Target-Side Augmentation for Document-Level Machine Translation

Guangsheng Bao, Zhiyang Teng, Yue Zhang
5/8/2023
cs.CL

Official code: baoguangsheng/target-side-augmentation (15 stars, MIT, last commit 2023-08-15).

Abstract

Document-level machine translation faces the challenge of data sparsity due to its long input length and a small amount of training data, increasing the risk of learning spurious patterns. To address this challenge, we propose a target-side augmentation method, introducing a data augmentation (DA) model to generate many potential translations for each source document. Learning on these wider range translations, an MT model can learn a smoothed distribution, thereby reducing the risk of data sparsity. We demonstrate that the DA model, which estimates the posterior distribution, largely improves the MT performance, outperforming the previous best system by 2.30 s-BLEU on News and achieving new state-of-the-art on News and Europarl benchmarks. Our code is available at https://github.com/baoguangsheng/target-side-augmentation.

View on arXivView PDF

Code Implementations(1)

baoguangsheng/target-side-augmentationOfficial100%
MIT

Code base for "Target-Side Augmentation for Document-Level Machine Translation"

153Shell, PythonMay 8, 20233 years agoMIT

Similar Papers

Textual Augmentation Techniques Applied to Low Resource Machine Translation: Case of Swahili

85% match2023

A Paradigm Shift in Machine Translation: Boosting Translation Performance of Large Language Models

85% match1 repo5882026

Understanding and Bridging the Modality Gap for Speech Translation

84% match1 repo162023

TreeSwap: Data Augmentation for Machine Translation via Dependency Subtree Swapping

83% match1 repo72023

Text Style Transfer Back-Translation

83% match2023

Building Accurate Translation-Tailored LLMs with Language Aware Instruction Tuning

83% match1 repo42024

Cite this paper

@article{bao2023targetside,
  title  = {Target-Side Augmentation for Document-Level Machine Translation},
  author = {Guangsheng Bao and Zhiyang Teng and Yue Zhang},
  year   = {2023},
  eprint = {2305.04505},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2305.04505v2}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2305.04505.svg)](https://codeofpaper.com/paper/2305.04505)

Discussion