A finite-state method, based on leftmost longest-match replacement, is presented for segmenting words into graphemes, and for converting graphemes into phonemes. A small set of hand-crafted conversion rules for Dutch achieves a phoneme accuracy of over 93%. The accuracy of the system is further improved by using transformation-based learning. The phoneme accuracy of the best system (using a large set of rule templates and a `lazy' variant of Brill's algoritm), trained on only 40K words, reaches 99% accuracy.
Cite
@article{arxiv.cs/0003074,
title = {A Finite State and Data-Oriented Method for Grapheme to Phoneme Conversion},
author = {Gosse Bouma},
journal= {arXiv preprint arXiv:cs/0003074},
year = {2007}
}