Misinformation Research Articles

COMPARATIVE ANALYSIS ON DEEP LEARNING METHODS FOR BOT DETECTION

Social bots are automated accounts on social media that are managed by software and overseen by people behind the scenes. While some bots are created for positive reasons, like sharing news updates or offering assistance in crises, others have been misused to spread false information, rumors, or influence political events. There are tools in place to identify and eliminate harmful bots automatically, but creators are continually evolving their techniques to evade detection. Hence, there is a pressing need for improved ways to differentiate between genuine and automated bot accounts. In recent years, several research studies have delved into the realm of social media bot detection, offering an extensive overview of different detection strategies, including advanced approaches such as machine learning (ML) and deep learning (DL) techniques. To the best of our knowledge, this study stands out as the first to solely focus on DL techniques, evaluating their efficacy and rationale in comparison to each other and to conventional ML methods. In this research, we offer a comprehensive classification of the characteristics utilized in deep learning research and provide information on the necessary preprocessing methods for creating appropriate training data for a deep learning model. We highlight the gaps identified in review papers discussing deep learning and machine learning studies, offering insights into future developments in this area. In general, deep learning methods have proven to be efficient in terms of computation and time for detecting social bots, demonstrating superior or comparable performance compared to traditional machine learning techniques. Key Words: Bot Detection, Social Media, Deep Learning, Computation, Automation

Multifaceted online coordinated behavior in the 2020 US presidential election

Organized attempts to manipulate public opinion during election run-ups have dominated online debates in the last few years. Such attempts require numerous accounts to act in coordination to exert influence. Yet, the ways in which coordinated behavior surfaces during major online political debates is still largely unclear. This study sheds light on coordinated behaviors that took place on Twitter (now X) during the 2020 US Presidential Election. Utilizing state-of-the-art network science methods, we detect and characterize the coordinated communities that participated in the debate. Our approach goes beyond previous analyses by proposing a multifaceted characterization of the coordinated communities that allows obtaining nuanced results. In particular, we uncover three main categories of coordinated users: (i) moderate groups genuinely interested in the electoral debate, (ii) conspiratorial groups that spread false information and divisive narratives, and (iii) foreign influence networks that either sought to tamper with the debate or that exploited it to publicize their own agendas. We also reveal a large use of automation by far-right foreign influence and conspiratorial communities. Conversely, left-leaning supporters were overall less coordinated and engaged primarily in harmless, factual communication. Our results also showed that Twitter was effective at thwarting the activity of some coordinated groups, while it failed on some other equally suspicious ones. Overall, this study advances the understanding of online human interactions and contributes new knowledge to mitigate cyber social threats.

LAW ENFORCEMENT AGAINST CRIMINAL ACTS OF FALSE INFORMATION THAT ENDANGER AVIATION SAFETY

Aviation safety is a problem that currently requires attention and has become a national and international issue. Because modes of transportation that have fast characteristics are becoming more and more crowded and thus the vulnerability to accidents is increasing. Recently, Indonesia has been one of the countries in the world whose aviation safety problems are very worrying related to criminal acts that endanger aviation safety. Several provisions have been issued by the government by adopting or converting international conventions to issuing specific laws and regulations, however, several regulations and deregulations of laws and regulations which carry criminal threats still contain overlap and are felt to not provide a sense of justice and legal uncertainty. This research aims to understand the regulation of criminal acts of false information that endanger aviation safety and the resolution process by researching and reviewing criminal threat norms regulated in several statutory regulations as well as the authorities in the resolution process. That this research was carried out using normative juridical methods is supported by field data.

Science-related populism referring to Europe: narratives of dominance and sovereignty

ABSTRACT In 2021–2022, the European Commission held the Conference on the Future of Europe, aiming to provide a platform for EU citizens to voice their visions for the Union's future. Simultaneously, concerns about the rise of populism and mis(dis)information have grown, significantly affecting the daily lives of Europeans and challenging democracy. This prompts the question: do cohesive populist visions for Europe exist? This article delves into a subset of populist narratives that express distrust in mainstream knowledge production frameworks and explores their connection to depictions of Europe through discourse analysis of 24 publicly available online documents opposing vaccination and 5G technology. It is argued that these narratives can be understood through science-related populism (Mede, N. G., & Schäfer, M. S. (2020). Science-related populism: Conceptualizing populist demands toward science. Public Understanding of Science, 29(5), 473–491). The analysed narratives consistently depict an antagonistic dichotomy between the noble people and corrupt elites, and they perform crisis in the face of the perceived dominance of illegitimate elites. However, both science and the EU do not unequivocally align with the contested elites. Their position in this dichotomous vision is influenced by the perceived morality of the scientists and a nationalist framework for interpreting Europe.

Deep Fake Detection System

Deep learning methods are used by the Deep Fake Detection System to recognize "deepfakes," or distorted media content. Deepfakes are artificial media produced by sophisticated artificial intelligence algorithms that threaten the credibility of media. The goal of our project is to create a reliable system that can discriminate between authentic and modified content in order to stop the spread of false information and protect media integrity. Our goal is to improve deepfake detection efficiency and accuracy by conducting a thorough evaluation of deep learning-based detection techniques. Our technology aims to offer real-time detection capabilities by utilizing sophisticated neural networks and machine learning techniques. This will aid in the continuous endeavors to tackle the widespread occurrence of deepfakes in digital media. Key Words: Deep Fake Detection, Deep Learning, Media Integrity, Misinformation, Neural Networks

The virtual reference radiologist: comprehensive AI assistance for clinical image reading and interpretation

ObjectivesLarge language models (LLMs) have shown potential in radiology, but their ability to aid radiologists in interpreting imaging studies remains unexplored. We investigated the effects of a state-of-the-art LLM (GPT-4) on the radiologists’ diagnostic workflow.Materials and methodsIn this retrospective study, six radiologists of different experience levels read 40 selected radiographic [n = 10], CT [n = 10], MRI [n = 10], and angiographic [n = 10] studies unassisted (session one) and assisted by GPT-4 (session two). Each imaging study was presented with demographic data, the chief complaint, and associated symptoms, and diagnoses were registered using an online survey tool. The impact of Artificial Intelligence (AI) on diagnostic accuracy, confidence, user experience, input prompts, and generated responses was assessed. False information was registered. Linear mixed-effect models were used to quantify the factors (fixed: experience, modality, AI assistance; random: radiologist) influencing diagnostic accuracy and confidence.ResultsWhen assessing if the correct diagnosis was among the top-3 differential diagnoses, diagnostic accuracy improved slightly from 181/240 (75.4%, unassisted) to 188/240 (78.3%, AI-assisted). Similar improvements were found when only the top differential diagnosis was considered. AI assistance was used in 77.5% of the readings. Three hundred nine prompts were generated, primarily involving differential diagnoses (59.1%) and imaging features of specific conditions (27.5%). Diagnostic confidence was significantly higher when readings were AI-assisted (p > 0.001). Twenty-three responses (7.4%) were classified as hallucinations, while two (0.6%) were misinterpretations.ConclusionIntegrating GPT-4 in the diagnostic process improved diagnostic accuracy slightly and diagnostic confidence significantly. Potentially harmful hallucinations and misinterpretations call for caution and highlight the need for further safeguarding measures.Clinical relevance statementUsing GPT-4 as a virtual assistant when reading images made six radiologists of different experience levels feel more confident and provide more accurate diagnoses; yet, GPT-4 gave factually incorrect and potentially harmful information in 7.4% of its responses.

Conflict Management Strategies used by Early Grade Teachers: A Qualitative Approach

This study presents the perceptions of teachers and head teachers toward the causes of conflict and effective strategies to resolve the conflict at preprimary schools. The research examined how conflicts are resolved by head teachers and teachers of preprimary schools. Primary data was collected through interviews with 6 head teachers and 20 teachers from purposively sampled preprimary schools. The results of the study revealed that conflicts are healthy for the institutes. Findings show that differences in beliefs between head teachers and teachers, management style, workload, dissatisfaction, lack of staff training, absence of reward systems for teachers, false information, lack of commitment, poor communication, lack of participation in decisions, and favoritism are the key causes that raised the conflicts between principal and teachers of preprimary schools. The research concludes that existing conflicts can be handled in preprimary schools by using the integrating, avoiding competing, compromising, and accommodating strategies. Keywords: conflict, heads of schools, preprimary schools

Differentiating Authentic News from Fabrications using Deep Learning: A New Approach

Fake news on social media disrupts perceptions and influences decision-making. Despite the importance of detection in politics, there's a lack of research on AI and ML-focused detection models. In this paper, we suggest a novel hybrid deep learning model (CNN-DNN) that combines the convolution neural network (CNN) and deep neural network (DNN). The proposed model was evaluated on the WELFake dataset, which consists of a total of 72,134 news articles. Out of the total number of articles, 35,028 are categorized as genuine news, while 37,106 are categorized as fabricated news. The proposed model achieved the highest accuracy at 0.9732, however, the LSTM model attained the lowest accuracy at 0.960. Our goal was to develop a detection model that would efficiently curb the dissemination of false information. The efficacy of this model, which depends on multiple data sources, has been demonstrated in the context of managerial decision-making. The study offers practical insights and indicates potential areas for future research.

AI in medical research

This review delves into the possible role of artificial intelligence (AI) in medical research, from planning to publication. AI can aid in idea generation, data analysis, and writing, with tools like chatbots and transcription systems enhancing efficiency. However, AI's limitations, including the "hallucination" problem in which it generates false information, require careful use and verification. Ensuring anonymity compliance with sensitive data is also vital. AI's transformative potential in research brings opportunities for innovation, necessitating mindful application to manage biases and data accuracy.

A Study on the Risks and Countermeasures of False Information Caused by AIGC

Generative artificial intelligence (AIGC) has changed the traditional information production mechanism and has a wide range of application scenarios. At the same time, the security risks it exposes such as data leakage, false content generation, and improper utilization have also attracted widespread attention from various countries. The development, application and governance of AIGC no longer seem to be a common challenge faced by one country but by the entire international community. In order to effectively respond to the challenges of AIGC to the false information governance system, this article uses multiple methods such as literature analysis and in-depth research to elaborate on the potential risks of AIGC, and conducts an in-depth analysis of the global challenges of false information risk governance. Finally, Proposing governance paths and countermeasures from various perspectives such as supervision and ecology provides intelligence reference for the healthy development of the AIGC industry.

Reality Check AI : Harnessing AI to Forecast and Unmask False Reporting

The internet has revolutionized the way people consume information, but it has also led to a rise in fake news, which is concerning because of the possible effects it may have on society. This study investigates whether it is possible to detect fake news only by looking at text using deep learning algorithms. The ability of three neural network architectures to identify false information on the internet is suggested and assessed: DistilBERT, Long Short-Term Memory networks (LSTMs), and Convolutional Neural Networks (CNNs). This dataset, called ISOT (In-Store Orders and Transactions), was first created for retail analytics but is now used as a standard for assessing false news detection algorithms. The goal of this research is to support continued initiatives to promote information integrity and fight false information

From Scientific Journals to Newspapers in Spain: Interest in Disinformation (2000–2023)

As disinformation has become a topic of conversation in the media in recent years, the theory of agenda setting is once again making its presence known. The aim of this research is to verify the degree of interest in disinformation by the media (mainstream press) and in academic writing (scientific communication journals) according to frequency and whether or not such disinformation can be observed in the field of science. The primary research has been carried out through quantitative content analysis of three Spanish newspapers (El País, Abc, El Mundo) and 32 Spanish scientific communication journals included in the SJR-SCImago Journal Rank database from the year 2000 to 2023. The results were 732 units of analysis. From those, it can be concluded that once again, the pandemic represents a before and after. Firstly, a general increase in disinformation has been observed, as well as a corresponding rise in false information in certain fields of science, especially that of health. Secondly, a gradual increase in public interest in disinformation has also been detected, which indicates that the issue is on the agenda of both the media and citizens.

Professional Values of Public Relations Specialists

The article is devoted professional values of public relations specialists in the context of the formation and management of public opinion in the context of current social challenges. The media space of media specialists is associated with certain professional risks. The high level of media activity of students increases the level of such risks. Negative consequences are associated with the possible occurrence of professional deformation, violation of a specialist’s psycho-emotional state of a (professional burnout, etc.), the destabilization of his system of value orientations. The analysis of theoretical and empirical material, including the results of the author’s research of future public relations specialists’ media literacy, indicates that there are certain concerns about the understanding and formation of professional value among future media specialists. Public relations specialists are directly involved in the formation and filling of the information space and its agenda. They have a significant impact on the mood in society and social reality, on the preferences and views of the target audience. The results of the study allow the authors to assert that in modern conditions it is necessary to create a system for the development of professional values of public relations specialists, taking into account the new risks of modern social reality. Such system should be aimed at creating a safe information space, protected from fake and deliberately false information, negative manipulation of public consciousness, destabilization of political, social, economic and other processes and the values of citizens.

Mindsets and politically motivated reasoning about fake news

False information may be published with the intention of misleading the public, and such fake news is often difficult to detect. Ideological fake news may pose a particular challenge, as people may be less able to detect false information that supports their prior beliefs. The difficulty of detecting fake news with an ideological slant may be compounded if people are motivated to defend their beliefs. Building on the mindset theory of action phases, we investigated motivational states as moderators of people’s ability to detect fake news. We tested two competing predictions to study the cognitive and motivational processes implicated in fake news detection. Both predictions concern an ideological belief bias, where people tend to accept information that agrees with their partisan identities and to reject information that disagrees with them. First, motivated reasoning accounts posit that deliberation should reinforce the ideological belief bias because reasoning primarily serves to defend and rationalize one’s own position. An opposing view, based on dual-process theory, assumes that deliberation attenuates the ideological belief bias by facilitating an unbiased assessment of new information. An online experiment (N = 497) tested these competing accounts. Participants were induced with deliberative/implemental/control mindsets prior to rating the veracity of (true/fake) news headlines. Some headlines favored a Republican view; others leaned toward a Democrat perspective. Based on self-reported political preference (Democrat vs. Republican), headlines were categorized as congruent or incongruent with participants’ political views. Consistent with an ideological belief bias, participants accepted more congruent than incongruent news, and they were more likely to fail to detect favorable fake news. In the main analysis, mindsets did not moderate the ideological belief bias, but showed interesting relationships with cognitive reflection and dishonest behavior. Further exploration using signal-detection theory suggested that the deliberative mindset might be associated with increased skepticism, thereby promoting fake news detection.

Persuading others in different communication media: appeals to logic, authority and emotion

PurposeScholars of persuasion have long made a distinction between appeals to logic, emotion and authority- logos, ethos and pathos- but ideas developed to account for live face-to-face conversation processes must also be tested in new media. We aimed to test the effectiveness of these three strategies in one-to-one chats through different communication media.Design/methodology/approachWith a 3 × 3 × 2 between-subject factorial design, we tested these three strategies in one-to-one chats (female–female or male–male pairs) through three communication media: face-to-face, Skype video or Skype text. The persuasion scenario was adapted from prior studies in which students were presented with the idea of requiring a comprehensive exam as part of their degree. The participants were all undergraduate students of a major university in USA.FindingsOur results showed trivial differences between female–female and male–male conditions. The logos appeal worked best overall in persuading the participants to change their reported attitudes. Additionally, the explanations provided by the participants for their own opinions were most like the persuasion scripts in the logos condition compared to the other two appeal conditions. Separately, participants indicated some disapproval of the pathos appeal in the text-based chat condition, although this did not seem to make a difference in terms of actual attitude change.Research limitations/implicationsOne major limitation of our study is that our subjects are college students and therefore are not representative of Internet users in general. Future research should test these three types of persuasion strategies on people of diverse backgrounds. For example, while logos seems to be most effective strategy in persuading college students (at least in our study), pathos or ethos may be more effective when one attempts to persuade people of different backgrounds.Practical implicationsAlthough it is enough for a statistical test, our sample size is still relatively small due to constraints on time, personnel and funding. We also recognize that it is challenging both conceptually and empirically to compare the effectiveness of three persuasion strategies separately.Social implicationsOur findings suggest it is helpful to use fact-checking tools to combat disinformation in cases where users may not have sufficient domain knowledge or may not realize the need to identify or examine the given information. Additionally, it may require more effort to negate the impact of the disinformation spread than correcting the information, as some users may not only believe false information but also may start to reason in ways similar to those presented in the disinformation messages.Originality/valuePast studies on online persuasion have limitedly examined whether and how communication media and persuasion strategies interact in one-to-one persuasion sessions. Our experiment makes an attempt to close this gap by examining the persuasion process and outcome in three different communication media and with three different persuasion strategies.

Fact or Fable: Your Guide to Fighting Misinformation

Nobody wants to believe false information and share it with others. Still, at some point, almost everyone will consider some fake news reports to be true, without even realizing it. This may happen because the information matches their world view but also because news producers use clever tricks to spread misinformation. In this article, we provide you with the tools to understand what misinformation is, why you may fall for it, and how to fight it. Given the ease with which fake news spreads on social media, it is important to know what you can do to resist its power.

Localization and Detection of Sinkhole Attacks in Wireless Sensor Networks Based on Denial of Service (DoS) Attacks

A wireless sensor network (WSN) uses uniformly deployed and positioned sensors to regularly send sensed data to a centralized station. Sinkhole attacks, in which a malicious node draws packets from other legitimate sensor devices and drops them, are the main danger to the WSN network layer and continue to be a difficult problem on wireless sensor networks. Security becomes a major issue since these networks have limited assets including less memory, less energy, and less transmission capacity. Additionally, given the dynamic context in which they are deployed, they are vulnerable to a variety of Denial of Service (DoS) assaults, including wormhole, sinkhole, and black hole attacks. The most dangerous routing attack at the network layer is known as a "sinkhole assault," which directs all network traffic onto a fake path by sending false information under the impression that it is the quickest one to the server. We explore sinkhole attacks to prevent them, and this work proposes a method of detection based on the redundancy process. Messages are transmitted across multiple paths to the suspicious nodes. Upon a thorough evaluation of the response, the attacked nodes are ultimately verified. Finally, a simulation is run to evaluate the method's efficacy. Furthermore, the simulation indicates that the strategy might be somewhat successful. Finally, it draws attention to difficulties and offers a prospective viewpoint for identifying sinkhole attacks.

Secure malicious node detection in flying ad-hoc networks using enhanced AODV algorithm

In wireless networking, the security of flying ad hoc networks (FANETs) is a major issue, and the use of drones is growing every day. A distributed network is created by a drone network in which nodes can enter and exit the network at any time. Because malicious nodes generate bogus identifiers, FANET is unstable. In this research study, we proposed a threat detection method for detecting malicious nodes in the network. The proposed method is found to be most effective compared to other methods. Malicious nodes fill the network with false information, thereby reducing network performance. The secure ad hoc on-demand distance vector (AODV) that has been suggested algorithm is used for detecting and isolating a malicious node in FANET. In addition, because temporary flying nodes are vulnerable to attacks, trust models based on direct or indirect reliability similar to trusted neighbors have been incorporated to overcome the vulnerability of malicious/selfish harassment. A node belonging to the malicious node class is disconnected from the network and is not used to forward or forward another message. The FANET security performance is measured by throughput, packet loss and routing overhead with the conventional algorithms of AODV (TAODV) and reliable AODV secure AODV power consumption decreased by 16.5%, efficiency increased by 7.4%, and packet delivery rate decreased by 9.1% when compared to the second ranking method. Reduced packet losses and routing expenses by 9.4%. In general, the results demonstrate that, in terms of energy consumption, throughput, delivered packet rate, the number of lost packets, and routing overhead, the proposed secure AODV algorithm performs better than the most recent, cutting-edge algorithms.

The effect of value on context and target recollection in memory for truth and falsity

Memory for truth and falsity has recently been investigated from the perspective of the dual-recollection theory, showing better context and target recollection for truth than falsity. In this paper, we examine whether these memory effects obtained for true statements are similar to the value effect, whereby true statements are given higher priority in encoding. For this purpose, we implemented value-directed remembering (VDR) into the conjoint-recognition paradigm. In our first experiment, the primary goal was to verify how VDR influences the processes defined by dual-recollection theory. At study, prioritized/important items were linked to higher numerical values (e.g., 10), while unimportant ones had lower values (e.g., 1). At test, the participants’ task was to recognize whether a particular sentence was important, unimportant, or new. We found that both context and target recollection were better for important items. In the second experiment, the main goal was to study the combined effects of importance and veracity on memory. In the between-subjects design, participants were monetarily rewarded for memorizing true or false sentences. The results demonstrated differences in the ability to prioritize truth over falsity. Specifically, we found a substantial increase in context recollection for prioritized true information but not for prioritized false information. Moreover, we found higher context recollection for true than false sentences in the true-prioritized condition, but not in the false-prioritized condition. These results indicated that people are able to prioritize true information better than false, and suggested that memory for truth may be a special case of the value effect.

Attitudes regarding vaccination on the example of COVID-19 prevention.

COVID-19 vaccines for public use were approved at the turn of 2020 and 2021. The level of vaccination coverage against COVID-19 in Poland is one of the lowest in Europe. Despite scientific studies there are people who are afraid of vaccinations and spread false information about them. Vaccinations against COVID-19 allowed to learn about the attitudes of people who decided to be vaccinated, which can be used to improve the effectiveness of the infectious diseases prevention in the work and public environment. The aim of the study was to analyze the main reasons determining a decision to get vaccinated against COVID-19. The study was carried on among all the people who decided to be vaccinated against COVID-19 in Nofer Institute of Occupational Medicine in Łódź. All vaccinated people completed an anonymous survey containing questions related to the most important factors, which encouraged them to undergo this vaccination. The study involved 945 people vaccinated against COVID-19. Statistical significance was revealed between 3 different reasons (one's own health, stress-free work, safe travelling and easier access to entertainment) for getting vaccinated and age, gender and a form of employment. The authors also found statistical significance of the correlation between the age, gender, education and form of employment of the respondents and the reason for choosing the specific vaccine product. Despite evidence presented by scientists about the effectiveness and safety of vaccinations, many people refused and still refuse to receive the vaccine. In order to increase the readiness of the population for vaccinations against COVID-19 and subsequent ones, if there is a need to limit new epidemics, the following should be done: introduce activities to raise health awareness, intensify strategies and campaigns aimed at increasing public trust in available vaccines against COVID-19, devolop individualized messages that contribute to vaccine hesitancy. Med Pr Work Health Saf. 2024;75(2):123-131.