Mental health problems not only affect students’ academic performance, but also may have a long-term impact on their personal growth and social adaptability. In this paper, a data-driven mental health management model for college students is proposed by combining fuzzy logic decision-making and risk warning mechanism. First, data mining methods are used to extract the characteristics of students’ consumption behavior and Internet behavior, and feature correlation analysis is performed to discover the behavioral differences between normal and abnormal mental health student groups. Then, based on the fuzzy clustering algorithm, the students’ mental health status is categorized. The experimental results show that when the number of clusters is 8, the clustering effect of the model is the best, and the error sum of squares is significantly reduced. Specifically, students’ behavioral characteristics such as dietary regularity, diligence, number of shared meals, and length of time spent on the Internet are strongly associated with their mental health status. Through clustering analysis and risk prediction, the accuracy of the model reaches more than 95%, which can effectively warn of mental health risks. The study shows that the combination of students’ daily behavioral data and advanced data analysis technology can provide strong support for mental health intervention in colleges and universities.
