Data-Driven Approach for Formality-Sensitive Machine Translation: Language-Specific Handling and Synthetic Data Generation

TL;DR

This paper presents a data-driven method for formality-sensitive machine translation that leverages language-specific data handling and synthetic data generation with large language models, significantly improving translation quality.

Contribution

It introduces a novel combination of language-specific data handling and synthetic data generation techniques for FSMT, enhancing translation performance.

Findings

Significant improvement over baseline models

Effective use of large-scale language models for synthetic data

Prompt engineering enhances translation quality

Abstract

In this paper, we introduce a data-driven approach for Formality-Sensitive Machine Translation (FSMT) that caters to the unique linguistic properties of four target languages. Our methodology centers on two core strategies: 1) language-specific data handling, and 2) synthetic data generation using large-scale language models and empirical prompt engineering. This approach demonstrates a considerable improvement over the baseline, highlighting the effectiveness of data-centric techniques. Our prompt engineering strategy further improves performance by producing superior synthetic translation examples.