to apply a variant of this to other deep neural embedding models.
format Article
id doaj-art-acae19d3f125427ca17de4814fb2f602
institution DOAJ
issn 0883-9514
1087-6545
language English
publishDate 2024-12-01
publisher Taylor & Francis Group
record_format Article
series Applied Artificial Intelligence
spelling doaj-art-acae19d3f125427ca17de4814fb2f6022025-08-20T02:49:22ZengTaylor & Francis GroupApplied Artificial Intelligence0883-95141087-65452024-12-0138110.1080/08839514.2024.2398920Distance Based Korean WordNet(alias. KorLex) Embedding ModelSeongReol Park0JoongMin Shin1Sanghyun Cho2Hyuk-Chul Kwon3Jung-Hun Lee4Department of Information Convergence Engineering, Pusan National University, Busan, South KoreaDepartment of Information Convergence Engineering, Pusan National University, Busan, South KoreaDepartment of Information Convergence Engineering, Pusan National University, Busan, South KoreaDepartment of Information Convergence Engineering, Pusan National University, Busan, South KoreaDepartment of Artificial Intelligence, Dong-Eui University, Busan, South KoreaThe objective of this study was to create graph embedding vectors using Korean WordNet (KorLex) and apply them to neural network word-embedding models. Semantic knowledge, especially lexical semantic knowledge in a language, can be represented by word-embedding vectors or graph structures of lexical databases, such as WordNet. Both representations capture common semantics; however, some semantic knowledge is only captured in a specific way or not at all. In a previous study, Path2vec mapped WordNet graphs to graph-embedding vectors using similarity scores between two words. In this study, we propose two main approaches. First, we mapped the knowledge in the Korean lexical database KorLex onto graph-embedding vectors. We then applied these embedding vectors to deep neural network word embeddings to capture additional semantic knowledge in the Korean language. On a custom test set, the proposed approach improved performance by capturing additional semantic knowledge in similarity and analogy analyses. We plan to apply a variant of this to other deep neural embedding models.https://www.tandfonline.com/doi/10.1080/08839514.2024.2398920
spellingShingle SeongReol Park
JoongMin Shin
Sanghyun Cho
Hyuk-Chul Kwon
Jung-Hun Lee
Distance Based Korean WordNet(alias. KorLex) Embedding Model
Applied Artificial Intelligence
title Distance Based Korean WordNet(alias. KorLex) Embedding Model
title_full Distance Based Korean WordNet(alias. KorLex) Embedding Model
title_fullStr Distance Based Korean WordNet(alias. KorLex) Embedding Model
title_full_unstemmed Distance Based Korean WordNet(alias. KorLex) Embedding Model
title_short Distance Based Korean WordNet(alias. KorLex) Embedding Model
title_sort distance based korean wordnet alias korlex embedding model
url https://www.tandfonline.com/doi/10.1080/08839514.2024.2398920
work_keys_str_mv AT seongreolpark distancebasedkoreanwordnetaliaskorlexembeddingmodel
AT joongminshin distancebasedkoreanwordnetaliaskorlexembeddingmodel
AT sanghyuncho distancebasedkoreanwordnetaliaskorlexembeddingmodel
AT hyukchulkwon distancebasedkoreanwordnetaliaskorlexembeddingmodel
AT junghunlee distancebasedkoreanwordnetaliaskorlexembeddingmodel