Factors Influencing The Communication Effect Of Rumor-Debunking Information On Social Media

Social media has gradually become an important channel for people to obtain and share information,thereby providing a fertile ground for the spread of rumors.The social characteristics of mass participation make rumors confusing to the public and trigger panic and anxiety,especially in crisis situations with many uncertainties,making rumors more destructive and negative.Therefore,to ensure healthy communication and stable development of the social order,effective improvement of the communication effect of rumor-debunking information has become topical for social communication managers and researchers.However,existing studies lack quantitative research exploring the factors influencing the communication effect of debunking information and the optimization of debunking strategies on Chinese social media.Consequently,based on Sina Weibo(or Chinese Microblog),this study uses deep learning,social network analysis,text mining,and econometric modeling methods that are based on the communicator,message,and recipient,three key elements that make up the communication process in the communication system model.The aim is to explore the factors influencing the communication effect of debunking information from three aspects: detecting debunking information and the characteristics of social networks,debunking methods,and rebuttal acceptance.Specifically,the following research contents were included.First,based on social network theory and deep learning methods,this study explored the influence of the social network characteristics of rumor refuters on their communication effect.Using 6 topics,176 events,and 49,278 postings of rumor data on Sina Weibo,deep learning algorithms were used to build a text classification model to identify debunking and non-debunking information,filter out rumor refuters and non-refuters,and investigate the differences in the rate of spread of debunking and non-debunking information between different rumor topics.This study used social network theory and analysis to construct three evaluation indices—the proportions of user types,size of the social network after k-core decomposition,and number of weakly connected components—to compare the social network characteristics of debunking and non-debunking information,and reveal the differences in the rate of spreading growth of debunking and non-debunking information between different rumor topics and the influence of the social network characteristics of the spreaders on their communication effect.The purpose of this study was to explore the social network characteristics of spreaders to improve their communication effect and provide references for more effective spread of debunking information.Second,using attitude change theory and deep learning methods,this study explored the influence of debunking methods on communication effect.A total of2053 original postings and 100,348 comments on 5 COVID-19 rumor events with the largest spread on Sina Weibo were selected.The study proposes a scheme for classifying debunking methods into six categories: denial,further fact-checking,refutation,personal response,organizational response,and combination methods.Deep learning algorithms were used to construct a text classification model for realizing the automatic recognition of audiences’ expressed stances in their comments on debunking posts.Based on the results of detecting audiences’ stances,this study used attitude change theory to construct a debunking effectiveness index to measure and compare the effectiveness of different debunking methods,with the aim of providing insights on how to effectively use debunking strategies to improve communication effect.Finally,based on the elaboration likelihood model of persuasion and source credibility theory,text mining and econometric methods were used to explore the influence of rebuttal acceptance on communication effect.Using COVID-19 rumor data,564,910 background posts from 3865 commenters were obtained three months before the outbreak of COVID-19.Employing the elaboration likelihood model and source credibility theory,this study constructed relationships between the central route(including information readability and argument quality),peripheral route(source credibility,including authority and influence),rebuttal acceptance,and the moderating effect of the receiver’s cognitive ability to explore factors affecting the degree of acceptance of debunking information.Text mining and econometric modeling methods were used to explore the factors affecting rebuttal acceptance of debunking information,revealing differences in rebuttal acceptance between audiences with different cognitive abilities,with the aim of providing support for optimizing the communication effect of debunking information from the perspective of analyzing the key elements of rebuttal acceptance.The findings of this dissertation thesis further extend the research field of social media rumor debunking and provide suggestions for improving the communication effect of debunking information,effective management of rumors,and management of the efforts at debunking information on Chinese social media platforms.