Abstract:

With the rapid development of artificial intelligence technology， the era of omnimedia has arrived. Network public opinion propagation has been endowed with new and unique characteristics， which also puts forward higher requirements for its guidance and governance. Therefore， it is of great significance to explore effective network public opinion guidance strategies and incentive mechanisms to build a healthy public opinion ecology.Considering the dynamic and sustainable propagation of network public opinion， combined with the actual propagation process and the research paradigm of differential game， five kinds of network public opinion propagation scenarios composed of regulator， media and netizens are constructed. By solving the equilibrium strategy of each subject， the propagation process and guiding effect of network public opinion under different incentive scenarios are theoretically analyzed. Combined with the simulation experiment， the influence of different guidance methods on the propagation and guidance effect of network public opinion is deeply analyzed， and then the key parameters in the guidance process of network public opinion are identified.The results show that the network public opinion guidance effect is the best under the dual incentive policy scenario， which can realize the Pareto optimization of the overall system income. And the network public opinion guidance effect is at the lowest level under the non-incentive policy scenario. It is worth noting that both explicit incentive and implicit incentive policies can effectively improve the guiding effect of network public opinions， but there are different intensity of influence：{\phi }_M（the cost subsidy coefficient provided by the regulator to the media）?\omega（implicit incentive coefficient）?{\phi }_N（the cost subsidy coefficient provided by the regulator to the netizens）.

Key words: network public opinion, omnimedia era, differential game, guidance strategy, incentive mechanism