Metraipolis › archive › dc_2bfa44cef03f

Tag-Aware Structured Text Translation: Towards a Systematic Understanding — arXiv:2609.29131

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.29131
License (as archived): CC0-1.0
Archived: 2026-10-08T04:27:18.836Z
SHA-256 of the archived text: 76626f9b55090af9d4d61cc369696b6ec1ab22e8519133f775dc23c0ede57424
Stable URL of this page: https://metraipolis.com/documento/dc_2bfa44cef03f

How to cite

Tag-Aware Structured Text Translation: Towards a Systematic Understanding — arXiv:2609.29131 — arXiv (en), https://arxiv.org/abs/2609.29131 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2bfa44cef03f (Metraipolis archive dc_2bfa44cef03f, sha256 76626f9b55090af9d4d61cc369696b6ec1ab22e8519133f775dc23c0ede57424)

BibTeX

@misc{dc2bfa44cef03f,
  title = {Tag-Aware Structured Text Translation: Towards a Systematic Understanding — arXiv:2609.29131},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.29131},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2bfa44cef03f; sha256 76626f9b55090af9d4d61cc369696b6ec1ab22e8519133f775dc23c0ede57424},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.29131. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Internet texts are replete with format tags that carry structural, semantic, and functional meaning. Current large language model (LLM)-based translation systems struggle to balance translation fluency with tag fidelity when processing tagged text. We argue that resolving this tension requires a systematic approach at three interconnected levels: data synthesis, capability building, and multi-objective alignment. At the data level, we identify and formalize a fundamental trade-off between structural tag diversity and translation naturalness in synthetic data generation; existing methods optimize for one at the expense of the other. We propose a hybrid synthesis strategy (Hy-LST) combining LLM-based synthesis tag method and Two-Stage LLM-based synthesis tag method to produce both diverse and natural tagged data. At the capability level, we decompose tag-aware translation into four sub-tasks of increasing difficulty in a multi-task supervised fine-tuning framework, enabling targeted capability acquisition and knowledge transfer. At the alignment level, we design three complementary reward functions under a group relative policy optimization framework, each targeting a distinct objective (fluency, tag fidelity, and tag-scoped translation quality), and show that joint optimization consistently outperforms single-reward alternatives. Experiments on six language directions (en2zh, en2ja, en2de, en2fr, en2ru, de2fr) demonstrate that each level contributes measurable improvements, and the complete system significantly outperforms existing methods. Qualitative analysis reveals specific error patterns and their mitigation after training with our method.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-25 in cs.AI).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.29131 — 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 2026-10-08T03:50:58.010Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-25?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).