A backpropagation neural network model with adaptive feature extraction for music emotion recognition in online music appreciation

With the rapid advancement of computer science and artificial intelligence, the integration of digital technologies in music appreciation has opened new avenues for understanding emotional responses to music. This study explores the relationship between students’ emotions and music categories in online music appreciation courses. To achieve this, a music emotion recognition model based on backpropagation neural networks (BP-NN) is proposed. The model extracts key musical features and classifies emotions using a combination of psychological and computational models, including the Hevner and Thayer emotion models. The study constructs a dataset comprising 500 pre-classified musical pieces, training the BP-NN on 324 samples and testing it on the remaining 176 pieces. Experimental results demonstrate the effectiveness of the proposed model, achieving high accuracy in emotion classification. The findings contribute to the development of intelligent music appreciation systems, enhancing personalized learning experiences by adapting content based on students’ emotional responses.