Abstract: 【Purpose/significance】Monitoring the big data information of university network public opinion and establishing a risk as⁃sessment and early warning system are extremely important for quickly and accurately identifying the risk level of university network public opinion crisis,and improving the supervision efficiency,response ability and governance level of universities and relevant gov⁃ernment departments on network public opinion【. Method/process】On the basis of researching in the influencing factors and evolution⁃ary laws of university network public opinion,construct a four-dimensional university network public opinion risk assessment index system,including the influence of public opinion publishers,public opinion popularity,public opinion intensity,and public opinion diffu⁃sion, using Random Forest algorithm and Entropy Weight method for index screening and weight calculation, comprehensive use of TOPSIS and Grey Relational Analysis method to construct a university network public opinion risk assessment and early warning
model,and divide the network public opinion risk level【. Result/conclusion】The research results show that the model has excellent ac⁃curacy and effectiveness,greatly simplifies the identification procedures of the risk assessment and crisis warning of university network public opinion,and provides reference for handling and coping with the university network public opinion reasonably and effectively.【Innovation/limitation】Increase the risk assessment index screening process, and the effective combination of the two assessment methods makes the risk assessment and early warning results more objective and accurate.