Community notes increase trust in fact-checking on social media

PNAS Nexus, Journal Year: 2024, Volume and Issue: 3(7)

Published: May 31, 2024

Community-based fact-checking is a promising approach to fact-check social media content at scale. However, an understanding of whether users trust community fact-checks missing. Here, we presented

Language: Английский

---

Differential impact from individual versus collective misinformation tagging on the diversity of Twitter (X) information engagement and mobility

Nature Communications, Journal Year: 2025, Volume and Issue: 16(1)

Published: Jan. 24, 2025

Abstract Fears about the destabilizing impact of misinformation online have motivated individuals and platforms to respond. Individuals increasingly challenged others’ claims with fact-checks in pursuit a healthier information ecosystem break down echo chambers self-reinforcing opinion. Using Twitter (now X) data, here we show consequences individual tagging: tagged posters had explored novel political expanded topical interests immediately prior, but being caused retreat into bubbles. These unintended were softened by collective verification system for moderation. In Twitter’s new feature, Community Notes, tagging was peer-reviewed other fact-checkers before revelation poster. With tagging, less likely from diverse engagement. Detailed comparison demonstrated differences toxicity, sentiment, readability, delay versus messages. findings provide evidence differential impacts moderation strategies on diversity engagement mobility across ecosystem.

Language: Английский

---

Did the Roll-Out of Community Notes Reduce Engagement With Misinformation on X/Twitter?

Proceedings of the ACM on Human-Computer Interaction, Journal Year: 2024, Volume and Issue: 8(CSCW2), P. 1 - 52

Published: Nov. 7, 2024

Developing interventions that successfully reduce engagement with misinformation on social media is challenging. One intervention has recently gained great attention X/Twitter's Community Notes (previously known as "Birdwatch"). a crowdsourced fact-checking approach allows users to write textual notes inform others about potentially misleading posts X/Twitter. Yet, empirical evidence regarding its effectiveness in reducing missing. In this paper, we perform large-scale study analyze whether the introduction of feature and roll-out U.S. around world have reduced X/Twitter terms retweet volume likes. We employ Difference-in-Differences (DiD) models Regression Discontinuity Design (RDD) comprehensive dataset consisting all corresponding source tweets since launch early 2021. Although observe significant increase fact-checks carried out via Notes, particularly for from verified many followers, find no significantly Rather, our findings suggest might be too slow effectively (and most viral) stage diffusion. Our work emphasizes importance evaluating field offers important implications enhance strategies media.

Language: Английский

---

A framework for detecting fake news by identifying fake text messages and forgery images

Enterprise Information Systems, Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 15, 2024

With the widespread adoption of social media, dissemination information has accelerated dramatically. However, this rapid spread poses challenges, as much content lacks verification, leading to proliferation fake news and manipulated visuals. Existing detection methods often emphasize text classification, overlooking visual forgeries. This study introduces FakeNews, a framework integrating modified BERT model (FaketextBERT) for CNN (ForgeryCNN) identifying forged images. Key contributions include broadening definition images, introducing modality-specific integrated approaches, achieving higher accuracy through Adam optimizer ELA Experimental results demonstrate enhanced performance improved recognition news.

Language: Английский

---

Community notes reduce the spread of misleading posts on X

Published: April 29, 2024

Community-based fact-checking is a promising approach to verify social media content and correct misleading posts at scale. Yet, causal evidence regarding its effectiveness in reducing the spread of misinformation on missing. Here, we performed large-scale empirical study analyze whether community notes reduce X. Using Difference-in-Differences design repost time series data for N=31,758 (community fact-checked) cascades that had been reposted more than 68 million times, found exposing users reduced by, average, 61.4%. However, our findings also suggest might be too slow intervene early (and most viral) stage diffusion. Our work offers important implications enhance community-based approaches media.

Language: Английский

---

Community notes increase trust in fact-checking on social media

Published: April 29, 2024

Community-based fact-checking is a promising approach to fact-check social media content at scale. However, an understanding of whether users trust community fact-checks missing. Here, we presented n = 1810 Americans with 36 misleading and non-misleading posts assessed their in different types interventions. Participants were randomly assigned treatments where was either accompanied by simple (i.e., context-free) misinformation flags formats (expert or flags), textual "community notes" explaining why the fact-checked post misleading. Across both sides political spectrum, notes perceived as significantly more trustworthy than flags. Our results further suggest that higher trustworthiness primarily stemmed from context provided explanations) rather generally towards fact-checkers. Community also improved identification posts. In sum, our work implies matters might be effective mitigate issues

Language: Английский

---

Readable and neutral? Reliability of crowdsourced misinformation debunking through linguistic and psycholinguistic cues

Frontiers in Psychology, Journal Year: 2024, Volume and Issue: 15

Published: Nov. 13, 2024

Background In the face of proliferation misinformation during COVID-19 pandemic, crowdsourced debunking has surfaced as a counter-infodemic measure to complement efforts from professionals and regular individuals. 2021, X (formerly Twitter) initiated its community-driven fact-checking program, named Community Notes Birdwatch). This program allows users create contextual corrective notes for misleading posts rate helpfulness others' contributions. The effectiveness platform been preliminarily verified, but mixed findings on reliability indicate need further research. Objective study aims assess by comparing readability language neutrality helpful unhelpful notes. Methods A total 7,705 2,091 spanning January 20, May 30, 2023 were collected. Measures reading ease, analytical thinking, affect authenticity derived means Wordless Linguistic Inquiry Word Count (LIWC). Subsequently, non-parametric Mann–Whitney U -test was employed evaluate differences between groups. Results Both groups are easy read with no notable difference. Helpful show significantly greater logical authenticity, emotional restraint than ones. As such, is validated in terms neutrality. Nevertheless, prevalence prepared, negative swear indicates manipulative abusive attempts platform. wide value range group overall limited consensus note also suggest complex information ecology within platform, highlighting necessity guidance management. Conclusion Based statistical analysis linguistic psycholinguistic characteristics, identified room improvement. Future endeavors could explore psychological motivations underlying volunteering, gaming, or even behaviors, enhance system integrate it broader infodemic

Language: Английский

---

JSocialFact: a Misinformation dataset from Social Media for Benchmarking LLM Safety

2021 IEEE International Conference on Big Data (Big Data), Journal Year: 2024, Volume and Issue: unknown, P. 3017 - 3025

Published: Dec. 15, 2024

Language: Английский

---