Development of English Speech Database Read by Japanese to Support CALL Research

This paper describes development of ERJ (English Read by Japanese) database, which is designed to support CALL (Computer Assisted Language Learning) research. The DB is divided into two parts, English read by Japanese and that by Americans with the same reading sheets in both. The reading material is composed of four sections, prosodic/phonetic × sentence/word sections. 202 Japanese students were randomly selected from various universities all over Japan and they were requested to repeat reading the sheets until they judged that they did the correct pronunciation. The reading sheets contained helpful guidelines to the correct pronunciation, such as phonemic symbols, prosodic marks, and so on. Before the recording, the students were allowed to do some practices. The resulting DB can be viewed as a volume of English pronunciations which are correct at least for Japanese students. After collecting speech samples, five American teachers of English were asked to rate utterances of a part of the DB based on three criteria, phonetic (segmental), rhythmic, and intonational aspects of pronunciation. This paper also describes several on-going projects with this DB.