Fake News Research Articles

Epidemic modeling for misinformation spread in digital networks through a social intelligence approach

Online digital networks, including social networks, have significantly impacted individuals’ personal and professional lives. Aside from exchanging news and topics of interest, digital networks play an essential role in the diffusion of information, which frequently significantly impacts worldwide societies. In this paper, we present a new mathematical epidemic model for digital networks that considers the sentiment of solitary misinformation in the networks and characteristics of human intelligence that play an important role in judging and spreading misinformation inside the networks. Our mathematical analysis has proved the existence and validity of the system in a real-time environment. Considering the real-world data, our simulation predicts how the misinformation could spread among different global communities and when an intervention mechanism should have to be carried out by the policyholders. Our simulation using the model proves that effective intervention mechanisms by isolating the fake news can effectively control the spread of misinformation among larger populations. The model can analyze the emotional and social intelligence of groups frequently subjected to disinformation and disseminating fake news.

Ujaran Kebencian dan Berita Bohong Berdasarkan Perspektif Sosiologi Hukum Ditinjau dari Undang - Undang Informasi dan Transaksi Elektronik

The purpose of this study is to examine and analyse the regulation of hate speech from a sociological perspective with a focus on Law Number 1 of 2024 which is the second amendment to Law Number 11 of 2008 concerning Electronic Information and Transactions. This research uses the normative juridical method. The results show that fake news that is deliberately spread to damage someone's reputation is regulated in Article 28 paragraph 1 of the ITE Law which prohibits the dissemination of electronic information that is untrue or misleading and has an impact on material losses for consumers in electronic transactions. On the other hand, hate speech is regulated in Article 28 paragraph 2 which prohibits the dissemination of electronic information that incites or influences others to cause hatred or hostility towards individuals or groups based on attributes such as race, nationality, religion or gender. This includes insults, defamation, provocation and slander, all of which aim to create hostility among the public.

BERTGuard: Two-Tiered Multi-Domain Fake News Detection with Class Imbalance Mitigation

BERTGuard: Two-Tiered Multi-Domain Fake News Detection with Class Imbalance Mitigation

Media literacy tips promoting reliable news improve discernment and enhance trust in traditional media

Media literacy tips typically encourage people to be skeptical of the news despite the small prevalence of false news in Western democracies. Would such tips be effective if they promoted trust in true news instead? A pre-registered experiment (N = 3919, US) showed that Skepticism-enhancing tips, Trust-inducing tips, and a mix of both tips, increased participants’ sharing and accuracy discernment. The Trust-inducing tips boosted true news sharing and acceptance, the Skepticism-enhancing tips hindered false news sharing and acceptance, while the Mixed tips did both. Yet, the effects of the tips were more alike than different, with very similar effect sizes across conditions for true and false news. We experimentally manipulated the proportion of true and false news participants were exposed to. The Trust and Skepticism tips were most effective when participants were exposed to equal proportions of true and false news, while the Mixed tips were most effective when exposed to 75% of true news - the most realistic proportion. Moreover, the Trust-inducing tips increased trust in traditional media. Overall, we show that to be most effective, media literacy tips should aim both to foster skepticism towards false news and to promote trust in true news.

Strategies and bottlenecks to tackle infodemic in public health: a scoping review.

The World Health Organization defines "infodemic" as the phenomenon of an uncontrolled spread of information in digital and physical environments during a disease outbreak, causing confusion and risk-taking behaviors that can harm health. The aim of this scoping review is to examine international evidence and identify strategies and bottlenecks to tackle health-related fake news. We performed a scoping review of the literature from 1 January 2018 to 26 January 2023 on PubMed, Web of Science, and Scopus electronic databases. We also performed a search of grey literature on institutional websites. The research question has been defined according to the PCC (population, concept, and context) mnemonic for constructing research questions in scoping reviews. The overall research in the scientific databases yielded a total of 5,516 records. After removing duplicates, and screening the titles, abstracts, and full texts, we included 21 articles from scientific literature. Moreover, 5 documents were retrieved from institutional websites. Based on their content, we decided to group recommendations and bottlenecks into five different and well-defined areas of intervention, which we called strategies: "foster proper communication through the collaboration between science and social media companies and users," "institutional and regulatory interventions," "check and debunking," "increase health literacy," and "surveillance and monitoring through new digital tools." The multidisciplinary creation of standardized toolkits that collect recommendations from the literature and institutions can provide a valid solution to limit the infodemic, increasing the health education of both citizens and health professionals, providing the knowledge to recognize fake news, as well as supporting the creation and validation of AI tools aimed at prebunking and debunking.

SARD: Fake news detection based on CLIP contrastive learning and multimodal semantic alignment

The automatic detection of multimodal fake news can be used to effectively identify potential risks in cyberspace. Most of the existing multimodal fake news detection methods focus on fully exploiting textual and visual features in news content, thus neglecting the full utilization of news social context features that play an important role in improving fake news detection. To this end, we propose a new fake news detection method based on CLIP contrastive learning and multimodal semantic alignment (SARD). SARD leverages cutting-edge multimodal learning techniques, such as CLIP, and robust cross-modal contrastive learning methods to integrate features of news-oriented heterogeneous information networks (N-HIN) with multi-level textual and visual features into a unified framework for the first time. This framework not only achieves cross-modal alignment between deep textual and visual features but also considers cross-modal associations and semantic alignments across different modalities. Furthermore, SARD enhances fake news detection by aligning semantic features between news content and N-HIN features, an aspect largely overlooked by existing methods. We test and evaluate SARD on three real-world datasets. Experimental results demonstrate that SARD significantly outperforms the twelve state-of-the-art competitors in fake news detection, with an average improvement of 2.89% in Mac.F1 score and 2.13% in accuracy compared to the leading baseline models across three datasets.

Beliefs trump facts: Effect of Frequent Corrections on Misinformation Beliefs during Extended Extreme Events

Providing corrections to people who have engaged with false claims in Online Social Networks (OSN) is a form of cognitive intervention employed to address the spread of misinformation. Although there is a large body of work that has studied the effectiveness of corrections for promoting accurate beliefs, there is still much uncertainty around the precise effects of corrections on individuals’ behaviors in OSNs. Notably, the effect of offering frequent corrections on discerning information and identifying misinformation remains uncertain. We conducted two laboratory experiments to test whether experiencing frequent corrections to misinformation improved peoples’ ability to discriminate between true and false news claims during extended extreme events like the COVID-19 pandemic. All participants recruited for the experiments were from USA. They received corrections at varying frequencies, depending on their assigned experimental condition. Results from both experiments suggest that increasing frequency of corrections may not affect people’s ability to correctly assess information (or misinformation). Participant’s beliefs (vaccine hesitancy, belief in mask effectiveness, and trust in fact-checking organization) were found to be the most significant contributing factors to the ability to learn from corrections. We discuss the implications of the findings from these experiments.

Fake News Detection in the 2024 Indonesian General Election Using Bidirectional Long Short-Term Memory (BI-LSTM) Algorithm

The advancement of information technology provides convenience, but it also brings about problems. One area affected by this is the election process in Indonesia, which has seen a rise in fake news often used to discredit political opponents. Fake news misleads the public into believing incorrect information related to the election. To address this issue, a system is needed to detect fake news in the 2024 election to help the public differentiate between true and false information. This system is developed using an artificial intelligence and deep learning approach trained to do text classification on fake news detection. The training data consists of 1999 entries obtained from the Global Fact-Check Database from turnbackhoax.id, detik.com, and cnnindonesia.com. The machine learning model is built using the Bidirectional Long Short-Term Memory (BI-LSTM) algorithm, which is suitable for processing text data. This study compares two types of feature representations: TF-IDF and contextual embeddings with the IndoBERT model. The study results in the best model for text classification with an accuracy of 92% and a loss of 42.92%, achieved by the model using TF-IDF feature representation. The implementation of this system aims to enhance the integrity of the election process by minimizing the spread of misinformation. Future work will focus on refining the model and expanding the dataset to include more diverse sources for improved accuracy and robustness.

MMHFND: Fusing Modalities for Multimodal Multiclass Hindi Fake News Detection via Contrastive Learning

Multimodal content contains more deception than unimodal information, causing significant social and economic impacts. Current techniques often focus on single modality, neglecting knowledge fusion. While most studies have concentrated on English fake news detection, this study explores multi-modality for low-resource languages like Hindi. This work introduces the MMHFND model, based on M-CLIP, which uses late fusion for coarse (Fake vs Real) and fine-grained (World vs India vs Politics vs News vs Fact-Check) configurations. We extract deep representations from image and text using image transformer ResNet-50, a BERT-based L3cube-HindRoberta text transformer handling headlines, content, OCR text, and image captions, paired M-CLIP transformers, and an ELA (Error Level Analysis) image forensic method incorporating EfficientNet B0 to analyse multimodal news in Hindi language based on Devanagari script. M-CLIP integrates crossmodal similarity mapping of images and texts with retrieved multimodal features. The extracted features undergo redundancy reduction before being channelled into the final classifier. The MAM (Modality Attention Mechanism) is introduced, which generates weights for each modality individually. The MMHFND model uses a computed modality divergence score to identify dissonance between modalities and a modified contrastive loss on the score. We thoroughly analyse HinFakeNews dataset in a multimodal context, achieving accuracy in coarse and fine-grained configurations. We also undertake an ablation study to evaluate outcomes and explore alternative fusion processes on three different setups. The results show that the MMHFND model effectively detects fake news in Hindi with an accuracy of 0.986, outperforming other existing multimodal approaches.

Efficacy of Verification Observatories in Mitigating the Psychological Impact of Rumors Disseminated via social media During wars: Gaza War on Telegram in 2023 as a case study

This study investigates the effectiveness of information verification observatories in mitigating the psychological impacts of rumors and false news during the 2023 Gaza War. Utilizing a descriptive-analytical approach, the research encompasses the entire audience of information verification observatories, including (THAQAQ, KASHIF, TAYQAN), on the Telegram platform, totaling 2893 subscribers as of December 7, 2023. The sample size within this community comprised 351 individuals. Data collection was conducted through an electronic questionnaire tool. Findings reveal that 82.5% of respondents experienced positive psychological effects from information received via Telegram during the 2023 Gaza War. Moreover, 72% perceived that the performance of information verification observatories reduced the adverse psychological impacts of rumors circulated on Telegram concerning the war. Furthermore, 72% attributed the alleviation of their psychological distress to their trust in the information disseminated by these observatories via Telegram, as well as the promptness with which they verified false news and rumors related to the Gaza War. Additionally, 71.5% of respondents reported a reduction in the severity of negative psychological effects resulting from rumors, while 65.8% emphasized the positive impact of information verification observatories via Telegram on their psychological well-being.

Generated or Not Generated (GNG): The Importance of Background in the Detection of Fake Images

Facial biometrics are widely used to reliably and conveniently recognize people in photos, in videos, or from real-time webcam streams. It is therefore of fundamental importance to detect synthetic faces in images in order to reduce the vulnerability of biometrics-based security systems. Furthermore, manipulated images of faces can be intentionally shared on social media to spread fake news related to the targeted individual. This paper shows how fake face recognition models may mainly rely on the information contained in the background when dealing with generated faces, thus reducing their effectiveness. Specifically, a classifier is trained to separate fake images from real ones, using their representation in a latent space. Subsequently, the faces are segmented and the background removed, and the detection procedure is performed again, observing a significant drop in classification accuracy. Finally, an explainability tool (SHAP) is used to highlight the salient areas of the image, showing that the background and face contours crucially influence the classifier decision.

Exposure to Higher Rates of False News Erodes Media Trust and Fuels Overconfidence

ABSTRACT In two online experiments (N = 2,735), we investigated whether forced exposure to high proportions of false news could have deleterious effects by sowing confusion and fueling distrust in news. In a between-subjects design where U.S. participants rated the accuracy of true and false news, we manipulated the proportions of false news headlines participants were exposed to (17%, 33%, 50%, 66%, and 83%). We found that exposure to higher proportions of false news decreased trust in the news but did not affect participants’ perceived accuracy of news headlines. While higher proportions of false news had no effect on participants’ overall ability to discern between true and false news, they made participants more overconfident in their discernment ability. Therefore, exposure to false news may have deleterious effects not by increasing belief in falsehoods, but by fueling overconfidence and eroding trust in the news. Although we are only able to shed light on one causal pathway, from news environment to attitudes, this can help us better understand the effects of external or supply-side changes in news quality.

Accurate Fake News Prediction by Comparing Performance of Machine learning algorithms

Accurate Fake News Prediction by Comparing Performance of Machine learning algorithms

“Middle East conflict in Berlin schools”: on the affectability of “fake news”

Abstract The article revisits my older work on moral panics over “Muslim homophobia” and youth crime to make sense of the arrival of a new criminalized figure: the Palestinian school kid who becomes prone to terrorism and antisemitism after consuming “fake news” on social and Arabic-language media. It addresses the current outbreak of anti-Palestinian racism in Germany as the latest expansion of an archive of criminalization that targets young people who, far from innocent and deserving of care and education, are constructed as folk devils that require heavy-handed punishment, segregation and deportation. While former panics have singled out the problematic heterosexualities of Muslim populations that fail to catch up with women-and-LGBT-friendly communities that care about diversity, the debate about the “Middle East conflict in Berlin schools” foregrounds media use as a central determinant of cultural pathology. This departure from languages of care and diversity, the article argues, reflects a dehumanizing tendency in racial capitalism that appears increasingly undisguised.

Effects of Intervention Timing on Health-Related Fake News: Simulation Study.

Fake health-related news has spread rapidly through the internet, causing harm to individuals and society. Despite interventions, a fenbendazole scandal recently spread among patients with lung cancer in South Korea. It is crucial to intervene appropriately to prevent the spread of fake news. This study investigated the appropriate timing of interventions to minimize the side effects of fake news. A simulation was conducted using the susceptible-infected-recovered (SIR) model, which is a representative model of the virus spread mechanism. We applied this model to the fake news spread mechanism. The parameters were set similarly to those in the digital environment, where the fenbendazole scandal occurred. NetLogo, an agent-based model, was used as the analytical tool. Fake news lasted 278 days in the absence of interventions. As a result of adjusting and analyzing the timing of the intervention in response to the fenbendazole scandal, we found that faster intervention leads to a shorter duration of fake news (intervention at 54 days = fake news that lasted for 210 days; intervention at 16 days = fake news that lasted for 187 days; and intervention at 10 days = fake news that lasted for 157 days). However, no significant differences were observed when the intervention was performed within 10 days. Interventions implemented within 10 days were effective in reducing the duration of the spread of fake news. Our findings suggest that timely intervention is critical for preventing the spread of fake news in the digital environment. Additionally, a monitoring system that can detect fake news should be developed for a rapid response.

Improving fake job description detection using deep learning-based NLP techniques

ABSTRACT In the modern online society, where social networks are increasingly developed, the authenticity of information is one of the essential needs. There are many problems with detecting fake information, such as fake news detection, fake review detection, etc. Fake job description (FJD) detection is an interesting problem many groups have studied recently. However, current studies still need improvement in predictability. Therefore, this study develops a new method named NLP2FJD that utilizes deep learning-based NLP techniques for improving FJD detection. Firstly, we utilize the pre-trained Word2Vec to extract features from textual information from the dataset. Next, combining textual information and meta-information in the experimental dataset to improve the performance of FJD detection. The above two improvements will help the recommender system significantly improve the predictive ability of the proposed model. Finally, the empirical experiments are conducted to confirm the effectiveness of the proposed method on the experimental dataset compared with cutting-edge methods. The experimental results demonstrate that the NLP2FJD framework transcends other experimental methods for FJD detection on the experimental dataset. Besides, this study also conducts ROC curve analysis to show how to determine the optimal threshold for distinguishing the fake or real job description on the experimental dataset.

Enhanced Feature Representation for Multimodal Fake News Detection Using Localized Fine-Tuning of Improved BERT and VGG-19 Models

Enhanced Feature Representation for Multimodal Fake News Detection Using Localized Fine-Tuning of Improved BERT and VGG-19 Models

ALFABETIZAÇÃO MIDIÁTICA E FAKE NEWS: DESAFIOS E ESTRATÉGIAS NO ENSINO DE PORTUGUÊS PARA A FORMAÇÃO DE ALUNOS CRÍTICOS

In the current context, media literacy has emerged as an essential skill for the formation of critical and well-informed citizens. This article addresses the importance of integrating this skill into the school curriculum, especially in the teaching of Portuguese. Initially, the challenges faced by educators when dealing with the vast amount of information available and the proliferation of fake news are presented. The lack of criteria for assessing the veracity of information and the influence of social networks make students vulnerable to misinformation, highlighting the urgent need for effective media education. The article then explores various didactic strategies that can be implemented in Portuguese teaching to combat disinformation. These include developing critical thinking, analyzing information sources and encouraging the production of responsible digital content. Practical activities such as news debates, fact-checking and school media projects are suggested to engage students in an active and meaningful way. Finally, the formation of critical learners as a central objective of contemporary education is discussed. Media literacy is presented not only as a complement to traditional Portuguese teaching, but as an integral component that prepares students to face the challenges of the digital age. The article concludes that by enabling students to recognize and question false information, schools play a fundamental role in promoting a more critical and informed society.

The COVID-19 infodemic as a major driver for vaccine hesitancy in conflict settings: a mixed-method study

Since COVID-19 emerged in China, social media has spread fake news and conspiracy theories faster than the virus, causing the population to take false actions and follow misleading information, which leads to hospitalization, and even deaths, especially in conflict settings. Therefore, efforts to address health misinformation on social media have become urgent. For these reasons, we monitored social media following the vaccine rollout in Yemen and assessed the impact of the infodemic on the population's vaccine hesitancy in Yemen. To triangulate our data, we executed a mixed method approach in which qualitative data via online social listening was gathered using the WHO Early AI-supported Response with Social Listening (EARS) platform, and then complemented and verified with a Knowledge, Attitude and Practices (KAP) survey. Our results showed that social media in Yemen were exhibited contradictory facts and information gaps about the COVID-19 vaccine. The unaddressed questions and concerns on Facebook and Twitter have been filled with all types of misinformation, including rumors and conspiracy theories about the vaccine. Furthermore, the government's weak Risk Communication and Community Engagement (RCCE) have caused an enormous mistrust in its public health actions and stimulated negative behaviours towards the COVID-19 vaccine. The online social listening results were complemented with the survey findings as one third (38%) of the respondents admitted that COVID-19 vaccine misinformation affected their health decisions and almost two-third (61%) showed unwillingness to take the COVID-19 vaccine which clearly demonstrate the Yemenis hesitation in making health decisions including vaccine uptake. Our study highlights the importance of strengthening health communication and infodemic management to mitigate vaccine hesitancy and increase public confidence in the COVID-19 vaccine.

Membangun Kewaspadaan Publik: Sosialisasi Anti-Hoaks Pemilu Melalui Brosur Sebagai Taktik Edukasi Yang Efektif

In the 2024 General Election, Indonesians will use their voting rights to elect five political positions, namely the president and vice president, members of the DPR, members of the DPD, members of the provincial DPRD and members of the district / city DPRD in accordance with Law Number 7 Year 2017. The activity of building public awareness with anti-hoax socialization through brochures ahead of the election was held in Padangsambian Klod Village, West Denpasar District, Denpasar City, Bali Province. With the massive socialization of anti-hoax through brochures in Padangsambian Klod Village, it is hoped that the community will be more vigilant and not easily provoked by SARA (ethnicity, religion, race, and intergroup) issues or misleading information spread by irresponsible individuals. This community service activity uses the participatory Action Research (PAR) method. The data collection method used in this research is observation, brochure distribution, and documentation. intergroup) or misleading information spread by irresponsible people. The purpose of this community service activity is to increase public awareness of hoax news on social media related to elections. to increase public awareness of hoax news on social media related to elections. The author hopes that in the future the village community will not easily believe and spread information from sources that are not clear about the truth.