Unveiling the truth: A systematic review of fact-checking and fake news research in social sciences

Online Journal of Communication and Media Technologies, Journal Year: 2024, Volume and Issue: 14(2), P. e202427 - e202427

Published: April 10, 2024

The current media ecosystem, marked by immediacy and social networks dynamics, has created a fertile field for disinformation. Faced with its exponential growth, since 2014, research focused on combating false content in the media. From descriptive approach, this study analyzed 200 documents fact-checking fake news published between 2014 2022 scientific journals indexed Scopus. This found that Europe United States are leading way number of authors publishing subject. universities ones host most significant working fact-checking, while methodologies used, mostly <i>ad hoc</i> due to novelty topic, allow reflect need promote work design, testing, evaluation prototypes or real experiences within field. common contributions include typologies manipulation mechanisms, models evaluating detecting disinformation, proposals combat strengthen verification studies role spread efforts develop literacy among public journalists, case fact-checkers, identification factors influence belief news, analysis relationship verification, politics, democracy. It is concluded it essential connects academy industry raise awareness address these issues different actors scenario.

Language: Английский

---

Liars know they are lying: differentiating disinformation from disagreement

Humanities and Social Sciences Communications, Journal Year: 2024, Volume and Issue: 11(1)

Published: July 31, 2024

Abstract Mis- and disinformation pose substantial societal challenges, have thus become the focus of a substantive field research. However, misinformation research has recently come under scrutiny on two fronts. First, political response emerged, claiming that aims to censor conservative voices. Second, some scholars questioned utility altogether, arguing is not sufficiently identifiable or widespread warrant much concern action. Here, we rebut these claims. We contend spread misinformation—and in particular willful disinformation—is demonstrably harmful public health, evidence-informed policymaking, democratic processes. also show outright lies can often be identified differ from good-faith contestation. conclude by showing how at least partially mitigated using variety empirically validated, rights-preserving methods do involve censorship.

Language: Английский

---

Examining the association between social media fatigue, cognitive ability, narcissism and misinformation sharing: cross-national evidence from eight countries

Scientific Reports, Journal Year: 2023, Volume and Issue: 13(1)

Published: Sept. 18, 2023

Abstract Several studies have explored the causes and consequences of public engagement with misinformation and, more recently, COVID-19 misinformation. However, there is still a need to understand mechanisms that cause propagation on social media. In addition, evidence from non-Western societies remains rare. This study reports survey eight countries examine whether media fatigue can influence users believe misinformation, influencing their sharing intentions. Our insights also build prior cognitive personality literature by exploring how this mechanism conditional upon users’ ability narcissism traits. The results suggest false beliefs which translates into We find those high levels are less likely share low most due fatigue. one first provide cross-national comparative highlighting adverse effects establishing relationship not universal but dependent both dark traits individuals.

Language: Английский

---

Identifying multimodal misinformation leveraging novelty detection and emotion recognition

Journal of Intelligent Information Systems, Journal Year: 2023, Volume and Issue: 61(3), P. 673 - 694

Published: June 6, 2023

Language: Английский

---

Liars Know They Are Lying: Differentiating Disinformation from Disagreement

Published: Jan. 25, 2024

Mis- and disinformation pose substantial societal challenges, have thus become the focus of a substantive field research. However, misinformation research has recently come under scrutiny on two fronts. First, political response emerged, claiming that aims to censor conservative voices. Second, some scholars questioned utility altogether, arguing is not sufficiently identifiable or widespread warrant much concern action. Here, we rebut these claims. We contend spread misinformation—and in particular willful disinformation—is demonstrably harmful public health,evidence-informed policymaking, democratic processes. also show outright lies can often be identified differ from good-faith contestation. conclude by showing how at least partially mitigated using variety empirically validated, rights-preserving methods do involve censorship.

Language: Английский

---

Domain- and category-style clustering for general fake news detection via contrastive learning

Information Processing & Management, Journal Year: 2024, Volume and Issue: 61(4), P. 103725 - 103725

Published: April 12, 2024

Language: Английский

---

Abstractive Summarizers Become Emotional on News Summarization

Applied Sciences, Journal Year: 2024, Volume and Issue: 14(2), P. 713 - 713

Published: Jan. 15, 2024

Emotions are central to understanding contemporary journalism; however, they overlooked in automatic news summarization. Actually, summaries an entry point the source article that could favor some emotions captivate reader. Nevertheless, emotional content of summarization corpora and behavior models still unexplored. In this work, we explore usage established methodologies study models. Using these methodologies, two widely used corpora: Cnn/Dailymail Xsum, capabilities three state-of-the-art transformer-based abstractive systems for eliciting generated summaries: Bart, Pegasus, T5. The main significant findings as follows: (i) persistent corpora, (ii) summarizers approach moderately well reference summaries, (iii) more than 75% introduced by novel words present ones. combined use has allowed us conduct a satisfactory

Language: Английский

---

Evaluating the generalisability of neural rumour verification models

Information Processing & Management, Journal Year: 2022, Volume and Issue: 60(1), P. 103116 - 103116

Published: Oct. 26, 2022

Research on automated social media rumour verification, the task of identifying veracity questionable information circulating media, has yielded neural models achieving high performance, with accuracy scores that often exceed 90%. However, none these studies focus real-world generalisability proposed approaches, is whether perform well datasets other than those which they were initially trained and tested. In this work we aim to fill gap by assessing top performing verification covering a range different architectures from perspectives both topic temporal robustness. For more complete evaluation generalisability, collect release COVID-RV, novel dataset Twitter conversations revolving around COVID-19 rumours. Unlike existing datasets, our COVID-RV contains rumours follow format prominent benchmarks, while being them in terms time scale, thus allowing better assessment robustness models. We evaluate model performance three popular understand limitations advantages architectures, training scenarios. find dramatic drop when testing used for training. Further, ability generalise few-shot learning setup, as word embeddings are updated vocabulary new, unseen rumour. Drawing upon experiments discuss challenges make recommendations future research directions addressing important problem.

Language: Английский

---

A Systematic Literature Review and Meta-Analysis of Studies on Online Fake News Detection

Information, Journal Year: 2022, Volume and Issue: 13(11), P. 527 - 527

Published: Nov. 4, 2022

The ubiquitous access and exponential growth of information available on social media networks have facilitated the spread fake news, complicating task distinguishing between this real news. Fake news is a significant barrier that has profoundly negative impact society. Despite large number studies detection, they not yet been combined to offer coherent insight trends advancements in domain. Hence, primary objective study was fill knowledge gap. method for selecting pertinent articles extraction created using preferred reporting items systematic reviews meta-analyses (PRISMA). This reviewed deep learning, machine ensemble-based detection methods by meta-analysis 125 aggregate their results quantitatively. primarily focused statistics quantitative analysis data from numerous separate investigations identify overall trends. were reported spatial distribution, approaches adopted, sample size, performance terms accuracy. According between-study variance high heterogeneity found with τ2 = 3.441; ratio true total observed variation I2 75.27% chi-square (Q) 501.34, degree freedom 124, p ≤ 0.001. A p-value 0.912 Egger statistical test confirmed absence publication bias. findings demonstrated satisfaction effectiveness recommended included. Furthermore, can inform researchers about various use detect online

Language: Английский

---

Fake Information Detection Using Deep Learning Methods: A Survey

Published: Jan. 27, 2023

Fake content has always existed, even before the internet was founded. Because social media is free to use and accessible, a great deal of information shared on these sites. These platforms play significant role in dissemination information, whether accurate or false. The unregulated proliferation fake creation that we've seen recent years poses constant threat democracy. articles have power persuade individuals, leaving them perplexed. Deep learning techniques are extremely useful for detecting information. This paper analyses multiple DL datasets used by different researchers analysis aids detection bogus

Language: Английский

---