Dynamic rumor spreading of public opinion reversal on Weibo based on a two-stage SPNR model
Introduction
Mobile Internet penetrates human daily lives as a convenient platform for acquiring and releasing information. People can acquire news about an event immediately after it transpires. Weibo, one of most important social media in China, attracts multiple netizens with its efficient sharing of text, pictures, voice, videos, and emojis. Weibo is apparently the main channel for social media news dissemination. The flexibility of sharing and forwarding blogs considerably accelerates the speed and scale of information fusion. However, the disseminated information is not always trustworthy. With the increasing speed and space of information spread, distinguishing the verity and faithfulness of facts [1] is difficult. This nature of social media provides a fertile ground for rumormongers to post and spread rumors that may result in major chaos and unpredictable reactions from involved individuals [2]. Understanding the dynamics of information spread can support the government to effectively publish actual information [3].
The release of official news regarding an incident tends to lag given the burstiness, uncertainty, and limited information. By comparison, a rumor is a story or statement whose truthfulness has not been verified [2], and thus always appears and spreads first on the Internet. Wrong, biased, and uncertain rumors often mislead the public or exert an adverse effect on social and public order. The rapid propagation of rumors through social media is a main factor in undermining social stability.
A real example of rumor reversal in a public event is the “the switch of the answer sheet in the college entrance examination” in Henan, China. Four parents questioned whether the answer sheets of test-takers were switched, which incited online public outrage. Although the rumor was debunked and reversed within several days, its spread in the early stages of the event exerted an unfortunate impact on the Department of Education and affected social stability. Such an adverse effect may have been avoided or minimized if the government has taken effective measures to control the rumor spread and guide the rumor reversal in the early stages. This example, including many similar real-life public events, shows that the government can pay more attention to the rumor spread and reversal.
This study explores the spread dynamics and reversal path of rumors in Sina Weibo regarding accidental events. We propose a new model that considers the impact of the hysteresis of official news and the “Follow the crowd” phenomenon on rumor development. The main contributions can be summarized as follows.
- •
We propose a new model called two-stage SPNR (S: Susceptible infected, : positive infected (people who do not know about the truth), : positive infected (people who know about the truth) N: negative infected, and R: Removal states) model to explore the development process of breaking news. Our model optimizes the SPNR model by considering the delay characteristics of official news released by authorities. To best simulate and predict the rumor spread and reversal, we divide this process into two stages.
- •
To better consider the effect of the official truth release on rumor spread, we classify the positive disseminators in the model into rational and truth disseminators based on their knowledge of the truth.
- •
We add the new transformation of rational (positive disseminators who do not know the truth) disseminators into rumor disseminators to consider the “follow the crowd” phenomenon. The probability of this transformation is positively correlated with the number of rumor disseminators, which is also in accordance with reality.
- •
The experimental results show that our proposed model can effectively simulate rumor spread and reversal, and helps in the understanding of rumor dynamics and then guiding its reversal.
Most current rumor propagation models are derived from the classic epidemic model [4]. A key to our understanding of epidemic theory are dynamic models [5], where the grouping of individuals in the same state can be regarded as a stock. The differential of stocks over time is the changing rate of stocks caused by state transition, which can be called a flow. Therefore, we create the model with system dynamics equations. Simultaneously, we use the classic simulation software, AnyLogic, to build and simulate the proposed two-stage SPNR model using real-life scenarios.
This paper is organized as follows. Section 2 presents a review or related literature. Section 3 describes the two-stage SPNR model, which considers the hysteresis of official news and “follow the crowd” phenomenon. Section 4 shows the model implementation using AnyLogic and simulation of opinion process in Weibo. Then the experimental results are compared with the real development process released by Section 5 discusses the experimental results and its implications, including its theoretical and practical significance. Section 6 concludes the paper, identifies research limitations, and highlights directions for future research.
Section snippets
Rumor spreading
Online social networks play an important role in spreading information on a large scale. Information diffusion occurs when information flows from one individual or community to another [6], and relevant models are widely used in research to understand this phenomenon. Related models can be divided into two categories, namely, explanatory and predictive. The former includes epidemics and influence models, whereas the latter includes independent cascade, linear threshold, game theory [6], SIS [7]
Proposed model: Two-stage SPNR model
Given that personality traits affect false rumor beliefs [27] and reposting negative information on microblogs [28], similar to the existing SPNR model [26], we divide rumor infection into two opposite states, positive-infected and negative-infected, according to whether people spread positive opinions or rumors, respectively. The difference is the addition of a secondary classification of positive public opinion disseminators, categorizing them as rational and truth reviewers on the basis of
Data description
The incident of the Chongqing bus plunging into a river is a typical case study of rumor reversal. At approximately 10 am on October 28, 2018, a bus plunged into a river in Wanzhou, Chongqing and a red car crashed on the scene. Shortly after the accident, several media, even mainstream, issued online reports that the female driver of the red car retrograded, causing the bus to plunge into the river. The news widely spread on Weibo and was forwarded by numerous Internet celebrity and the media.
Discussion
Analysis of the parameters affecting rumor spread lead to the following conclusions.
First, the low rumor-infected rate results in strong inhibition of rumor spread and the small maximum number of rumor disseminators. Apart from not believing rumors, not spreading rumors is equally important. Therefore, netizens not believing and not spreading rumors can better suppress its spread and maintain social stability.
Second, high probability of rumor disseminators transforming into truth disseminators
Conclusions and future research
With the Internet development, the information disseminated in online social networks is increasing more than in offline channels. Several characteristics of social media can facilitate posting of unverified information and the rapid diffusion among netizens across the world. In this study, we propose the two-stage SPNR based on the original model and consider the hysteresis of official news and the “follow the crowd” to examine the rumor spread and reversal about emergencies on Weibo. The
CRediT authorship contribution statement
Guoyin Jiang: Conceptualization, Methodology, Writing - review & editing. Saipeng Li: Programing, Writing - original draft. Minglei Li: Validation, Writing - review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgment
This work was partially supported by a grant from the National Natural Science Foundation of China (No. 71671060)
References (36)
- et al.
Detecting breaking news rumors of emerging topics in social media
Inf. Process. Manage.
(2020) - et al.
A statistical model of the impact of online rumors on the information quantity of online public opinion
Physica A
(2020) - et al.
SIR rumor spreading model in the new media age
Physica A
(2013) - et al.
SIHR rumor spreading model in social networks
Phys. A
(2012) - et al.
Rumor spreading model considering forgetting and remembering mechanisms in inhomogeneous networks
Physica A
(2013) - et al.
SIRaRu rumor spreading model in complex networks
Phys. A
(2014) - et al.
SICR rumor spreading model in complex networks: Counterattack and self-resistance
Phys. A
(2014) - et al.
2SI2R rumor spreading model in homogeneous networks
Phys. A
(2014) - et al.
Minimizing rumor influence in multiplex online social networks based on human individual and social behaviors
Inform. Sci.
(2020) - et al.
Stability analysis of an SDILR model based on rumor recurrence on social media
Phys. A
(2019)
How do users respond to online rumor rebuttals?
Comput. Hum. Behav.
Modeling of agent-based complex network under cyber-violence
Phys. A
Rumor spreading models with random denials
Physica A
Reposting negative information on microblogs: Do personality traits matter?
Inf. Process. Manage.
An analysis of rumor propagation based on propagation force
Physica A
Rumor spreading model considering hesitating mechanism in complex social networks
Physica A
Theory of rumour spreading in complex social networks
Physica A
Clicking position and user posting behavior in online review systems: A data-driven agent-based modeling approach
Inform. Sci.
Cited by (53)
A novel group multi-criteria sorting approach integrating social network analysis for ability assessment of health rumor-refutation accounts
2024, Expert Systems with ApplicationsEvolution of online public opinions on major accidents: Implications for post-accident response based on social media network
2024, Expert Systems with ApplicationsOnline attention dynamics: The triangle framework of theory, big data and simulations
2023, Expert Systems with ApplicationsDynamic modeling and simulation of rumor propagation based on the double refutation mechanism
2023, Information SciencesSimulating rumor spreading and rebuttal strategy with rebuttal forgetting: An agent-based modeling approach
2023, Physica A: Statistical Mechanics and its Applications