There is a disconnect between mental health intervention and value guidance in modern ideological and political education of college students, for this reason, this paper proposes a big data-driven mental health assessment model for college students. It constructs the college students’ mental health assessment index system, defines the text similarity calculation model and weight allocation rules. Optimize the search process of FCM clustering center based on the firefly algorithm to avoid the local optimal problem of the traditional algorithm. Taking 5000 students in a university as the research object, relying on expert assessment to determine the weights of indicators. The effectiveness of this paper’s algorithm is tested through controlled experiments, and the characteristics of different psychological state levels are analyzed with the help of clustering results. The clustering effect of the FCM-FA algorithm proposed in this paper has obvious advantages compared with the FCM algorithm and the Grid_PFcm algorithm, with short time-consuming and smooth movement, and the time is controlled within 0.1s. The difference between the career impact dimension scores of different mental health status groups is small, and the difference in the emotion regulation dimension is the largest, i.e., students with lower scores in the emotion regulation dimension are more likely to be categorized as “poor” or “bad”. The application of fuzzy cluster analysis in the analysis of college students’ mental health can help colleges and universities to carry out early prevention of college students and formulate corresponding strategies for the intervention of psychological disorders.
