Research on Network Public Opinion Combination Prediction Based on Analytic Hierarchy Process

In response to the difficulty of accurately predicting the three-stage evolution process of complex network public opinion using traditional single prediction models, this paper conducts research on the combination prediction method of network public opinion. Firstly, based on the evolution process of network public opinion, a three-stage mathematical model of incubation diffusion regression was constructed; Secondly, taking into account the characteristics of the Logistic model, exponential smoothing model, and grey prediction, a combination prediction model judgment matrix is constructed based on the scheme layer, and a network public opinion combination prediction method based on the analytic hierarchy process is proposed; Finally, taking the Pelosi hot topic public opinion event as an example, the accuracy and effectiveness of the combined prediction algorithm in public opinion prediction were verified.