Due to the existing “island” state of psychological and behavioral data, there is no way for anyone to access students’ psychological and behavioral histories. This limits the comprehensive understanding and effective intervention of college students’ mental health status. Therefore, this article constructs an artificial intelligence-based psychological health and intervention system for college students. Firstly, this article obtains psychological health testing data of college students through online platforms or on-campus system design, distribution of questionnaires, feedback from close contacts of students, and internal campus resources. Then, the architecture of a mental health monitoring system is designed. Its overall architecture includes a data collection layer, a data processing layer, a decision tree algorithm layer, and an evaluation display layer. The system uses the C4.5 decision tree algorithm to calculate the information gain of the processed sample data, selects the attribute with the maximum value, and constructs a decision tree structure model to evaluate students’ mental health. Finally, this article studies the evaluation of students’ mental health status by combining multidimensional information such as the SCL-90 scale, self-assessment scale, and student behavior data. Experimental data shows that the system can effectively identify students’ mental health problems and provide precise intervention measures based on their situation, with high accuracy and practicality.