Development of Japanese Speech Database Read by Non-native Speakers for Constructing CALL System

This paper describes the construction and evaluation of Japanese speech database read by non-native speakers in order to develop CALL systems by a research project. The project has been organized with interdisciplinary members such as specialists in second language acquisition, phonetics and speech processing. The database recorded 140 non-native speakers, who were all overseas students at eight universities in Japan, to recognize distinctive features of pronunciation and prosody. The project also constructed a native Japanese speech database, which was used as a control data based on the same text read by the non-native speakers. The corpus of the database is distributed on 5CD-ROMs, and includes the reading text sentences, words and dialogues. Professional language teachers and linguists evaluated this database. 1