Undergraduate self-consciousness is still in the formative stage and is highly susceptible to external environment. Schools must adapt to the new forms and requirements, improve their ability to search, analyze, and prevent online public opinion through big data technologies; we should make early discovery, early action, and early resolution, so as not to give time and space for public opinion crisis fermentation and consolidate the dominant position of the mainstream ideology. This paper focuses on the university network public opinion monitoring and early warning system based on big data and analyzes the university public opinion from the perspective of the specificity, sensitivity, identification, early warning and prevention ability of the system to deal with public opinion, and effectively solves the network public opinion crisis. Combined with the current actual situation and specific needs, this paper explores the intervention measures and guidance strategies for the network public opinion of college emergencies, so as to lay a solid foundation for the healthy, stable, and sustainable development of colleges and universities.

With the popularity of the Internet and the coverage of new media technologies, the widespread spread of online public opinion on public emergencies has a certain amount of negative impact on public sentiment and social development. In order to systematically grasp the research dynamics and cutting-edge themes of the literature from a macro perspective, this paper took the literature related to online public opinion on public emergencies in CSSCI journals as the research object based on the database of China National Knowledge Infrastructure (CNKI), and comprehensively applied statistical measurement methods to sort out the development lineage, institutional distribution, journal carriers and author influence of this field. Secondly, this study used citespace knowledge mapping technology to visualize and present keywords. Based on this, the clustering outcomes of the LLR algorithm were analyzed to summarize keyword evolution and topic variation from technical, informational, and managerial multipe dimensions, and to contrast the variations in methodologies and topic selection in various literature. Finally, the study identified cutting-edge research directions  for the current and future long-term development of the subject, including empirical analysis of big data, information ecological perspectives, and government management.