Research On The Division And Governance Methods Of The Life Cycle Of Internet Rumors

With the vigorous development of network technology in recent years,Internet information has also shown explosive growth.While information brings convenience to people,it also has its disadvantages-it is difficult for people to distinguish between true and false information on the Internet.When people spread false information,rumors arise.The spread of rumors has brought great adverse effects to the society,the most typical of which is the online rumors about COVID-19.In public health emergencies,people are more prone to panic and anxiety than usual,and it is difficult to distinguish the true and false information in a short period of time,which can easily be misleading.In this paper,the BERT model is established,and the machine learning technology based on two-way testing is used to train the natural language model.For the downstream tasks,the steps of feature extraction,network building,overfitting control and parameter tuning are omitted,and the categories of online rumors are quickly classified,which can quickly and effectively improve the quality and efficiency of natural language processing problem solving,and finally obtain the accuracy Recall rate and F1 value.Analyze four typical types of online rumors,fit the heat curve of online rumors,and finally summarize that the spread stages of online rumors are generally divided into the embryonic stage,the outbreak stage,the variation stage,the extinction stage and the recurrence stage.Through in-depth research on each stage of online rumors spread under COVID-19,it is concluded that online rumors have the characteristics of concealment of the main body in the embryonic stage,and diversity of communication channels in the outbreak stage,In the period of mutation,there is diversity of transmission subjects,and in the period of extinction,there is latency of transmission heat;In the recurrent period,there is disorder in the dissemination of content;Finally,according to the different periods of the spread of online rumors,the corresponding governance methods are proposed: aiming at the characteristics of the concealment of the main body during the budding period of online rumors,it is proposed to carry out the online real-name system to enhance the self-restraint of netizens;According to the characteristics of the diversity of communication channels during the outbreak of online rumors,the following suggestions are proposed: 1.Carry out information hedging to mitigate the harm of rumors;2.Improve network information processing technology and accurately identify rumors;3.Improve the timeliness of information release and enhance the credibility of the government;Aiming at the characteristics of the diversity of subjects during the period of variation of online rumors,this paper proposes to enhance the media literacy of netizens and create a good network environment;According to the characteristics of the latent heat in the disappearance period of online rumors,the following suggestions are proposed: 1.Do a good job in popularizing science among the public and improve the ability to identify online rumors;2.Explore the public psychology behind online rumors and manage the source of rumors;In response to the characteristics of disorderly content during the recurrent period of online rumors,it is proposed to strengthen the governance of the platform and disclose the truth of online rumors;2.Accurately crack down on online rumors against drug rumor brands.