Global Emergency Map AI

An automated, production-style project that tracks live global emergencies (earthquakes, wildfires, storms, floods), enriches them with a lightweight NLP pipeline (optional translation via Hugging Face), and publishes an interactive map with filters, a date range picker, severity legend, and a details panel.

• Zero servers — everything runs in GitHub Actions and serves from docs/.
• Auto-refresh — data pipeline runs daily at 9:00 AM ET (manual run supported).
• Live demo — hosted on GitHub Pages.

Features

• Sources: USGS (earthquakes), NASA EONET (open natural events), ReliefWeb (best-effort).
• NLP (optional): translate non-English titles/descriptions to English using Hugging Face Inference API.
• UI:
  • Map pins by type, color-coded.
  • Date range filter (from–to calendars).
  • Severity legend for earthquakes (magnitude bins).
  • Details panel with title, time, severity, summary, coordinates, and source link.

Quick Start (GitHub)

README.md requirements.txt docs/ index.html styles.css app.js events.json pipeline/ init.py fetchers.py normalize.py hf_client.py run.py .github/ workflows/ ingest.yml deploy.yml # optional, if using Pages via Actions

2. Settings → Actions → General → Workflow permissions → Read and write → Save.

3. Pages

• Either Deploy from a branch (Branch: main, Folder: /docs)
• Or set Source: GitHub Actions and include deploy.yml (recommended).

4. (Optional) Hugging Face token:
   Settings → Secrets and variables → Actions → New repository secret
   Name: HF_TOKEN → Value: your token.

5. Run it: Actions → Global Emergency Map → Run workflow.
   After success, your site will update automatically.

Local Dev (optional)

python -m venv .venv && source .venv/bin/activate # Windows: .venv\Scripts\activate pip install -r requirements.txt python pipeline/run.py

open docs/index.html in a browser

Tech Stack Data: USGS, NASA EONET, ReliefWeb

NLP: Hugging Face Inference API (e.g., facebook/nllb-200-distilled-600M)

Frontend: Leaflet + vanilla JS/CSS (static)

CI/CD: GitHub Actions (daily cron + optional Pages deploy)

Roadmap Event-type classifier + severity normalization (storms/floods)

Email/Slack alerts for high-severity areas

Country/region filters

Download CSV / CSV snapshot per run