Powered by OpenAIRE graph
Found an issue? Give us feedback
addClaim

Performance Comparison of Word2vec and fastText Embedding Models

Word2vec 및 fastText 임베딩 모델의 성능 비교
Authors: Hyungsuc Kang; Janghoon Yang;

Performance Comparison of Word2vec and fastText Embedding Models

Abstract

Word2vec 임베딩 모델은 단순하고 성능이 우수하기 때문에, 자연어 처리 분야에서 가장 널리 쓰이는 모델 중 하나이지만 몇 가지 한계도 있다. 이런 한계를 극복하기 위해 일반적인 언어에 적용 가능한 fastText 임베딩 모델이 제안되었고, 이후 한국어에 적합한 특정한 fastText 모델도 제안되었다. 본 연구는 유사도 검사, 유추 검사 및 감정 분석을 통해 몇 가지 word2vec 및 fastText 모델의 성능을 비교 평가하는 것을 목표로 한다. fastText 모델을 제안한 이전 연구의 결과와는 달리, 최소한 유추 검사와 감정 분석의 측면에서는 fastText 모델이 word2vec 모델보다 더 우수하다고 단정 지을 수 없다는 것이 본 연구의 결론이다. 특히 특정 다운스트림 태스크의 경우, word2vec 임베딩으로도 fastText 임베딩보다 작은 복잡도를 가지면서 유사하거나 더 나은 성능을 얻을 수 있음을 확인하였다.

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    6
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Top 10%
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Average
Powered by OpenAIRE graph
Found an issue? Give us feedback
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
6
Top 10%
Average
Average
Upload OA version
Are you the author of this publication? Upload your Open Access version to Zenodo!
It’s fast and easy, just two clicks!