|
SYDNEY, AUSTRALIA - Appen Speech Technology, a provider of multi-language transcription, lexicon (phonetic dictionaries) and grammar services, has announced the completion of their Korean lexicon. The Korean lexicon is the latest offering in Appen's roll-out of lexicon material across a range of languages. The lexicon contains around 21,000 unique words across a range of categories, including place names, personal (family and given) names, company names and commonly used Korean words. It is available in both Hangul and romanised form. The lexicon is designed for both recognition and synthesis, containing multiple variants for each headword where applicable. It is also designed to be flexible for conversion into multiple phonetic formats. Korean is the latest language to be undertaken by Appen in their roll-out program for multi-language lexicons. Appen has already completed multiple projects in languages from Asia, Europe and the Middle East, as well as multiple dialects of English. Dr. Julie Vonwiller, Appen principal, says: "Appen has invested heavily in developing highly sophisticated tools, processes and skills for lexicon projects. These tools and processes are targeted at achieving efficiency, high through-put and quality. Combine this with the fact that we only use native speakers of the language or dialect, and we can complete large-scale lexicons in rapid time, at competitive cost, and to high quality standards." "To give an indication of what we can achieve", adds Vonwiller, "Appen completed a 300,000 word lexicon in a European language in 6 weeks. This included multiple variants for each headword. And we did this with an error rate of less than 1%."
|