Saved in:
Bibliographic Details
Main Authors: Wu, Si, Wieting, John, Smith, David A.
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2412.18707
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909511453769728
author Wu, Si
Wieting, John
Smith, David A.
author_facet Wu, Si
Wieting, John
Smith, David A.
contents While a source sentence can be translated in many ways, most machine translation (MT) models are trained with only a single reference. Previous work has shown that using synthetic paraphrases can improve MT. This paper investigates best practices for employing multiple references by analyzing the semantic similarity among different English translations of world literature in the Par3 dataset. We classify the semantic similarity between paraphrases into three levels: low, medium, and high, and fine-tune three different models (mT5-large, LLaMA-2-7B, and Opus-MT) for literary MT tasks. Across different models, holding the total training instances constant, single-reference but more source texts only marginally outperforms multiple-reference with half of the source texts. Moreover, when fine-tuning an LLM, using paraphrases with medium and high semantic similarity outperforms an unfiltered dataset, with improvements in BLEU (0.3-0.5), COMET (0.1-0.9), and chrF++ (0.17-0.32). Our code is publicly available on GitHub.
format Preprint
id arxiv_https___arxiv_org_abs_2412_18707
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Multiple References with Meaningful Variations Improve Literary Machine Translation
Wu, Si
Wieting, John
Smith, David A.
Computation and Language
While a source sentence can be translated in many ways, most machine translation (MT) models are trained with only a single reference. Previous work has shown that using synthetic paraphrases can improve MT. This paper investigates best practices for employing multiple references by analyzing the semantic similarity among different English translations of world literature in the Par3 dataset. We classify the semantic similarity between paraphrases into three levels: low, medium, and high, and fine-tune three different models (mT5-large, LLaMA-2-7B, and Opus-MT) for literary MT tasks. Across different models, holding the total training instances constant, single-reference but more source texts only marginally outperforms multiple-reference with half of the source texts. Moreover, when fine-tuning an LLM, using paraphrases with medium and high semantic similarity outperforms an unfiltered dataset, with improvements in BLEU (0.3-0.5), COMET (0.1-0.9), and chrF++ (0.17-0.32). Our code is publicly available on GitHub.
title Multiple References with Meaningful Variations Improve Literary Machine Translation
topic Computation and Language
url https://arxiv.org/abs/2412.18707