Propensity detection model based on micro-message process and flow detection instruction distribution

Since the media services such as microblog,WeChat flourish,risk prediction becomes the primary problem of micro-message public opinion management.This paper designs public opinion tendency detection model based on conceptual framework of SDN and MapReduce combining virtual honey net technology.For the front-end honeypot,this paper sets up public opinion monitoring task instruction set,and arranges detection strategy,then completes distributed traffic detection task.Through the virtual suspected theme,and in the face of the difficulties of large data sparse,it designs user sensitive behavior feature set,and realizes priori algorithm about interest tendency of micro-message circle.Finally,algorithm model is tested by practice.The practice proves that propensity subject detection based on flow level and process level association is efficient,and pertinence is strong,and it can obtain good monitoring results.It provides important support for the active prevention of micro-message public opinion and regulation control of public opinion.