The CJK Dictionary Institute (CJKI), which specializes in the compilation of large-scale CJK (Chinese, Japanese and Korean) and Arabic lexical resources, is pleased to announce the release of the Japanese Phonological Database (JPD). For the first time, JPD provides phonetic transcriptions in IPA (International Phonetic Alphabet) and accent codes that accurately indicate how Japanese words and names are pronounced in actual speech.
- Precise phonetic transcriptions in IPA include accent codes for each entry.
- Designed specifically for speech technology applications.
- Developed in collaboration with The National Language Research Institute.
- Compiled by native editors and linguists trained in Japanese phonology.
- Hundreds of thousands of variants cover the major allophonic variation patterns.
This the first database of its kind both in terms of quantity and quality. Covering the core vocabulary of Japanese and beyond, JPD was developed for The National Language Research Institute for use in a sophisticated CAL (computer-aid learning) system for Japanese language education. Parts of JPD is already been used by one of the world's leading speech technology companies for its embedded speech recognition products.
Developed by a team of experienced native Japanese editors trained in Japanese phonetics, JPD is a product of CJKI's ongoing research on CJK and Arabic phonetics and phonology and the development of phonological databases, which will culminate in the release of comprehensive IPA phonetic database for Chinese, Korean and Arabic as well.
A description of JPD and an paper describing the complex issues of Japanese phonology can be found here:
- Speech synthesis systems (text-to-speech generation or TTS).
- Speech recognition systems.
- Voice user interfaces (VUI) for mobile search applications.
- Japanese ASL applications and pedagogical research.
- Research and development of Japanese speech technology.