About Grapheme to Phoneme Conversion
The Grapheme to Phoneme Conversion technology converts written words into their phonetic pronunciations. Given a list of words, it returns possible pronunciations for each word. Each pronunciation result includes the phonetic representation (phonemes separated by spaces) and the source of the pronunciation:
- Dictionary — the pronunciation was found in the system's built-in dictionary
- Generated — the pronunciation was automatically generated by the system
- Partly generated — part of the pronunciation was generated by the system, rest was found in the system's built-in dictionary
This technology is particularly useful for generating pronunciations for keywords not found in the dictionary, such as brand names, technical terms, or foreign words. Grapheme to Phoneme Conversion can be used as a supporting tool for the following technologies:
- Keyword Spotting — by generating phonetic pronunciations of out-of-vocabulary keywords, enabling the system to search for words it has never seen before.
- Speech to Text — by expanding the pronunciation dictionary with new words, improving transcription accuracy for domain-specific or uncommon vocabulary.