Detecting rumors in social media using emotion based deep learning approach.

Social media, an undeniable facet of the modern era, has become a primary pathway for disseminating information. Unverified and potentially harmful rumors can have detrimental effects on both society and individuals. Owing to the plethora of content generated, it is essential to assess its alignment with factual accuracy and determine its veracity. Previous research has explored various approaches, including feature engineering and deep learning techniques, that leverage propagation theory to identify rumors. In our study, we place significant importance on examining the emotional and sentimental aspects of tweets using deep learning approaches to improve our ability to detect rumors. Leveraging the findings from the previous analysis, we propose a Sentiment and EMotion driven TransformEr Classifier method (SEMTEC). Unlike the existing studies, our method leverages the extraction of emotion and sentiment tags alongside the assimilation of the content-based information from the textual modality, i.e., the main tweet. This meticulous semantic analysis allows us to measure the user's emotional state, leading to an impressive accuracy rate of 92% for rumor detection on the "PHEME" dataset. The validation is carried out on a novel dataset named "Twitter24". Furthermore, SEMTEC exceeds standard methods accuracy by around 2% on "Twitter24" dataset.

The eco-evolutionary dynamics of two strategic species: From the predator-prey to the innocent-spreader rumor model.

Species frequently engage in both competitive and cooperative interactions, delicately balancing these dynamics to optimize their chances of survival and reproduction. While competition drives individuals to compete for limited resources, cooperation can emerge as a strategic response, mitigating risk and enhancing collective payoff. To bridge theoretical game approaches such as payoff, cooperation, and defections in ecological systems, we propose a two-species predator-prey model inspired by the principles and variations of the prisoner's dilemma game. We comprehensively address and analytically verify all stable strategic states, exploring the role of payoff parameters both individually and collectively. Additionally, we investigate the effect of free space. Beyond ecological contexts, we present a model of rumor propagation within a social system to establish connections with the prisoner's dilemma game. In both systems, our primary focus is to discuss strategies and enhance the cooperative factor within the system, given its crucial importance across diverse environments.

SAMGAT: structure-aware multilevel graph attention networks for automatic rumor detection.

The rapid dissemination of unverified information through social platforms like Twitter poses considerable dangers to societal stability. Identifying real versus fake claims is challenging, and previous work on rumor detection methods often fails to effectively capture propagation structure features. These methods also often overlook the presence of comments irrelevant to the discussion topic of the source post. To address this, we introduce a novel approach: the Structure-Aware Multilevel Graph Attention Network (SAMGAT) for rumor classification. SAMGAT employs a dynamic attention mechanism that blends GATv2 and dot-product attention to capture the contextual relationships between posts, allowing for varying attention scores based on the stance of the central node. The model incorporates a structure-aware attention mechanism that learns attention weights that can indicate the existence of edges, effectively reflecting the propagation structure of rumors. Moreover, SAMGAT incorporates a top-k attention filtering mechanism to select the most relevant neighboring nodes, enhancing its ability to focus on the key structural features of rumor propagation. Furthermore, SAMGAT includes a claim-guided attention pooling mechanism with a thresholding step to focus on the most informative posts when constructing the event representation. Experimental results on benchmark datasets demonstrate that SAMGAT outperforms state-of-the-art methods in identifying rumors and improves the effectiveness of early rumor detection.

Factors influencing college students' online rumor refuting behavior during major public health crises: the moderating effect of group norms.

This study integrates SOR (Stimuli-Organism-Response) theoretical framework and rational behavior theory within a theoretical framework, incorporating group norms as a moderating factor to investigate the psychological mechanisms influencing Chinese college students' online rumor-refutation behavior amidst public health crises. Using the structural equation modeling research method, data was collected via questionnaires from 1,254 participants in the context of the COVID-19 pandemic. The findings indicate that both online and offline information seeking are positively correlated with college students' attitudes and subjective norms. Moreover, the attitudes and subjective norms of college students are positively correlated with the online rumor refuting behavior. Furthermore, group norms serve to strengthen the connection between college students' attitudes and their engagement in online refuting rumors. These results illuminate the psychological underpinnings driving college students' online rumor-refuting actions, offering practical and policy implications for effectively managing rumor behaviors.

The dynamics and control of a multi-lingual rumor propagation model with non-smooth inhibition mechanism.

In this paper, the dynamic behaviors and control strategies of a rumor propagation model are studied in multi-lingual environment. First, an S2E2I2R rumor propagation model is proposed, which incorporates a non-smooth inhibition mechanism. Meanwhile, the existence and stability of the equilibrium are analyzed, grounded in the spreader threshold of the government intervention. Finally, the optimal control and the event-triggered impulsive control strategies are proposed to mitigate the spread of rumors, and the comparison of their effectiveness is further presented by the numerical simulation and a practical case.

Corrigendum: How can the collaborative participation of regulators, whistleblowers, and parties effectively promote rumor management in public health emergencies?

[This corrects the article DOI: 10.3389/fpubh.2023.1290841.].

A synergistic study of policy and literature on rumor governance.

Rumor governance is an important guarantee for social stability and public safety. Based on the life cycle and crisis cycle model, this paper conducts a synergistic analysis of China's rumor governance policies and regulations and the core scientific research literature on rumor governance in WOS and CNKI. In this paper, we use the TF-IDF algorithm to count the word frequencies of 326 policy and regulation texts, the Jieba-RoBERTa-Kmeans model to cluster high-frequency keywords, and CiteSpace software and the LLR clustering algorithm are utilized to extract and cluster keywords from 391 documents in the WOS database and from 703 documents in the CNKI database. Based on the synergistic analysis of the life cycle model, it is found that the research on policies and regulations precedes the research on literature, and both are in the period of refinement.Based on the synergistic analysis using the co-occurrence comparison of subject terms in the crisis cycle model, it is found that there is a lack of research in the stages of prevention, monitoring, and governance, and this paper proposes the systematic governance mechanism and strategy for crisis resolution that conforms to the trend of life cycle evolution and is synergistic with policy and literature. This study has only selected Chinese policies and regulations, and the proposed governance strategies have not yet been verified in practice; future research can expand the scope and depth of the study and conduct empirical research and pilot projects.

Research on domain ontology construction based on the content features of online rumors.

Online rumors are widespread and difficult to identify, which bring serious harm to society and individuals. To effectively detect and govern online rumors, it is necessary to conduct in-depth semantic analysis and understand the content features of rumors. This paper proposes a TFI domain ontology construction method, which aims to achieve semantic parsing and reasoning of the rumor text content. This paper starts from the term layer, the frame layer, and the instance layer, and based on the reuse of the top-level ontology, the extraction of core literature content features, and the discovery of new concepts in the real corpus, obtains the core classes (five parent classes and 88 subclasses) of the rumor domain ontology and defines their concept hierarchy. Object properties and data properties are designed to describe relationships between entities or their features, and the instance layer is created according to the real rumor datasets. OWL language is used to encode the ontology, Protégé is used to visualize it, and SWRL rules and pellet reasoner are used to mine and verify implicit knowledge of the ontology, and judge the category of rumor text. This paper constructs a rumor domain ontology with high consistency and reliability.

Survival analysis of the duration of rumors during the COVID-19 pandemic.

BACKGROUND: The emergence of the COVID-19 pandemic towards the end of 2019 triggered a relentless spread of online misinformation, which significantly impacted societal stability, public perception, and the effectiveness of measures to prevent and control the epidemic. Understanding the complex dynamics and characteristics that determine the duration of rumors is crucial for their effective management. In response to this urgent requirement, our study takes survival analysis method to analyze COVID-19 rumors comprehensively and rigorously. Our primary aim is to clarify the distribution patterns and key determinants of their persistence. Through this exploration, we aim to contribute to the development of robust rumor management strategies, thereby reducing the adverse effects of misinformation during the ongoing pandemic. METHODS: The dataset utilized in this research was sourced from Tencent's "Jiao Zhen" Verification Platform's "Real-Time Debunking of Novel Coronavirus Pneumonia" system. We gathered a total of 754 instances of rumors from January 18, 2020, to January 17, 2023. The duration of each rumor was ascertained using the Baidu search engine. To analyze these rumors, survival analysis techniques were applied. The study focused on examining various factors that might influence the rumors' longevity, including the theme of the content, emotional appeal, the credibility of the source, and the mode of presentation. RESULTS: Our study's results indicate that a rumor's lifecycle post-emergence typically progresses through three distinct phases: an initial rapid decline phase (0-25 days), followed by a stable phase (25-1000 days), and ultimately, an extinction phase (beyond 1000 days). It is observed that half of the rumors fade within the first 25 days, with an average duration of approximately 260.15 days. When compared to the baseline category of prevention and treatment rumors, the risk of dissipation is markedly higher in other categories: policy measures rumors are 3.58 times more likely to perish, virus information rumors have a 0.52 times higher risk, epidemic situation rumors are 4.86 times more likely to die out, and social current affairs rumors face a 2.02 times increased risk. Additionally, in comparison to wish rumors, bogie rumors and aggression rumors have 0.26 and 0.27 times higher risks of dying, respectively. In terms of presentation, graphical and video rumors share similar dissolution risks, whereas textual rumors tend to have a longer survival time. Interestingly, the credibility of the rumor's source does not significantly impact its longevity. CONCLUSION: The survival time of rumors is strongly linked to their content theme and emotional appeal, whereas the credibility of the source and the format of presentation have a more auxiliary influence. This study recommends that government agencies should adopt specific strategies to counter rumors. Experts and scholars are encouraged to take an active role in spreading health knowledge. It's important for the public to proactively seek trustworthy sources for accurate information. Media platforms are advised to maintain journalistic integrity, verify the accuracy of information, and guide the public towards improved media literacy. These actions, collectively, can foster a collaborative alliance between the government and the media, effectively combating misinformation.

A Serious Game ("Fight With Virus") for Preventing COVID-19 Health Rumors: Development and Experimental Study.

BACKGROUND: Health rumors arbitrarily spread in mainstream social media on the internet. Health rumors emerged in China during the outbreak of COVID-19 in early 2020. Many midelders/elders (age over 40 years) who lived in Wuhan believed these rumors. OBJECTIVE: This study focused on designing a serious game as an experimental program to prevent and control health rumors. The focus of the study was explicitly on the context of the social networking service for midelders/elders. METHODS: This research involved 2 major parts: adopting the Transmission Control Protocol model for games and then, based on the model, designing a game named "Fight With Virus" as an experimental platform and developing a cognitive questionnaire with a 5-point Likert scale. The relevant variables for this experimental study were defined, and 10 hypotheses were proposed and tested with an empirical study. In total, 200 participants were selected for the experiments. By collecting relevant data in the experiments, we conducted statistical observations and comparative analysis to test whether the experimental hypotheses could be proved. RESULTS: We noted that compared to traditional media, serious games are more capable of inspiring interest in research participants toward their understanding of the knowledge and learning of health commonsense. In judging and recognizing the COVID-19 health rumor, the test group that used game education had a stronger ability regarding identification of the rumor and a higher accuracy rate of identification. Results showed that the more educated midelders/elders are, the more effective they are at using serious games. CONCLUSIONS: Compared to traditional media, serious games can effectively improve midelders'/elders' cognitive abilities while they face a health rumor. The gameplay effect is related to the individual's age and educational background, while income and gender have no impact.

BiMGCL: rumor detection via bi-directional multi-level graph contrastive learning.

The rapid development of large language models has significantly reduced the cost of producing rumors, which brings a tremendous challenge to the authenticity of content on social media. Therefore, it has become crucially important to identify and detect rumors. Existing deep learning methods usually require a large amount of labeled data, which leads to poor robustness in dealing with different types of rumor events. In addition, they neglect to fully utilize the structural information of rumors, resulting in a need to improve their identification and detection performance. In this article, we propose a new rumor detection framework based on bi-directional multi-level graph contrastive learning, BiMGCL, which models each rumor propagation structure as bi-directional graphs and performs self-supervised contrastive learning based on node-level and graph-level instances. In particular, BiMGCL models the structure of each rumor event with fine-grained bidirectional graphs that effectively consider the bi-directional structural characteristics of rumor propagation and dispersion. Moreover, BiMGCL designs three types of interpretable bi-directional graph data augmentation strategies and adopts both node-level and graph-level contrastive learning to capture the propagation characteristics of rumor events. Experimental results on real datasets demonstrate that our proposed BiMGCL achieves superior detection performance compared against the state-of-the-art rumor detection methods.

Stability Analysis of a Delayed Rumor Propagation Model with Nonlinear Incidence Incorporating Impulsive Vaccination.

The presence of information asymmetry can hinder the public's ability to make well-informed decisions, resulting in unwarranted suspicion and the widespread dissemination of rumors. Therefore, it is crucial to provide individuals with consistent and dependable scientific education. Regular popular science education is considered a periodic impulsive intervention to mitigate the impact of information asymmetry and promote a more informed and discerning public. Drawing on these findings, this paper proposes a susceptible-hesitant-infected-refuting-recovered (SHIDR) rumor-spreading model to explain the spread of rumors. The model incorporates elements such as time delay, nonlinear incidence, and refuting individuals. Firstly, by applying the comparison theorem of an impulsive differential equation, we calculate two thresholds for rumor propagation. Additionally, we analyze the conditions of global attractiveness of the rumor-free periodic solution. Furthermore, we consider the condition for the rumor's permanence. Finally, numerical simulations are conducted to validate the accuracy of our findings. The results suggest that increasing the proportion of impulsive vaccination, reducing the impulsive period, or prolonging the delay time can effectively suppress rumors.

Research on the Dynamical Behavior of Public Opinion Triggered by Rumor Based on a Nonlinear Oscillator Model.

In public opinion triggered by rumors, the authenticity of the information remains uncertain, and the main topic oscillates between diverse opinions. In this paper, a nonlinear oscillator model is proposed to demonstrate the public opinion triggered by rumors. Based on the model and actual data of one case, it is found that a continuous flow of new information about rumors acts as external forces on the system, probably leading to the chaotic behavior of public opinion. Moreover, similar features are observed in three other cases, and the same model is also applicable to these cases. Based on these results, it is shown that our model possesses generality, revealing the evolutionary trends of a certain type of public opinion in real-world scenarios.

Influence of Tweets Indicating False Rumors on COVID-19 Vaccination: Case Study.

BACKGROUND: As of December 2022, the outbreak of COVID-19 showed no sign of abating, continuing to impact people's lives, livelihoods, economies, and more. Vaccination is an effective way to achieve mass immunity. However, in places such as Japan, where vaccination is voluntary, there are people who choose not to receive the vaccine, even if an effective vaccine is offered. To promote vaccination, it is necessary to clarify what kind of information on social media can influence attitudes toward vaccines. OBJECTIVE: False rumors and counterrumors are often posted and spread in large numbers on social media, especially during emergencies. In this paper, we regard tweets that contain questions or point out errors in information as counterrumors. We analyze counterrumors tweets related to the COVID-19 vaccine on Twitter. We aimed to answer the following questions: (1) what kinds of COVID-19 vaccine-related counterrumors were posted on Twitter, and (2) are the posted counterrumors related to social conditions such as vaccination status? METHODS: We use the following data sets: (1) counterrumors automatically collected by the "rumor cloud" (18,593 tweets); and (2) the number of COVID-19 vaccine inoculators from September 27, 2021, to August 15, 2022, published on the Prime Minister's Office's website. First, we classified the contents contained in counterrumors. Second, we counted the number of COVID-19 vaccine-related counterrumors from data set 1. Then, we examined the cross-correlation coefficients between the numbers of data sets 1 and 2. Through this verification, we examined the correlation coefficients for the following three periods: (1) the same period of data; (2) the case where the occurrence of the suggestion of counterrumors precedes the vaccination (negative time lag); and (3) the case where the vaccination precedes the occurrence of counterrumors (positive time lag). The data period used for the validation was from October 4, 2021, to April 18, 2022. RESULTS: Our classification results showed that most counterrumors about the COVID-19 vaccine were negative. Moreover, the correlation coefficients between the number of counterrumors and vaccine inoculators showed significant and strong positive correlations. The correlation coefficient was over 0.7 at -8, -7, and -1 weeks of lag. Results suggest that the number of vaccine inoculators tended to increase with an increase in the number of counterrumors. Significant correlation coefficients of 0.5 to 0.6 were observed for lags of 1 week or more and 2 weeks or more. This implies that an increase in vaccine inoculators increases the number of counterrumors. These results suggest that the increase in the number of counterrumors may have been a factor in inducing vaccination behavior. CONCLUSIONS: Using quantitative data, we were able to reveal how counterrumors influence the vaccination status of the COVID-19 vaccine. We think that our findings would be a foundation for considering countermeasures of vaccination.

Dynamic selection of clarification channels in rumor propagation containment.

Rumors refer to spontaneously formed false stories. As rumors have shown severe threats to human society, it is significant to curb rumor propagation. Rumor clarification is an effective countermeasure on controlling rumor propagation. In this process, anti-rumor messages can be published through multiple media channels, including but not limited to online social platforms, TV programs and offline face-to-face campaigns. As the efficiency and cost of releasing anti-rumor information can vary from media channel to media channel, provided that the total budget is limited and fixed, it is valuable to investigate how to periodically select a combination of media channels to publish anti-rumor information so as to maximize the efficiency (i.e., make as many individuals as possible know the anti-rumor information) with the lowest cost. We refer to this issue as the dynamic channel selection (DCS) problem and any solution as a DCS strategy. To address the DCS problem, our contributions are as follows. First, we propose a rumor propagation model to characterize the influences of DCS strategies on curbing rumors. On this basis, we establish a trade-off model to evaluate DCS strategies and reduce the DCS problem to a mathematical optimization model called the DCS model. Second, based on the genetic algorithm framework, we develop a numerical method called the DCS algorithm to solve the DCS model. Third, we perform a series of numerical experiments to verify the performance of the DCS algorithm. Results show that the DCS algorithm can efficiently yield a satisfactory DCS strategy.

Stochastic Stabilization of Dual-Layer Rumor Propagation Model with Multiple Channels and Rumor-Detection Mechanism.

With the development of information technology, individuals are able to receive rumor information through various channels and subsequently act based on their own perceptions. The significance of the disparity between media and individual cognition in the propagation of rumors cannot be underestimated. In this paper, we establish a dual-layer rumor propagation model considering the differences in individual cognition to study the propagation behavior of rumors in multiple channels. Firstly, we obtain the threshold for rumor disappearance or persistence by solving the equilibrium points and their stability. The threshold is related to the number of media outlets and the number of rumor debunkers. Moreover, we have innovatively designed a class of non-periodic intermittent noise stabilization methods to suppress rumor propagation. This method can effectively control rumor propagation based on a flexible control scheme, and we provide specific expressions for the control intensity. Finally, we have validated the accuracy of the theoretical proofs through experimental simulations.

Understanding the role of social media usage and health self-efficacy in the processing of COVID-19 rumors: A SOR perspective.

Apart from the direct health and behavioral influence of the COVID-19 pandemic itself, COVID-19 rumors as an infodemic enormously amplified public anxiety and cause serious outcomes. Although factors influencing such rumors propagation have been widely studied by previous studies, the role of spatial factors (e.g., proximity to the pandemic) on individuals' response regarding COVID-19 rumors remain largely unexplored. Accordingly, this study, drawing on the stimulus-organism-response (SOR) framework, examined how proximity to the pandemic (stimulus) influences anxiety (organism), which in turn determines rumor beliefs and rumor outcomes (response). Further, the contingent role of social media usage and health self-efficacy were tested. The research model was tested using 1246 samples via an online survey during the COVID-19 pandemic in China. The results indicate that: (1)The proximity closer the public is to the pandemic, the higher their perceived anxiety; (2) Anxiety increases rumor beliefs, which is further positively associated rumor outcomes; (3) When the level of social media usage is high, the relationship between proximity to the pandemic and anxiety is strengthened; (4) When the level of health self-efficacy is high, the effect of anxiety on rumor beliefs is strengthened and the effect of rumor beliefs on rumor outcomes is also strengthened. This study provides a better understanding of the underlying mechanism of the propagation of COVID-19 rumors from a SOR perspective. Additionally, this paper is one of the first that proposes and empirically verifies the contingent role of social media usage and health self-efficacy on the SOR framework. The findings of study can assist the pandemic prevention department in to efficiently manage rumors with the aim of alleviating public anxiety and avoiding negative outcomes cause by rumors.

A Rumor Detection Model Incorporating Propagation Path Contextual Semantics and User Information.

Currently, social media is full of rumors. To stop rumors from spreading further, rumor detection has received increasing attention. Recent rumor detection methods treat all propagation paths and all nodes on the paths as equally important, resulting in models that fail to extract the key features. In addition, most methods ignore user features, leading to limitations in the performance improvement of rumor detection. To address these problems, we propose a Dual-Attention Network model on propagation Tree structures named DAN-Tree, where a node-and-path dual-attention mechanism is designed to organically fuse deep structure and semantic information on the propagation structures of rumors, and path oversampling and structural embedding are employed to enhance the learning of deep structures. Finally, we deeply integrate user profiles into the propagation trees in DAN-Tree, thus proposing the DAN-Tree++ model to further improve performance. Empirical studies on four rumor datasets have shown that DAN-Tree outperforms the state-of-the-art rumor detection models learning on propagation structures, and the results on two datasets with user information validate the superior performance of DAN-Tree++ over other models using both user profiles and propagation structures. What's more, DAN-Tree, especially DAN-Tree++, has achieved the best performance on early detection tasks.

Exploring rumor behavior during the COVID-19 pandemic through an information processing perspective: The moderating role of critical thinking.

In the midst of the pervasive disruption caused by the proliferation of rumors, it is unclear how individuals react to such information. Guided by the SOR theory (Stimuli-Organism-Response), our study investigates the association between different information sources (stimuli), emotions experienced by individuals (organism), and resulting rumor behaviors such as sharing and refuting (response). Furthermore, we examine the moderating role of individual critical thinking in this process. Using the COVID-19 pandemic as a study scenario, we collected questionnaire data from 4588 respondents. Our results reveal a large positive association between pandemic-related information and feelings of fear. Additionally, a medium negative correlation between fear and rumor sharing was observed while a moderate positive correlation between fear and rumor refuting was identified. Moreover, we found that individual critical thinking abilities can effectively moderate the relationship between fear and online COVID-19-related information and strengthen the link between fear and rumor sharing while weakening the link between fear and rumor refuting. Additionally, our study indicates that an individual's fear plays a mediating role in the relationship between information sources and rumor behavior. Our findings shed light on the information processing mechanisms underlying rumor behaviors and yield practical and policy implications for managing them.

A survey on rumor detection and prevention in social media using deep learning.

In the current digital era, massive amounts of unreliable, purposefully misleading material, such as texts and images, are being shared widely on various web platforms to deceive the reader. Most of us use social media sites to exchange or obtain information. This opens a lot of space for false information, like fake news, rumors, etc., to spread that could harm a society's social fabric, a person's reputation, or the legitimacy of a whole country. Therefore, preventing the transmission of such dangerous material across platforms is a digital priority. However, the main goal of this survey paper is to thoroughly examine several current state-of-the-art research works on rumor control (detection and prevention) that use deep learning-based techniques and to identify major distinctions between these research efforts. The comparison results are intended to identify research gaps and challenges for rumor detection, tracking, and combating. This survey of the literature makes a significant contribution by highlighting several cutting-edge deep learning-based models for rumor detection in social media and critically evaluating their effectiveness on recently available standard datasets. Furthermore, to have a thorough grasp of rumor prevention to spread, we also looked into various pertinent approaches, including rumor veracity classification, stance classification, tracking, and combating. We also have created a summary of recent datasets with all the necessary information and analysis. Finally, as part of this survey, we have identified some of the potential research gaps and challenges that need to be addressed in order to develop early, effective methods of rumor control.