The development of classroom teaching evaluation in China is still in its infancy, despite the country’s rapid growth in music education. The quality of music instruction is not scientifically and reasonably evaluated based on actual teaching quality. When it comes to teaching and evaluating music, music teachers simply follow the classroom teaching evaluation indicators and models of any other topic in the school. An evaluation system that is merely a form, and does not have any meaningful impact on evaluation, feedback, or promotion is inevitable as a result of this. That’s why a deep-level, in-depth study of music classroom teaching’s quality evaluation system is still needed, along with a theoretical-and-practical combination. This work takes the music teaching as the starting point to carry out the corresponding research. First, this work builds a music teaching quality evaluation system via B/S architecture. It includes three parts: client, application unit, and database. The application unit includes user management, online evaluation, data management, evaluation result query and analysis. After users enter the system, they will score the teaching quality. Second, this work proposes the use of multi-scale convolution to extract data features, and proposes an improved attention mechanism to construct MS-IATT-CNN for evaluating music teaching quality. Third, this work has carried out a comprehensive and systematic experiment, and the experimental data has verified the feasibility of this work.
Juhui Qiu (Fri,) studied this question.