利用 GPT-4 進行自動翻譯後編輯

Leveraging GPT-4 for Automatic Translation Post-Editing

May 24, 2023

作者: Vikas Raunak, Amr Sharaf, Hany Hassan Awadallah, Arul Menezes

cs.AI

摘要

儘管神經機器翻譯（NMT）代表著機器翻譯（MT）的主流方法，NMT 模型的輸出仍需要經過翻譯後編輯來糾正錯誤並提升質量，特別是在關鍵情境下。在這項研究中，我們將翻譯後編輯的任務與大型語言模型（LLMs）形式化，並探索使用 GPT-4 自動編輯 NMT 輸出在多種語言對之間。我們的結果顯示，GPT-4 擅長進行翻譯後編輯，即使目標語言不是英語，也能產生有意義的編輯。值得注意的是，我們在 WMT-22 英中、英德、中英和德英語言對上，使用基於 GPT-4 的編輯後翻譯實現了最先進的表現，經由最先進的機器翻譯質量指標評估。

English

While Neural Machine Translation (NMT) represents the leading approach to Machine Translation (MT), the outputs of NMT models still require translation post-editing to rectify errors and enhance quality, particularly under critical settings. In this work, we formalize the task of translation post-editing with Large Language Models (LLMs) and explore the use of GPT-4 to automatically post-edit NMT outputs across several language pairs. Our results demonstrate that GPT-4 is adept at translation post-editing and produces meaningful edits even when the target language is not English. Notably, we achieve state-of-the-art performance on WMT-22 English-Chinese, English-German, Chinese-English and German-English language pairs using GPT-4 based post-editing, as evaluated by state-of-the-art MT quality metrics.