This paper focuses on the correlation analysis of the thinking operation model, learning motivation, and ideological and political education practices. According to the correlation between the thinking operation model and learning motivation, internal motivation and external motivation are analyzed for motivation to learn. The core elements of the ideological and political education practice system in colleges and universities are finely classified, and students’ learning motivation and academic achievement are selected as the main research objects. A combination of stack autoencoder and Gaussian mixture model is used to constitute a deep clustering model of learning motivation for intervention analysis. The change in achievement was significant before and after the intervention twice, p<0.1, which means that the achievement after the intervention was significantly higher. Colleges and universities should mobilize students’ learning motivation to enhance the effectiveness of ideological and political education practices in colleges and universities.