Research on the Public Opinion Pre-warning Based on the Logistic Model

At present, public opinion research on early warning model always use almost all history data, which can't predict the development trend of events timely. The paper presents an early warning model of public opinion based on the logistic model to dynamically describe the propagation rule and warn timely on public opinion. Regarding the hot words in micro-index as the research object, the paper studied the logistic model by mathematical modeling method, researched the characteristic point-in-time calculation of public opinion diffusion curve and monitoring and early warning of public opinion. The method was verified correctly and efficiently with the time series of the development trend of an incident in Sina micro-blog and the result showed that the study is practically useful.