MEDIA INTELLIGENCE DASHBOARD

The Same War,
Seven Narratives

How Al Jazeera, BBC, Reuters, RT, Xinhua and others describe the same events in the US-Israel-Iran war — analyzed by AI

● LIVE ARTICLES ANALYZED SOURCES MONITORED
Live narrative divergence
Source
Frame
These four indicators summarize the current state of media coverage. The divergence score shows how fragmented the narrative is across sources. The higher it is, the more sources disagree on how to frame the conflict.
DIVERGENCE SCORE TODAY
DOMINANT FRAME: MILITARY
ALARMIST TONE
highest since conflict began
CONFLICT ONGOING
since Feb 28, 2026
How each source drifted from center
Score 0 = narrative identical to global average | Score 100 = maximum divergence
This chart shows the daily evolution of the divergence score for each source. A rising line means the source is using frames increasingly distinct from the global average. Hover over the chart to see details.
The narrative DNA of each source
Frame distribution and characteristic vocabulary this week
Each card shows how a source frames the conflict. The colored bar represents the proportion of each narrative frame. The vocabulary below shows terms with the highest TF-IDF score — words this source uses much more than the others.
Loading...
Editorial Compass
Where each source sits on two axes: tone (neutral → propagandistic) and frame lean (factual/military → narrative/geopolitical)
Compare two sources
Select two sources to see how they describe the same events
The comparator shows side by side the frame distribution, vocabulary and tone of two sources. The "narrative distance" in the center is calculated using the same Bhattacharyya metric as the main chart — the higher it is, the more divergent the sources are in how they describe the conflict.
vs
Loading...
How it works
Collection via GDELT Project

The Global Database of Events, Language, and Tone (GDELT) indexes news from thousands of outlets in real time. We query the API hourly for conflict-related articles, filtering by domain for the 7 selected sources.

AI Classification

Each headline and lead is sent to Claude Haiku (Anthropic) with a prompt that extracts: narrative frame (7 categories), emotional tone (5 categories), and revealing vocabulary. The model does not judge factual accuracy — it classifies only how the text was written.

Divergence Score (Bhattacharyya)

We compare each source's frame distribution against the daily global median using Bhattacharyya distance — a statistical metric measuring overlap between two probability distributions. Score 0 = the source describes the conflict exactly like the average. Score 100 = completely different framing from all others.

Pipeline de dados
GDELT collector ~168 arts SQLite articles.db scorer.py Bhattacharyya analyzer.py Claude Haiku Dashboard Flask+Chart.js
Important limitations
• GDELT returns mainly title+lead, not full text
• Frame analysis is probabilistic, not deterministic
• English sources only — Arabic/Hebrew versions may differ
• Divergence score measures framing difference, not truthfulness
Latest analyzed articles
Click on any article to expand and see the AI reasoning: why was this headline classified in this frame? The explanation is generated in real time by Claude Haiku.
Carregando artigos...
Transparency & Audit