README.md

metadata

title: Madrid Content Analyzer
emoji: 🏛️
colorFrom: blue
colorTo: green
sdk: gradio
sdk_version: 4.44.0
app_file: app.py
pinned: false
license: mit

🏛️ Madrid Content Analyzer

Automated analysis of language clarity in Madrid City Council communications.

📋 What It Does

This application:

• 📥 Fetches content from Madrid City Council RSS feeds and Open Data Portal
• 🔍 Analyzes language clarity using the Aclarador system
• 📊 Tracks trends over time in a DuckDB database
• 📈 Visualizes results in an interactive dashboard

🎯 Features

📊 Dashboard

• Real-time statistics on analyzed content
• Clarity score distribution charts
• Timeline of content and scores
• Category breakdown

📝 Content Browser

• Search and filter content by date, category, and clarity score
• View detailed analysis for each item
• Identify areas for improvement

📈 Analytics

• Find low-clarity items that need improvement
• Track trends over time
• Export data for further analysis

⚙️ Settings

• Manual content fetch trigger
• Database statistics
• Fetch logs and history

🔄 Automatic Updates

Content is fetched and analyzed automatically every 6 hours. The app runs continuously on Hugging Face Spaces with persistent data storage.

🛠️ Technology Stack

• Frontend: Gradio 4.44
• Database: DuckDB (analytics-optimized, 16GB storage)
• Scheduler: APScheduler (background tasks)
• Visualization: Plotly
• Analysis: Aclarador language clarity analyzer

💾 Data Storage

Data is stored in /data/madrid.duckdb which persists across Space restarts. The database can hold millions of content items within the 16GB Space storage limit.

🚀 Usage

Simply visit the Space URL and explore the tabs:

1. Dashboard - See overall statistics and trends
2. Browse Content - Search and filter analyzed content
3. Analytics - Find improvement opportunities
4. Settings - Trigger manual updates, view logs

📈 Data Sources

• RSS Feed: https://diario.madrid.es/feed
• Open Data Portal: https://datos.madrid.es/portal/site/egob

🔒 Privacy

This Space analyzes publicly available content from Madrid City Council. No personal data is collected or stored.

📄 License

MIT License - Feel free to fork and adapt!

🤝 Contributing

Contributions are welcome! This project helps improve government communication clarity.

---

Built with 🤗 Hugging Face Spaces | Free forever 💰