Improving Mandarin Chinese Learning in Tibetan Second-Language Learning by Artificial Intelligent Speech Technology
Weitong Guo, Hongwu Yang, Zhenye Gan · 2019 International Joint Conference on Information, Media and Engineering (IJCIME) · 2019
Mandarin plays an important role in the world and it is also the national common language in China that means mandarin need to be also popularized in ethnic minority areas. Tibetan is located in remote areas and their language is mainly Tibetan, which makes it difficult for Tibetan students to learn Mandarin and it has seriously affected the popularization of Mandarin in minority areas. Therefore, this paper proposed an effective method to assist Tibetan students' second language learning by using intelligent speech technology. Firstly, speech recognition algorithm is adopted to detect pronunciation errors. Secondly, speech synthesis method is utilized to correct pronunciation errors. And thirdly, the STRAIGHT algorithm is used to adjust speaking rate to further help Tibetan students understand the correct pronunciation. We carry out the experiment with 20 Tibetan students and the results shows that the method can effectively reduce the error rate of Mandarin pronunciation and improve the accuracy of learning Mandarin. The result also shows that as the speaking rate becomes slower, the comprehension accuracy increases.