User interactions and engagement patterns

Study of user behavior and engagement with news content and misinformation, including commenting, sharing, liking, and other interaction signals on social media platforms.

Approaches

Network-based analysis:
Constructing user-content networks from interaction patterns (comments, reactions, shares) to understand propagation and characterize content veracity.

Temporal analysis:
Examining how engagement patterns change over time to detect misinformation spread and identify coordinated behavior.

User characterization:
Building models of user preferences, susceptibility, and influence from historical behavior patterns.

Synthetic data generation:
Using language models to simulate diverse user perspectives and generate synthetic interactions when real data is unavailable or incomplete.

Graph neural networks:
Applying GNNs to user-content networks to jointly model content and behavioral signals for detection tasks.

Key papers in this wiki

• DELL: Generating Reactions and Explanations for LLM-Based Misinformation Detection (2024) — Generates synthetic user-news interaction networks from diverse demographic perspectives using LLMs; applies graph neural networks on these networks to improve misinformation detection across multiple tasks.

Related concepts

• Fake news sharing behavior — how users decide to spread content
• Social media and misinformation — platforms as distribution channels
• Misinformation spread and diffusion — propagation dynamics