Abstract With the development of the information age, big data analysis has become the basis for social management and the development of new ideas for intelligent governance. In this paper, the governance process of colleges and universities is presented and stored in the form of data by simulating the content of governance through the full domain data set, and the relationship between different subjects and object features is mapped through the generalized Cartesian product. From there, various data patterns are clustered to effectively manage various types of operations’ linkage. The multi-objective optimization algorithm is used to solve the Pareto optimal solution for different governance subjects, i.e., the mutual coordination between teachers and students. The objective optimization function is established to identify the Pareto frontiers of various governance programs and emergency plans. The results show that in the driving pathway of college big data governance, the correct rate of classification of policing grid governance events in 28 colleges and universities is 0.8279 on average, which effectively improves governance efficiency. Big data governance in colleges and universities enables college governance to continuously move from decentralized to centralized, from part to whole, and to achieve accurate and efficient governance modes.