Abstract: 【Purpose/significance】In order to effectively restrict the dangerous information dissemination in social networks, a dynam⁃
ic information behavior management mechanism based on reputation is proposed.【Method/process】The study calculated the reputa⁃
tion of node by its identity authentication and historical behavior; and classified the risk types after calculating the risk index of con⁃
tent through the risk weight value of sensitive words. Finally, the information dissemination control model based on reputation (RBDC)
is proposed, which is an authority rule according to comprehensive reputation and content risk index.【Result/conclusion】The simula⁃
tion results show that compared with the unrestricted information dissemination scenarios, the forwarding permission mapping relation⁃
ship, which is based on the fine-grained classification of content risk and user reputation, can significantly limit the propagation speed
and scope of high-risk information. 【Innovation / limitation】The reputation authorization propagation inhibition model needs to be op⁃
erated under the condition of ensuring the authenticity of social network user attribute data and the effective maintenance of sensitive
word risk vocabulary. The RBDC model can realize the dynamic adjustment mechanism of information dissemination, inhibit risk infor⁃
mation dissemination, and promote the self-discipline of network nodes.