Misinformation
Information Disorder Taxonomy
- Misinformation: False or inaccurate information, regardless of intent. The spreader may genuinely believe it to be true.
- Disinformation: Deliberately false information spread with intent to deceive. Includes propaganda, hoaxes, and strategic manipulation.
- Malinformation: Genuine information shared with malicious intent, such as leaking private information to cause harm.
The Key Distinction
The difference lies in intent and truth value: misinformation is accidentally wrong, disinformation is deliberately wrong, and malinformation is deliberately harmful even when true.
Key Properties / Variants
| Type | Intent | Truth Value | Example |
|---|---|---|---|
| Misinformation | Unintentional | False | Sharing outdated medical advice |
| Disinformation | Deliberate deception | False | State-sponsored propaganda |
| Malinformation | Malicious | True | Doxxing, revenge porn |
IR’s Role in the Information Ecosystem
Amplification Risks
IR systems may inadvertently promote false content because:
- Engagement optimization: Sensational content generates more clicks
- Controversy bias: Controversial topics drive engagement
- Filter bubbles: Personalization reinforces existing beliefs
- Feedback loops: Popular content becomes more visible, amplifying initial spread
Mitigation Opportunities
IR systems can potentially:
- Downrank unreliable sources in search results
- Promote authoritative information for sensitive queries
- Label content with credibility indicators
- Diversify results to break filter bubbles
Technical Approaches
Source Credibility Assessment
- Domain-level reliability scores
- Author expertise evaluation
- Citation and linking patterns
- Historical accuracy tracking
Content-Based Detection
- Claim verification against knowledge bases
- Stylistic indicators of unreliability
- Contradiction detection across sources
- Multimodal analysis (text, images, metadata)
User-Facing Interventions
- Warning labels on disputed content
- Related articles providing context
- “Read before sharing” friction
- Fact-check panels
Significant Challenges
- Misinformation evolves to evade detection
- “Ground truth” is contested for many claims
- Interventions may backfire (reactance, distrust)
- Cultural and linguistic variation in what constitutes misinformation
- Who decides what is “true”? (epistemological challenge)
Connections
- Related to: Algorithmic Fairness (who is harmed by misinformation spread)
- Requires: Explainability (explaining why content was flagged)
- Critiqued by: Emancipatory IR (questions who controls truth-determination)
- Technical approaches use: BERT for IR, Cross-Encoder for claim verification