🌍 A multilingual research initiative detecting, classifying & countering COVID-19 misinformation on social media through AI-driven tools, open datasets and community workshops.
🌐 Web Platform • 🎞️ Platform Demo • 🎓 Workshop Series
|
|
The "Multilingual COVID-19 Fake News Detection and Intervention on Social Media" project represents a groundbreaking interdisciplinary initiative that bridges machine learning, data journalism, and public health research. By leveraging cutting-edge AI technologies and comprehensive multilingual datasets, this project empowers researchers, journalists, and policy-makers to effectively identify, analyze, and combat the spread of COVID-19 misinformation across diverse social media platforms.
"To provide actionable insights and solutions to mitigate the risks associated with fake news and foster a more reliable and informed digital ecosystem through advanced machine learning and natural language processing technologies."
Our comprehensive approach encompasses real-time detection systems, extensive data collection, user-friendly verification tools, and evidence-based intervention strategies to address the complex challenges of the global infodemic.
graph TD
A[🔍 Data Collection] --> B[🤖 AI Model Development]
B --> C[🛠️ Tool Creation]
C --> D[🗣️ Intervention Strategies]
D --> E[📊 Open Research]
E --> A
A --> A1[52+ Languages<br/>Multilingual Corpus]
B --> B1[State-of-the-art<br/>Detection Models]
C --> C1[Fact-checking<br/>Web Platform]
D --> D1[Community Workshops<br/>& Surveys]
E --> E1[Open-access<br/>Resources]
- 🔍 Build the world's largest multilingual COVID-19 misinformation corpus spanning 52+ languages
- 🤖 Develop state-of-the-art detection & verification models for text and social context analysis
- 🛠️ Design intuitive fact-checking and visual analytics tools for end-users
- 🗣️ Evaluate intervention strategies through comprehensive surveys & community workshops
- 📊 Provide open-access resources for global research collaboration
An interactive web application that leverages advanced machine learning algorithms to classify COVID-19 related claims in real-time. The tool provides instant verdicts with confidence scores and detailed explanations of the decision-making process, making it accessible for both researchers and general users.
- ✅ Real-time claim verification - Instant analysis of user-submitted content
- 🌍 Multilingual support - Processing in 52+ languages
- 📊 Confidence scoring system - Transparent reliability indicators
- 🔍 Transparent decision explanations - Detailed reasoning behind classifications
- 🎯 User-friendly interface - Accessible for both researchers and general public
Our comprehensive dataset represents the most extensive collection of multilingual COVID-19 misinformation data available for academic research. Updated weekly with rich social engagement metadata, this resource enables researchers worldwide to develop and test their own detection models.
| Category | Metric | Value |
|---|---|---|
| 📰 Content | Total Records | 45,904 news articles |
| 🌍 Coverage | Languages | 52 languages |
| 🐦 Social Media | Twitter Posts | 90,000+ tweets |
| 📺 Publishers | News Sources | 56 verified sources |
| ✅ Verification | Fact-check Sites | 235 international platforms |
- 🔄 Weekly automated updates - Fresh data continuously added
- 📊 Rich metadata - Social engagement metrics, timestamps, source tracking
- ✅ Verified fact-check labels - Professional fact-checker annotations
- 🆓 Open access for academic use - Free for research purposes
- 🔧 Standardized format - Cross-platform compatibility
A comprehensive stratified online survey designed to understand public exposure patterns, belief systems, and behavioral responses to COVID-19 misinformation. Deployed across Indonesia and Australia, this research component provides crucial insights into the real-world impact of misinformation on diverse communities.
- 🌏 Cross-cultural comparative analysis - Indonesia vs Australia
- 👥 Demographic stratification - Age, education, location factors
- 📈 Behavioral pattern identification - Information consumption habits
- 🎯 Intervention effectiveness measurement - Strategy impact assessment
- 📋 Policy recommendation development - Evidence-based guidelines
🤝 Conducted in partnership with Deakin University and Universitas Gadjah Mada
Our knowledge transfer initiative includes comprehensive research workshops that bring together international scholars, practicing journalists, and frontline health communicators. These collaborative sessions facilitate the sharing of methodologies, early research findings, and evidence-based policy recommendations.
|
🎯 First Workshop
|
🚀 Second Workshop
|
- 📹 Workshop Recording → Watch on YouTube
- 📊 Ongoing Webinars → Regular updates and community engagement
- 📋 Workshop Materials → Presentations and resources available
- Machine Learning: Advanced NLP models for multilingual text analysis
- Deep Learning: Neural networks for pattern recognition and classification
- Natural Language Processing: Text preprocessing and feature extraction
- Web Development: Real-time platform for user interaction
- Data Engineering: Scalable pipelines for large-scale data processing
- 🎯 Detection Accuracy: Achieved 95%+ accuracy across multiple languages
- 🌍 Cross-cultural Insights: Identified unique misinformation patterns by region
- 📱 Platform Engagement: Over 10,000+ fact-checks performed by users
- 🔬 Academic Impact: Multiple publications in top-tier conferences
- Conference papers on multilingual misinformation detection
- Workshop presentations at international venues
- Open-access datasets for research community
- Policy briefs for public health organizations
|
🇦🇺 Deakin University |
🇮🇩 Universitas Gadjah Mada |
We extend our gratitude to:
- 🏅 VESKI - Study Melbourne Research Partnership for funding support
- 🌍 International research community for collaboration and expertise
- ✅ Fact-checking organizations for data and verification support
- 📱 Social media platforms for API access and partnership
- 👥 Workshop participants for valuable feedback and engagement
- 🌐 Explore the Platform: veski.counterinfodemic.org
- 📊 Access the Dataset: Contact us for academic use
- 🛠️ Try the Fact-Checker: Real-time verification tool
- 📋 Join the Community: Participate in surveys and workshops
Interested in collaborating or accessing our resources? We welcome partnerships with:
- 🎓 Academic researchers
- 📰 Journalists and media organizations
- 🏛️ Policy makers and health organizations
- 💻 Tech developers and AI practitioners
Last Updated: June 2025 | Project Status: Active Research Phase
Fighting the Infodemic Through Science 🔬
Supported by VESKI - Study Melbourne Research Partnership