Abstract: 【Purpose/significance】With the increasing number of network public opinion events, whether to effectively deal with them has become a new requirement for the governing ability of the government. In order to provide a scientific reference for the government to deal with network public opinion timely and accurately, it is necessary to systematically explore the internal differences of network public opinion events and summarize the objective law of public opinion dissemination.【Method/process】This paper selects 15 hot net⁃ work public opinion events from 2015 to 2018, establishes an evaluation index system for government response to public opinion events, and uses DEA model to analyze the transmission process and internal law of these events.【Result/conclusion】It is found that the model presented in this paper can directly give the efficiency value of government response in public opinion events. When the effi⁃ ciency value is not high, it can further point out the inappropriate link of government response and give suggestions for improvement. 【Innovation/limitation】The next step of this paper is to establish a more accurate index system, in order to make a more in-depth thinking on the influencing factors of government response to public opinion events.The limitation is that some indexes in the evalua⁃ tion system can only be assigned by Delphi method. Considering the subjective factors, it is difficult to accurately quantify, and it is greatly limited by time and conditions.