Tracking the Temporal Dynamics of News Coverage of Catastrophic and Violent Events
A new study quantifies how media framing shifts during violent events using semantic drift analysis.
Researchers Emily Lugos and Maurício Gruppi have published a significant study titled 'Tracking the Temporal Dynamics of News Coverage of Catastrophic and Violent Events' on arXiv. The paper presents a comprehensive analysis of how media narratives form and evolve during crises, using a massive corpus of 126,602 news articles from online publishers. By applying computational linguistics techniques, the researchers developed metrics to quantify narrative change through publication volume, semantic drift, semantic dispersion, and term relevance.
The study reveals that sudden, impactful events follow structured and predictable news-cycle patterns. These patterns are characterized by rapid surges in coverage volume immediately following an event, early semantic drift as initial reports give way to analysis and context, and gradual declines toward baseline coverage levels. The researchers also identified the specific terms that drive these temporal patterns, providing insight into how language evolves during crisis reporting. This work represents an important advancement in understanding how digital media shapes public discourse during moments of collective trauma and crisis.
- Analyzed 126,602 news articles to track media narrative evolution during crises
- Found predictable patterns: rapid coverage surges, early semantic drift, gradual decline
- Quantified narrative change using semantic drift, dispersion, and term relevance metrics
Why It Matters
Provides tools to understand and potentially counteract sensationalism or misinformation during breaking news events.