The concept of digital education has produced an unnoticeable impact on the teaching of Civics courses in colleges and universities. The study starts from the digital transformation of the curriculum ideological and political education in college courses to analyze and explore, based on the Internet of Things (IoT) technology to collect data from students’ campus card data, combined with the improved K-means clustering algorithm and correlation matrix, to establish a digital assessment model of curriculum ideological and political education based on the correlation matrix. The model’s data mining performance is evaluated through simulation tests. Then, students of a higher vocational college are selected as the research object, and the model is used to realize the construction of student portraits and the cluster analysis of student behavior and to mine the association rules between student behavior and the effect of curriculum ideological and political education. The data fluctuation range of this paper’s method is small and stable, within 0.35~0.50, with better data analysis performance. The model categorizes the sample students into 6 types of consumption regularity behaviors, 8 types of life regularity behaviors and 5 types of study effort level behaviors, and obtains 9 association rules between students’ behaviors and the effectiveness of Civic and Political teaching. The digital evaluation of curriculum ideological and political education can be effectively achieved using the campus card data mining model based on matrix association, as evidenced by the empirical analysis.