Metraipolis › archive › dc_90cbdbe79659

STAR : Sentence Translation Alignment Rate for Document-to-Document Machine Translation — arXiv:2608.27161

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.27161
License (as archived)
CC0-1.0
Archived
2026-10-17T15:20:23.645Z
SHA-256 of the archived text
4e046defcdcca6979f8bf4948aa97bca1a0c4d91aedd91d04cae78d887852c77
Stable URL of this page
https://metraipolis.com/documento/dc_90cbdbe79659

How to cite

STAR : Sentence Translation Alignment Rate for Document-to-Document Machine Translation — arXiv:2608.27161 — arXiv (en), https://arxiv.org/abs/2608.27161 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_90cbdbe79659 (Metraipolis archive dc_90cbdbe79659, sha256 4e046defcdcca6979f8bf4948aa97bca1a0c4d91aedd91d04cae78d887852c77)
BibTeX
@misc{dc90cbdbe79659,
  title = {STAR : Sentence Translation Alignment Rate for Document-to-Document Machine Translation — arXiv:2608.27161},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.27161},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_90cbdbe79659; sha256 4e046defcdcca6979f8bf4948aa97bca1a0c4d91aedd91d04cae78d887852c77},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.27161. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

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 Sentence Translation Alignment Rate (STAR), an auxiliary metric that explicitly quantifies sentence-level structural fidelity. Building on this, we propose STAR-masked Preference Optimization (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 to surpass the performance of massive proprietary systems like GPT-4o while maintaining superior token efficiency.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-28 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2608.27161 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-17T15:15:18.089Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-08-28?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).