Skip to content
This repository has been archived by the owner on Mar 27, 2022. It is now read-only.

Sequence-to-sequence Pre-training with Data Augmentation for Sentence Rewriting #6

Open
tm4roon opened this issue Oct 11, 2019 · 0 comments
Open

Sequence-to-sequence Pre-training with Data Augmentation for Sentence Rewriting #6

tm4roon opened this issue Oct 11, 2019 · 0 comments
Labels
Data Augmentation Data Augmentation Grammatical Error Correction Grammatical Error Correction Style Transfer Style Transfer
Projects
arXiv

Comments

@tm4roon
Copy link
Owner

tm4roon commented Oct 11, 2019 •
edited
Loading

Sequence-to-sequence Pre-training with Data Augmentation for Sentence Rewriting

文書き換えタスク(論文では、文法誤り訂正とスタイル変換)におけるデータ拡張手法を提案。生成した擬似データと教師データを同時に用いて学習させるのではなく、擬似データを学習したのちに、教師データでfine-tuningを行っている。

擬似データは逆翻訳によって生成を行うが、言語モデル(文法誤り訂正)や二値分類器(スタイル変換)を用いてフィルタリングすることで、学習に効果的なデータのみを抽出している。

文献情報
@tm4roon tm4roon added Data Augmentation Data Augmentation Grammatical Error Correction Grammatical Error Correction Style Transfer Style Transfer labels Oct 11, 2019
@tm4roon tm4roon added this to 2019 in arXiv Oct 11, 2019
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Assignees
No one assigned
Labels
Data Augmentation Data Augmentation Grammatical Error Correction Grammatical Error Correction Style Transfer Style Transfer
Projects
arXiv
2019
Milestone
No milestone
Development

No branches or pull requests
1 participant
@tm4roon