STAR : Sentence Translation Alignment Rate for Document-to-Document Machine Translation

ACL ARR 2026 January Submission5983 Authors

Keywords: Document-level Machine Translation, Preference Optimization, Large Language Models, Quality Estimation

Abstract: Large Language Models (LLMs) have enabled a shift from sentence-level to document-to-document (Doc2Doc) machine translation, promising improved global coherence. However, document-to-document generation in a single pass frequently suffers from structural misalignment, manifesting as sentence omissions or hallucinations that violate the core requirement of source-target correspondence. To address this, we introduce **S**entence **T**ranslation **A**lignment **R**ate (**STAR**), an auxiliary metric that explicitly quantifies sentence-level structural fidelity. Building on this, we propose **STAR**-masked **P**reference **O**ptimization (**StarPO**), a framework that ranks document-level hypotheses by structural quality and utilizes a dynamic alignment mask to focus optimization on misaligned segments. Experimental results across news and literary domains demonstrate that StarPO significantly enhances translation quality and structural integrity. Notably, StarPO allows compact models (e.g., Qwen3-4B) to surpass the performance of massive proprietary systems like GPT-4o while maintaining superior token efficiency. We will release our code and datasets.

Paper Type: Long

Research Area: Machine Translation

Research Area Keywords: Machine Translation

Contribution Types: Model analysis & interpretability, NLP engineering experiment

Languages Studied: English, Chinese, Russian, German, Spanish

Submission Number: 5983