Abstract: Unconfirmed information is one of the most important research topics in the field of crisis management, because of its fast, wide and derivative dissemination. The “echo chamber” effect in social media sharply increases the complexity of propagation of unconfirmed information, bringing new challenges to management decision-making. Therefore, exploring the propagation mechanism and intervention strategies of unconfirmed information in social media is significant to social governance and policy making. This paper takes the “Shuanghuanglian incident” spread on the Weibo platform during the Covid-19 period as an example to carry out an experimental analysis. Firstly, we use complex network theories to build a micro network of user forwarding and propagation, and analyze the process of information propagation through clustering models and the designed indicators under the “echo chamber” effect. Secondly, we explore the topic tipping points in the propagation of unconfirmed information, and establish an opinion leader identification model based on the shallow and deep propagation of “echo chamber”. Thirdly, we validate the proposed model through various datasets of multiple platforms and some other unconfirmed information datasets. Finally, we propose unconfirmed information intervention strategies and policy recommendations from the perspectives of government agencies, enterprise platforms and platform users, in order to help decision makers to better manage and control the spread of unconfirmed information.

Key words: social media, unconfirmed information, echo chamber effect, opinion leader, propagation mechanism