Abstract: 【 Purpose/significance】 In recent years, there has been a surge in network public opinion incidents related to the ethics and conduct of university faculty. Due to the unique nature of the entities involved, these incidents garner higher attention and pose el⁃ evated risks. Assessing and forecasting the risks associated with such events is beneficial for enhancing the crisis management capa⁃ bilities of universities in handling public opinion.【 Method/process】The research focused on network public opinion incidents pertain⁃ ing to the ethics and conduct of university faculty. Drawing from pertinent research findings, it employed an information ecology per⁃ spective to establish a three-dimensional network public opinion risk assessment indicator system, covering information content, infor⁃ mation people, and information environment. The entropy weight TOPSIS method was applied to determine indicator weights, followed by a thorough evaluation of multi-indicator risks. The K-means clustering method categorized public opinion incidents into risk lev⁃ els, resulting in the establishment of a risk assessment and early warning model.【 Result/conclusion】The research, taking into consid⁃ eration the characteristics of network public opinion incidents related to the ethics and conduct of university faculty, constructs a risk assessment indicator system. The crisis risk is categorized into three levels: low-risk type, moderate-risk type, and high-risk type. Corresponding to the different risk levels, it identifies analytical differences in indicators and characteristics of university network pub⁃ lic opinion and proposes corresponding public opinion control strategies. 【Innovation/limitation】This research selects indicators and optimizes quantification methods in combination with the characteristics of specific types of network public opinion. Future research may consider verifying the accuracy of the assessment results and improving the universality of assessment and early warning by com⁃ bining more instance data.