This paper establishes a psychological cognition-based model of Civic Education and obtains a basic training corpus through contour coefficients, which is preprocessed according to the basic classification criteria. Based on the Chinese participle classifier, a data mining algorithm was used to identify the standardized labels in the text. In the test application stage, the overall scope of course recommendations is set, and the best teaching program is produced according to the degree of influence of teaching quality. The teaching evaluation system’s performance can be improved by using gray correlation based on user feedback on classification results. In the analysis of correlation and influence factors, students focus on learning Introduction to the Fundamental Principles of Marxism and Mao Zedong Thought, with a learning rate of 93.3%, and the weight value of the combination of teaching and research business force and teaching reform is higher than the standard of 0.6. The loading value of curriculum setting and educational resource allocation is 0.865 and 0.898, respectively. Colleges and universities can enhance their teaching programs by integrating students’ psycho-cognitive characteristics.