📊 Visualization & Data Exploration
📰 About GDELT Data
The Global Database of Events, Language, and Tone (GDELT) monitors global news media in over 100 languages, cataloging every political action, statement, and interaction worldwide. Our visualizations use 342,754 GDELT events (2021-2025) where any country acted on Taiwan, revealing global engagement patterns and China's reactions.
Event Codes (1-20): GDELT categorizes actions from cooperation (e.g., Code 05: Diplomatic Cooperation, Code 06: Material Aid) to conflict (e.g., Code 13: Threaten, Code 15: Military Posture, Code 19: Fight). Goldstein Scale (-10 to +10): Measures conflict vs. cooperation intensity—positive scores indicate cooperative actions, negative scores indicate hostile actions.
🛩️ ADIZ Database Visualizer
Interactive timeline showing PLA ADIZ baseline with arms sales, diplomatic events, bills, and ship transits overlays. Explore trigger-response patterns.
ActiveData: 2020-2025 ADIZ incidents + trigger events
🗺️ GDELT Interactive Map
Scrub through time to see where global actions toward Taiwan originate. Events display in their actor's home country (e.g., US actions appear in USA). Filter by country, event type (📢 statements, 🤝 cooperation, ⚠️ threats), and get direct links to news sources.
✅ ActiveBest for: Exploring specific events and temporal patterns
🌐 3D Network Globe
Visualize global relationships with Taiwan on a rotating Earth. Animated arcs flow from countries to Taiwan showing cooperation (green), conflict (red), and verbal actions (blue). See China's reactions alongside other nations' engagements—reveals global impact patterns.
✅ ActiveBest for: Understanding relationship networks and spatial patterns
🔄 Sankey Flow Diagram
Identify causal relationships by visualizing how countries and event types flow through Taiwan. Width = frequency of events. Compare which countries drive cooperation vs. conflict, and aggregate patterns by time period.
✅ ActiveBest for: Comparative analysis and causality patterns
🔍 Analysis Modules
Module 1: Single Event
Analyze ADIZ response around a specific event. See impact metrics, time-to-peak, persistence, and confounders.
MVP ActiveModule 2: Event Stack Analyzer
Stack multiple events of the same type to see average ADIZ patterns. Reveals typical response curves you can't see in the regular visualizer.
✅ ActiveModule 3: A/B Comparison
Compare two event groups side-by-side. Which triggers stronger reactions: arms vs ships? Offensive vs defensive? Get timing, severity metrics, and AI-like recommendations.
✅ ActiveModule 4: Heatmap View
Event-aligned and calendar heatmaps showing intensity patterns over time.
Coming SoonModule 5: Rhetoric Analyzer
Analyze PRC rhetoric shifts around events using GDELT tone, volume, and key themes correlation with ADIZ.
Coming SoonModule 6: AI Analyzer
Dual chat interface: Global LLM (Gemini + geopolitical knowledge) and Local RAG (WebLLM analyzing only your dataset). Natural language Q&A about ADIZ spikes.
✅ ActiveModule 7: Predictor
Trigger ranking by impact score and what-if simulator to predict ADIZ response to hypothetical events.
Coming SoonModule 8: Reactive vs Pre-Planned
Classify if PLA response was genuinely triggered or pre-planned using event as excuse. Analyzes 5 signals: timing, pre-buildup, magnitude, symbolic dates, pattern shape.
✅ ActiveModule 9: Actions Roundup
Intelligent analysis of top ADIZ spike days. Automatically identifies likely triggers across all event categories with confidence scores. Highlights unexplained reactions for intelligence gap analysis.
✅ Activeℹ️ About This Dashboard
China Tripwire analyzes Taiwan ADIZ incursions in relation to political, military, and diplomatic events. Like a security tripwire that detects intrusions, this dashboard identifies the geopolitical triggers that precede China's military responses, helping analysts understand cause-and-effect relationships in cross-strait tensions.
📊 Data includes: ADIZ baseline, arms sales, diplomatic events, legislative bills, ship transits, and political/symbolic events from 2020-2025.