Self-Guided Curriculum Learning for Neural Machine Translation

Lei Zhou*, Liang Ding, Kevin Duh, Shinji Watanabe, Ryohei Sasano, Koichi Takeda

*この研究の対応する著者

研究成果: Conference contribution

2 被引用数 (Scopus)

抄録

In supervised learning, a well-trained model should be able to recover ground truth accurately, i.e. the predicted labels are expected to resemble the ground truth labels as much as possible. Inspired by this, we formulate a difficulty criterion based on the recovery degrees of training examples. Motivated by the intuition that after skimming through the training corpus, the neural machine translation (NMT) model “knows” how to schedule a suitable curriculum according to learning difficulty, we propose a self-guided curriculum learning strategy that encourages the NMT model to learn from easy to hard on the basis of recovery degrees. Specifically, we adopt sentence-level BLEU score as the proxy of recovery degree. Experimental results on translation benchmarks including WMT14 English?German and WMT17 Chinese?English demonstrate that our proposed method considerably improves the recovery degree, thus consistently improving the translation performance.

本文言語English
ホスト出版物のタイトルIWSLT 2021 - 18th International Conference on Spoken Language Translation, Proceedings
出版社Association for Computational Linguistics (ACL)
ページ206-214
ページ数9
ISBN(電子版)9781954085749
出版ステータスPublished - 2021
外部発表はい
イベント18th International Conference on Spoken Language Translation, IWSLT 2021 - Virtual, Bangkok, Thailand
継続期間: 2021 8月 52021 8月 6

出版物シリーズ

名前IWSLT 2021 - 18th International Conference on Spoken Language Translation, Proceedings

Conference

Conference18th International Conference on Spoken Language Translation, IWSLT 2021
国/地域Thailand
CityVirtual, Bangkok
Period21/8/521/8/6

ASJC Scopus subject areas

  • 人間とコンピュータの相互作用
  • 言語および言語学
  • 言語学および言語

フィンガープリント

「Self-Guided Curriculum Learning for Neural Machine Translation」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル