This paper divides the Course on the Ideological and Political Dimension into four dimensions based on data mining technology and establishes a four-dimensional classroom teaching system. By analyzing the main algorithms of data mining, cluster analysis is selected as the main method to divide the Ideological and Political Science dimensions among the three main types of algorithm analysis. The expectation maximization algorithm in the clustering algorithm is used to introduce the approximate values of each dimension parameter. The unknown categories corresponding to the dimensions were calculated using the EM algorithm, which combined with Jensen’s inequality to turn the random variables in the set into constants to ensure that the samples of the same category have high similarity. The results show that the percentage of integration of the school’s Civic content with the profession is low, only 0.1, and the ideal figure should be 0.4.