Analysis and Prediction of New Media Information Dissemination of Police Microblog

Abstract

: This paper aims to analyze the microblog data published by the official account in a certain province of China, and finds out the rule of Weibo that is easier to be forwarded in the new police media perspective. In this paper, a new topic-based model is proposed. Firstly, the LDA topic clustering algorithm is used to extract the topic categories with forwarding heat from the microblogs with high forwarding numbers, then the Naive Bayesian algorithm is used to topic categories. The sample data is processed to predict the type of microblog forwarding. In order to evaluate this method, a large number of microblog online data is used to analysis. The experimental results show that the proposed method can accurately predict the forwarding of Weibo. on this, we propose an experimental method to predict the forwarding behavior of Weibo. The method is based on the LDA model and is modeled using the Naïve Bayes algorithm for prediction. Experiments show that there are two popular forwarding themes in public security police microblog: social hotspot case notification and life safety. From the final recall and precision of the model, this experimental method has certain accurate prediction ability. Through the predictions of the model, the life warning class (preventing fraud, etc.) is the most popular type of microblog tweets that can be forwarded by users. It can be seen from the displayed topic category keywords that the user forwards relevant content before and after the college entrance examination.