World Trends Today — Top Global Events
Real-time analysis of global news trends detected from over 300,000 daily articles. WorldTrend identifies breaking events, developing stories, and emerging trends using 17 anomaly detectors and co-occurrence graph analysis. Last updated: 2026-04-24.
Frequently Asked Questions
How does WorldTrend detect trending events?
WorldTrend analyzes over 300,000 news articles daily using 17 independent anomaly detectors across 10 axes (volume, salience, slope, changepoint, novelty, breadth, acceleration, sentiment, distribution, and graph structure). Entities flagged by multiple detectors are clustered using NPMI co-occurrence analysis and Leiden community detection to form coherent events.
What data sources does WorldTrend use?
WorldTrend processes the GDELT Global Entity Graph, which covers news articles from virtually every country in over 100 languages. This provides entity-level analysis with salience scoring, enabling detection of both major headlines and emerging niche trends.
What are the three trend layers?
WorldTrend classifies events into three layers: Breaking (anomaly-detected events happening right now), Developing (stories that persist and evolve over days), and Emerging (long-term trends showing sustained growth patterns detected by 11 trend signals including Mann-Kendall, floor shift, and cluster evolution).
How often is the data updated?
The pipeline processes new data daily, analyzing the previous day's global news coverage. The 3D visualization shows the latest detected events, developing stories, and emerging trends.