Method for predicting and simulating network public opinion in emergency based on SOAR model

Abstract

The invention discloses a method for predicting and simulating a network public opinion in an emergency based on an SOAR model. The method comprises the steps of based on the SOAR model, taking net citizens as agents, and taking a net citizen group behavior transformation process in the network public opinion as a continuous transformation process of state with time in a public opinion problem space; in combination with network public opinion development and government response related theories, dividing a net citizen group into Me-formers and In-formers, and dividing a network public opinion development phase into generation, bursting, ripening and attenuation; designing working memory, long-term memory, decision-making process and learning mechanisms of the net citizen agent; constructing a net citizen group behavior transformation rule base; and establishing a net citizen group behavior transformation SOAR agent model based on the SOAR model. The attenuation phase in the network public opinion development process is taken as an example, the prediction and simulation for the network public opinion in the emergency are realized, and thus the influence of different government responses on different types of net citizen group behaviors is evaluated by the method.

Description

technical field [0001] The invention relates to a method for predicting and simulating microblog network public opinion in emergencies, and in particular to a simulation method based on behavior rules of netizen groups established on a SOAR model. Background technique [0002] With the continuous popularization of Web2.0 technology and related Internet applications, new media such as Weibo have become an important public opinion field of Internet public opinion. Weibo has the characteristics of a large user base, fast transmission speed, and convenient information uploading. It has become the main source and medium of public opinion outbreaks in my country, spreading like a virus to every corner of the Internet, businesses, and personal life. How to study an effective microblog public opinion prediction and simulation method has become a top priority in view of the massive unstructured text data, large number of users and strong real-time characteristics of microblog. [00...

AI Technical Summary

Problems solved by technology

[0005] At present, domestic scholars' research on government emergency management in Internet public opinion often focuses on theoretical suggestions. Foreign scholars study government emergency measures in crisis management from the perspective of public management, but lack of analysis and modeling of real-time data. The development trend of public opinion to provide clear suggestions for the government to take emergency measures

Images