Abstract:

In the progress of teaching Chinese pronunciation for students, teachers find out that it is quite difficult. This work takes a lot of time to reach the right pronunciation for each student, and it also seems to be the same to test Chinese pronunciation. To solve these difficulties, we would like to propose a new approach in Chinese pronunciation test at elementary level, this approach takes the advantage of AI voice recognition to recognize the right pronunciation. With this study, we can save time of teachers and improve the quality of Chinese pronunciation for each individual student when they are guided and shown by the approach. In this approach, we build an application using AI voice recognition. The app using a set of input data with the sample of 100 words, each word has about 20 to 30 qualified pronunciation voices which is used to be trained in AI technology. After the experiment results, we received positive feedbacks from students and teachers while they were carrying the Chinese pronunciation test at elementary level. Comparing to the manual results of the Chinese lecturers / experts from HCMC University of Education (HCMUE), the accuracy of the proposal using AI voice recognition is up to 95%. Through the experiment, most of lecturers were extremely satisfied with the consequence of the new proposed approach, and it is opening a new methodology in Chinese pronunciation test.

Keywords:

Chinese pronunciation, Pronunciation test, AI voice recognition, learning Chinese.