Introduction
Disinformation in online environments is primarily spread by malicious actors or entities with biased agendas. Fake news can have severe consequences across political, social, business, and media sectors [[1]](#references). Detecting fabricated information is critical in global contexts, as misleading content can propagate rapidly through news outlets, amplifying its impact.
While extensive research exists on fake news detection [[2]](#references)—employing machine learning (ML), deep learning (DL), and natural language processing (NLP) techniques [[3]](#references)—current methods often fall short of achieving high accuracy.
The Trust Crisis in Media
- Declining Trust: Only 19% of EU adults reported high trust in news media in 2019, while 40% expressed low or no trust [[4]](#references).
- Economic Incentives: Click-driven ad revenue models prioritize engagement over factual rigor, blurring lines between real and fake news.
Blockchain technology emerges as a solution, offering transparency, traceability, and decentralization via distributed ledger technology (DLT) [[5]](#references). Its immutable, cryptographically secured data structure ensures integrity, making it ideal for combating misinformation.
Proposed Solution: FiDisD Project
The FiDisD project leverages:
- Blockchain: For decentralized trust and data validation.
- Crowd Wisdom + Federated AI: To analyze and score news credibility.
- Decentralized Retrieval: Multiple independent actors verify content via majority rule.
Key Innovations
- Separate Crawling/Scraping Processes: Enhances scalability by decoupling URL extraction (crawlers) from article data extraction (scrapers).
- Community Involvement: Third-party actors perform crawling/scraping, ensuring decentralization.
- Majority-Rule Validation: Malicious actors are flagged if their submissions deviate from consensus.
System Architecture
1. Core Components
- OffchainCore: Central hub for managing crawlers/scrapers and validating data.
- WebCrawler: Extracts URLs and identifies article pages.
- WebScraper: Retrieves article details (title, content, author, etc.).
2. Decentralized Workflow
- URL Allocation: OffchainCore assigns URLs to random crawler/scraper subsets.
- Data Validation: Hashes of extracted content are compared; majority consensus determines correctness.
- Blockchain Integration: Validated article hashes are stored on-chain as immutable proof.
3. Extraction Template
JSON-based templates pinpoint article elements using CSS selectors:
{
"featured_image": ["img", "src", false, "\\S+"],
"author": [".author-name", "text", true]
}- Keys: Define extractable features (e.g.,
featured_image). - Values: Specify HTML element location, property type (text/attribute), and optional regex filters.
Deployment & Scalability
Cloud Deployment Options
| Service | OpenStack | AWS | GCP | Azure |
|------------------|--------------------|------------------|----------------------|-------------------|
| Compute | Nova | EC2 | Compute Engine | Virtual Machines |
| Database | Trove | RDS | Cloud SQL | Azure SQL |
| Storage | Swift | S3 | Cloud Storage | Blob Storage |
Optimization: Use containerization (Docker/Kubernetes) for crawler/scraper instances to streamline deployment.
Toward Full Decentralization
Future Enhancements
- IPFS Integration: Store article data on decentralized file systems like IPFS [[6]](#references).
- Decentralized Oracles: Use hybrid smart contracts for autonomous majority-rule validation [[7]](#references).
- Sybil Attack Prevention: Dynamic actor rotation and cryptographic identity verification.
Case Study & Results
Testing Environment
- Platform: OpenStack (Rocky Linux, 2 vCPUs, 4GB RAM).
- Websites: 7 Romanian news portals (e.g., Digi24, Hotnews).
Key Metrics
| Website | Avg. Processing Time | Article Pages (%) |
|--------------|-------------------------|----------------------|
| AgerPres | 217 ms | 22% |
| Stiripesurse | 887 ms | 96% |
Finding: Majority-rule validation effectively filters malicious submissions, with scrapers processing URLs 16x faster than crawlers.
FAQs
Q: How does blockchain prevent fake news?
A: By storing immutable hashes of validated content, ensuring tamper-proof records.
Q: What happens if crawlers disagree on a URL?
A: OffchainCore flags discrepancies; repeated deviations result in actor penalties.
Q: Can the system handle non-English content?
A: Yes—UTF-8 support allows multilingual extraction.
Conclusion
The FiDisD system combines blockchain, crowd wisdom, and AI to create a transparent, scalable, and decentralized news-validation framework. Future work will focus on IPFS integration and oracle-based automation for full decentralization.
👉 Explore blockchain solutions for media integrity
References
- Wu et al. (2022). Internet Research, 32, 1662–1699.
- Bondielli & Marcelloni (2019). Information Sciences, 497, 38–55.
- Ciampaglia et al. (2015). PLoS ONE, 10, e0141938.
- Guttmann (2019). Statista Survey on EU Media Trust.
- Soltani et al. (2022). Applied Sciences, 12, 7898.
- Trautwein et al. (2022). ACM SIGCOMM.
- Breidenbach et al. (2021). Chainlink 2.0 Whitepaper.
### Key SEO Enhancements:
- **Headings**: Hierarchical structure (`H2`–`H4`) for readability.
- **Keywords**: "Blockchain news verification," "decentralized crawler," "fake news detection."
- **Tables**: Comparative cloud deployment options for skimmability.
- **FAQ**: Targets search intent for "how blockchain fights disinformation."
- **Anchor Text**: Engaging CTA linking to a relevant resource.