In the Web 2.0 age, mass media disseminates the disinformation of companies and exerts considerable influence. How to manage this trend in a timely and effective fashion in this big data era has become difficult. In this study, we delve into this issue by trying to identify the core disseminators in the dissemination process. We propose the concept of a disinformation channel and quantitatively analyse these company-related disinformation channels among media outlets. By empirically analysing 4,689 disinformation news values and 330 channels in 2018, we reveal that the disinformation values and negative news values are characteristics. We also build automatic identification models to identify these channels from the media combined with machine learning algorithms. Our study sheds light on disinformation, thus providing managers with an empirical basis upon which to analyse the media and help them address the disinformation problem.

Suicide events may have a negative impact on all of society. The media plays a significant role in suicide prevention. Therefore, the aims of this study are (a) to understand the association between characteristics of suicide events and characteristics of who committed suicide, and event impact indexes (EIIs) of suicide reported on the internet; (b) to analyze violation of recommendations for reporting suicide by Weibo, and (c) to investigate the effect of online reports of suicide on public opinion. We carried out a content analysis of online reports of suicide. This study analyzed 113 suicide events, 300 news reports of suicide, and 2,654 Weibo comments about suicide collected from the WeiboReach between 2015 and 2020. We used a t-test and analysis of variance (ANOVA) to explore the potential factors associated with the EIIs of suicide events. The results found that (a) The suicide events reported on the internet during COVID-19 and those related to celebrities and students tend to have higher EIIs; (b) suicide reports on Weibo frequently violated WHO recommendations for suicide reporting in the media; and (c) public opinion of suicide reporting in the online media was mostly emotional and irrational, which is not beneficial for public mental health and suicide prevention. In conclusion, first, the situation of many people working from home or studying from home and spreading more time online during COVID-19 may lead to suicide events obtain more public attention. Online media could further improve public responsible reporting and daily media-content surveillance, especially taking particular care in those suicide events during COVID-19, and related to celebrities and students, which may have a higher event impact on the internet. Second, health managers should regular assessment of observance of the WHO recommendations for suicide reporting by online social media to prevent suicide. Third, health communication managers should use big data to identify, assess, and manage harmful information about suicide; and track anyone affected by suicide-related reports on social media to reduce the negative impact of public opinion to intervene suicide in the early stage of suicide.

When a major public health incident breaks out, in order to prevent the explosion of multiple types of public opinion, relevant government departments need to guide the online public opinion according to the needs and characteristics of different audiences in order to achieve reasonable regulation and control. In this process, gender differences among the participating public in areas such as comprehension ability often affect the effectiveness of government guidance. A proper understanding of these differences will enable the government to allocate resources on the basis of needs to save resources and achieve the same goals with half the effort. This paper takes the outbreak of the COVID-19 as an example to analyze the gender differences among users in terms of the overall volume of participation and specific participation behaviors from the dimension of time and geographical locations. A total of 735,271 comments posted by users in responding to tweets published by 144 official government accounts on Weibo during the COVID-19 outbreak were collected and analyzed with a combination of the methods of natural language processing and propensity score analysis. The results show that in comparison to male users, female users participated more, and their responses were more emotionally expressive. Female users tended to respond faster than male users by 30 minutes to an hour, which allowed female users to play a more important role in the process of government guidance of public opinion during major public health incidents. Therefore, this study further provides policy recommendations for the government to provide reasonable guidance of public opinion and give future direction.