Abstract:Abstract To improve the teaching effect of Civics courses in colleges and universities, this paper uses multi-attribute regression to compute the nonlinear representation ability, combined with the perception machine deep learning network, to get the teaching resources and interaction mechanism of Civics courses. The weighted k-shell decomposition method is used to determine the weighted undirected graph of the Civics and Political Science course, and the shortest path between nodes is used to map the link to complete the teacher’s teaching feedback and personalized teaching. The results show that the hit rate and normalized gain rate of the proposed method in the Civics course dataset in colleges and universities are as high as 0.58 and 0.39, respectively, and the performance of the Civics course in colleges and universities is excellent, which indicates that the proposed method can promote the cultivation of students’ quality of Civics and politics, and it has great significance in the field of teaching and learning.

Design of virtual simulation practice teaching platform for ideological and political course based on Perceptron model