This paper proposes a statistical mapping-based technique for guessing pronunciations of novel words from their spellings. The technique is based on the automatic determination and utilization of unidirectional mappings between n-tuples of characters and n-tuples of phonemes, and may be viewed as a statistical extension of analogy-based pronunciation guessing algorithms.
Bibliographic reference. Singh, Rita / Gouvêa, Evandro B. / Raj, Bhiksha (2007): "Probabilistic deduction of symbol mappings for extension of lexicons", In INTERSPEECH-2007, 1789-1792.