Metraipolis › archive › dc_a30a1e5ff247

Dictionary-Constrained Grapheme-to-Phoneme for Unsegmented Languages from LLM-Annotated Data — arXiv:2609.19805

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.19805
License (as archived)
CC0-1.0
Archived
2026-10-10T03:48:07.774Z
SHA-256 of the archived text
b09ae3d0218f8ca5e23a3a2cd09d7655888e8e5c7e2ec094e84ca6f50f007b27
Stable URL of this page
https://metraipolis.com/documento/dc_a30a1e5ff247

How to cite

Dictionary-Constrained Grapheme-to-Phoneme for Unsegmented Languages from LLM-Annotated Data — arXiv:2609.19805 — arXiv (en), https://arxiv.org/abs/2609.19805 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_a30a1e5ff247 (Metraipolis archive dc_a30a1e5ff247, sha256 b09ae3d0218f8ca5e23a3a2cd09d7655888e8e5c7e2ec094e84ca6f50f007b27)
BibTeX
@misc{dca30a1e5ff247,
  title = {Dictionary-Constrained Grapheme-to-Phoneme for Unsegmented Languages from LLM-Annotated Data — arXiv:2609.19805},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.19805},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_a30a1e5ff247; sha256 b09ae3d0218f8ca5e23a3a2cd09d7655888e8e5c7e2ec094e84ca6f50f007b27},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.19805. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Grapheme-to-phoneme (G2P) conversion turns raw text into its phonemic form and is an essential part of both text-to-speech (TTS) and automatic speech recognition (ASR) systems. It is required to be fast, stable and context-aware. For unsegmented languages such as Japanese, G2P additionally couples word segmentation with highly context-dependent polyphone disambiguation, and the scarcity of accurately annotated data remains a bottleneck. In this paper, we present a context-aware, segmentation-agnostic neural G2P framework that models the joint segmentation-and-reading hypothesis space, scoring paths of a discriminative conditional random field (CRF) over a dictionary-derived word lattice. To tackle data scarcity, we utilize large language models (LLMs) to generate more than 2 million sentences. Experimental results demonstrate that our method substantially outperforms conventional morphological analyzer-based methods and neural sequence models. On the Joyo-Kanji-Yomi benchmark, our method reaches 99.62% target word reading accuracy, 0.32% target word phoneme error rate (PER) and 0.14% sentence PER.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-18 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2609.19805 — 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 2026-10-10T03:24:08.133Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-18?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).