Improving Word Embedding Using Variational Dropout

Pre-trained word embeddings are essential in natural language processing (NLP). In recent years, many post-processing algorithms have been proposed to improve the pre-trained word embeddings. We present a novel method - Orthogonal Auto Encoder with Variational Dropout (OAEVD) for improving word embe...

Full description

Saved in:
Bibliographic Details
Main Authors: Zainab Albujasim, Diana Inkpen, Xuejun Han, Yuhong Guo
Format: Article
Language:English
Published: LibraryPress@UF 2023-05-01
Series:Proceedings of the International Florida Artificial Intelligence Research Society Conference
Subjects:
Online Access:https://journals.flvc.org/FLAIRS/article/view/133326
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850270973374234624
author Zainab Albujasim
Diana Inkpen
Xuejun Han
Yuhong Guo
author_facet Zainab Albujasim
Diana Inkpen
Xuejun Han
Yuhong Guo
author_sort Zainab Albujasim
collection DOAJ
description Pre-trained word embeddings are essential in natural language processing (NLP). In recent years, many post-processing algorithms have been proposed to improve the pre-trained word embeddings. We present a novel method - Orthogonal Auto Encoder with Variational Dropout (OAEVD) for improving word embeddings based on orthogonal autoencoders and variational dropout.  Specifically, the orthogonality constraint encourages more diversity in the latent space and increases semantic similarities between similar words, and variational dropout makes it more robust to overfitting.   Empirical evaluation on a range of downstream NLP tasks, including semantic similarity, text classification, and concept categorization shows that our proposed method effectively improves the quality of pre-trained word embeddings. Moreover, the proposed method successfully reduces the dimensionality of pre-trained word embeddings while maintaining high performance.
format Article
id doaj-art-8c87ffa94e2e4e1f9c0485795cf87826
institution OA Journals
issn 2334-0754
2334-0762
language English
publishDate 2023-05-01
publisher LibraryPress@UF
record_format Article
series Proceedings of the International Florida Artificial Intelligence Research Society Conference
spelling doaj-art-8c87ffa94e2e4e1f9c0485795cf878262025-08-20T01:52:22ZengLibraryPress@UFProceedings of the International Florida Artificial Intelligence Research Society Conference2334-07542334-07622023-05-013610.32473/flairs.36.13332669632Improving Word Embedding Using Variational DropoutZainab Albujasim0https://orcid.org/0009-0003-3569-6802Diana Inkpen1https://orcid.org/0000-0002-0202-2444Xuejun Han2https://orcid.org/0000-0001-5238-3291Yuhong Guo3https://orcid.org/0000-0002-7621-342XCarleton UniversityUniversity of OttawaCarleton UniversityCarleton UniversityPre-trained word embeddings are essential in natural language processing (NLP). In recent years, many post-processing algorithms have been proposed to improve the pre-trained word embeddings. We present a novel method - Orthogonal Auto Encoder with Variational Dropout (OAEVD) for improving word embeddings based on orthogonal autoencoders and variational dropout.  Specifically, the orthogonality constraint encourages more diversity in the latent space and increases semantic similarities between similar words, and variational dropout makes it more robust to overfitting.   Empirical evaluation on a range of downstream NLP tasks, including semantic similarity, text classification, and concept categorization shows that our proposed method effectively improves the quality of pre-trained word embeddings. Moreover, the proposed method successfully reduces the dimensionality of pre-trained word embeddings while maintaining high performance.https://journals.flvc.org/FLAIRS/article/view/133326word embeddingpost-processingvariational dropoutdimensionality reduction
spellingShingle Zainab Albujasim
Diana Inkpen
Xuejun Han
Yuhong Guo
Improving Word Embedding Using Variational Dropout
Proceedings of the International Florida Artificial Intelligence Research Society Conference
word embedding
post-processing
variational dropout
dimensionality reduction
title Improving Word Embedding Using Variational Dropout
title_full Improving Word Embedding Using Variational Dropout
title_fullStr Improving Word Embedding Using Variational Dropout
title_full_unstemmed Improving Word Embedding Using Variational Dropout
title_short Improving Word Embedding Using Variational Dropout
title_sort improving word embedding using variational dropout
topic word embedding
post-processing
variational dropout
dimensionality reduction
url https://journals.flvc.org/FLAIRS/article/view/133326
work_keys_str_mv AT zainabalbujasim improvingwordembeddingusingvariationaldropout
AT dianainkpen improvingwordembeddingusingvariationaldropout
AT xuejunhan improvingwordembeddingusingvariationaldropout
AT yuhongguo improvingwordembeddingusingvariationaldropout