[Purpose/significance] Aiming at the current public opinion response problem of multimedia network in China, this paper combs and proposes the mechanism of government organization response path integration matching and response workflow model based on crisis risk classification. In order to improve the resource integration ability for management decision-makers, the network public opinion crisis provides a reference for the accurate response effectiveness.[Method/process] This paper analyzes the role of the dynamic factors of network public opinion crisis in big data environment, extracts the basic coefficients of network public opinion crisis risk classification, and extracts the network public opinion crisis risk classification superposition coefficient from the perspective of the multi-structure of network public opinion information carried by multimedia technology. Then the paper analyzes the work points of the network public opinion crisis response under each risk classification one by one.[Result/conclusion] According to the ordering combination of the main structural elements, the media effectiveness factors and the object attribute elements, the credit base level of the multimedia network is established on the relationship nodes. According to the different influence effects of the ontology component on the public opinion crisis, the public opinion crisis of the big data multimedia network is established. Establishing a multimedia network public opinion crisis risk classification model, deconstructing the public opinion crisis risk into a ranking coefficient and an addition coefficient, and a more accurate description of the lyric crisis is conducive to judging the future development of the grievance crisis and improving the speed and accuracy of matching with the previous grievance crisis case.