초등영어교육의 음성인식 소프트웨어 활용과 전망

A study on the prospect of using speech recognition softwares in elementary English education

The study holds two purposes: 1) to review the problems of speech recognition softwares that have been used for beginning English education; 2) to suggest features that should be considered in developing future speech recognition softwares. The study has found that the present speech recognition softwares are mostly computer-directed, hindering the self-motivation of the student in learning activities. In these softwares, the student is usually involved in: 1) comparing wave forms of his pronunciation with those of the computer; 2) speaking a limited number of sentences that are already assigned in a role play; 3) selecting his answer from some choices, which are not clearly distinguished in speech recognition; 4) practicing some useful expressions, which are often mistaken as another phrase in speech recognition. In order to facilitate interactive activities between the computer and students, prospective softwares should focus more on developing softwares: 1) to apply the branching method to allow more student-directed conversations; 2) to be able to identify various expressions of the same illocution and those of different structures; 3) to provide immediate and various feedbacks for the student's performance; 4) to employ stimulating games which require many oral activities; and finally 5) to utilize a user-friendly hardware which requires a less amount of student adjustment to the machine.