The Development of Internet Intelligent Platform in Art Education and Emotional Interaction Therapy for High School Students in Shanghai

Abstract

Human sentiment represents a complex aspect of psychology and consistently attracts attention across various scientific disciplines. This type of communication, inherently embedded in human interaction, conveys not just explicit messages but also a wealth of underlying emotional elements. This multifaceted nature of emotional communication is leveraged extensively in the study of emotions. In modern society, the unprecedented development of social economy and the unprecedented fierce competition for human survival have destroyed the harmony of human itself, between human beings and between human beings and nature. In the final analysis, it breaks the harmony of people's emotions, which leads to the negative development of emotions, so emotional cultivation is put on the agenda. Especially the emotional problems of senior high school students are particularly prominent. Based on the Internet intelligence platform, this paper analyzes the development of art education and emotional interaction therapy of senior high school students in Shanghai, and analyzes it through BP (Back-Propagation) neural network and DNN (Deep Neural Networks) Model. The results of the experimental evaluation indicate that the mean accuracy for identifying emotions stands at 61.4%. Within the spectrum of error assessment, the algorithm introduced in this study successfully mitigates 75.4% of potential inaccuracies. Therefore, a 60.2% improvement in the unweighted recognition rate of discrete emotions is quite meaningful, and also handles the unweighted case well. On the other hand, the recognition rate of emotional features under different samples has an average analysis strength of 67.4%.