Metraipolis › archive › dc_625fa640a868

Phonemizing User-Generated Text: A Benchmark, Taxonomy, and Compositional Approach — arXiv:2609.27205

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.27205
License (as archived)
CC0-1.0
Archived
2026-10-08T09:12:16.238Z
SHA-256 of the archived text
43e6aaea5bc9949c7ffc5b4e91d03cb327dc49355a4b04e1c454cd64922fa72a
Stable URL of this page
https://metraipolis.com/documento/dc_625fa640a868

How to cite

Phonemizing User-Generated Text: A Benchmark, Taxonomy, and Compositional Approach — arXiv:2609.27205 — arXiv (en), https://arxiv.org/abs/2609.27205 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_625fa640a868 (Metraipolis archive dc_625fa640a868, sha256 43e6aaea5bc9949c7ffc5b4e91d03cb327dc49355a4b04e1c454cd64922fa72a)
BibTeX
@misc{dc625fa640a868,
  title = {Phonemizing User-Generated Text: A Benchmark, Taxonomy, and Compositional Approach — arXiv:2609.27205},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.27205},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_625fa640a868; sha256 43e6aaea5bc9949c7ffc5b4e91d03cb327dc49355a4b04e1c454cd64922fa72a},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.27205. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Text-to-speech systems increasingly process user-generated text (UGT) such as ppl and imo, whose pronunciation must be inferred from the canonical rather than surface form. We introduce UGTPhon, the first grapheme-to-phoneme (G2P) benchmark for UGT in English, Vietnamese, and Korean, together with an inference-grounded taxonomy for fine-grained diagnosis. Existing G2P models and frontier LLMs exhibit a systematic canonical-to-non-canonical performance gap, reaching up to 66.8 PER points. As a benchmark baseline, we propose a simple compositional G2P approach that incorporates canonical-form evidence through exact-match lookup and staged decoding. Across matched ByT5 and Qwen2.5-0.5B backbones, explicit canonical-form modeling consistently reduces non-canonical G2P errors. The 0.5B variant also performs competitively with much larger few-shot frontier LLMs, highlighting the benefit of explicitly modeling canonical-form inference for UGT phonemization.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-24 in cs.AI).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.27205 — 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 2026-10-08T08:13:26.629Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-24?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).