Traditional teaching methods are difficult to meet the personalized learning needs of students, and the teaching effect is uneven. The application of intelligent algorithms in the field of education provides a new idea to solve this problem, using computer technology to analyze students’ learning characteristics and knowledge mastery to realize the intelligent generation and personalized push of the content of ideological and political education. In this study, we constructed a personalized intelligent grouping model of ideological and political education content based on artificial fish swarm algorithm and a personalized test question recommendation model PMF-CD based on probability matrix decomposition, and verified through experiments that the distribution of knowledge points in the test paper generated by the personalized intelligent grouping model is more targeted and aggregated, while the distribution of knowledge points in the traditional grouping method is more dispersed. In the test question recommendation model test, PMF-CD model in DATASET2 dataset 40% test set conditions of the accuracy rate of 98.26%, much higher than the traditional DINA model of 54.39%. Practical experiments show that the experimental class using the model of this study has a significantly higher level of ideology and morality than the control class in the six dimensions of healthy life, ecological civilization, patriotism, scientific spirit, social responsibility and civic literacy, of which the average value of the experimental class in the dimension of patriotism reaches 4.876, while the control class is 4.372, with a significant difference (P<0.000). The results of the study show that the ideological and political education content generation and personalized push strategy based on intelligent algorithms can effectively improve students' ideological and moral level and provide a new path for the modernization of ideological and political education.
