Since the increasing development of information technology, its integration and curriculum resources have become a new point of view in education and teaching reform. As a result, educational concepts, teaching models, and learning methods have been changed. Moreover, new requirements for the teaching reform of ideological and political courses in colleges and universities have also been put forward. Aiming at the current educational mechanism of ideological education courses in colleges and universities, which has a single form, is not well targeted and lacks synergy. This paper studies and establishes a brand-new three-dimensional hybrid education concept. The characteristics of this concept include (1) timeliness; that is, the basic views of Marxism are unified with the characteristics of the times; (2) pertinence, that is, the combination of ideological education and the characteristics of the students’ growth stage; (3) openness; that is, the new curriculum content and creative thinking are connected. Therefore, the three-dimensional hybrid teaching that combines networked teaching with the traditional teacher-centered teaching model has become an inevitable trend of classroom teaching reform at this stage. This paper develops a college ideological education course recommendation system based on deep learning, based on a hybrid collaborative filtering algorithm, and by introducing the effectiveness of the gradually forgetting curve based on changes in user feature, it better solves the shortcomings of traditional collaborative filtering algorithms, such as low efficiency and weak adaptability. Further, a corresponding recommendation system for ideological guidance courses in colleges and universities has been developed. The system runs stably and has strong practicability and robustness. It is of positive significance for creating an ideological and educational atmosphere with different forms and innovation for teachers and students.