Dataset of Chinese language beginning learners reading speech and text-to-speech

• Main Author: Yoke Lian Lau
• Format: Proceedings
• Language: English; English
• Published: Figshare 2023
• Subjects: P1-1091 Philology. Linguistics; QP1-345 General Including influence of the environment
• Online Access: https://eprints.ums.edu.my/id/eprint/39039/1/ABSTRACT.pdf; https://eprints.ums.edu.my/id/eprint/39039/2/FULL%20TEXT.pdf; https://eprints.ums.edu.my/id/eprint/39039/; https://figshare.com/articles/conference_contribution/Dataset_of_Chinese_Language_Beginning_Learners_Reading_Speech_and_Text-to-Speech/25239643

Three categories of audio recordings, a web service platform, and commercial software were involved in this analysis process. The three categories of audio recordings included eight filtered student recordings, an audio recording provided by a Mandarin instructor from University Malaysia Sabah, and an audio recording generated by Text-to-Speech MP3, available at https://ttsmp3.com/. All ten recordings will be uploaded into the NCH WavePad Sound Editor to build a spectrum image. This will allow us to compare and identify differences in tone reading, emotion reading, etc. You can download the NCH WavePad Sound Editor from https://www.nch.com.au/WavePad/index.html. This information is beneficial for all Mandarin language beginners and their teachers, as it can be used for comparison purposes. With this knowledge, they can further undertake a pilot study. This study provides a foundation for investigating differences in human and machine reading from multiple perspectives. Furthermore, this dataset can be used to analyze reading speed. It serves as a crucial starting point for spectrum, voice wave, and reader behavior research