Contextualized context2vec

Kazuki Ashihara, Yuki Arase, Tomoyuki Kajiwara, Satoru Uchida

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Lexical substitution ranks substitution candi- dates from the viewpoint of paraphrasability for a target word in a given sentence. There are two major approaches for lexical substitu- tion: (1) generating contextualized word em- beddings by assigning multiple embeddings to one word and (2) generating context embed- dings using the sentence. Herein we propose a method that combines these two approaches to contextualize word embeddings for lexi- cal substitution. Experiments demonstrate that our method outperforms the current state-of- the-art method. We also create CEFR-LP, a new evaluation dataset for the lexical substi- tution task. It has a wider coverage of sub- stitution candidates than previous datasets and assigns English proficiency levels to all target words and substitution candidates.

Original languageEnglish
Title of host publicationW-NUT@EMNLP 2019 - 5th Workshop on Noisy User-Generated Text, Proceedings
PublisherAssociation for Computational Linguistics (ACL)
Pages397-406
Number of pages10
ISBN (Electronic)9781950737840
Publication statusPublished - 2019
Event5th Workshop on Noisy User-Generated Text, W-NUT@EMNLP 2019 - Hong Kong, China
Duration: Nov 4 2019 → …

Publication series

NameW-NUT@EMNLP 2019 - 5th Workshop on Noisy User-Generated Text, Proceedings

Conference

Conference5th Workshop on Noisy User-Generated Text, W-NUT@EMNLP 2019
Country/TerritoryChina
CityHong Kong
Period11/4/19 → …

All Science Journal Classification (ASJC) codes

  • Information Systems
  • Computer Science Applications
  • Computational Theory and Mathematics

Fingerprint

Dive into the research topics of 'Contextualized context2vec'. Together they form a unique fingerprint.

Cite this