ScoopRank
Political news intelligence: real-time scraping, NLP & a deep-learning classifier
What needed solving
Tracking political sentiment and entity popularity across news and social media in real time is a data-and-ML problem most teams cannot operationalize.
How I built it
Built a proactive scraper + Twitter listener feeding an NLP pipeline (DBpedia/Wikipedia entity extraction, sentiment analysis, a trained Keras classifier) with live WebSocket ranking updates.
The details
Political news-intelligence and ranking platform. A proactive Node.js scraper and Twitter listener feed a Django NLP pipeline: named-entity recognition, sentiment analysis and a trained Keras deep-learning model for news categorization, with real-time WebSocket ranking updates.
- Real-time scraper + Twitter/X listener
- DBpedia/Wikipedia named-entity recognition
- Trained Keras model for news categorization
- Sentiment analysis (pos/neg/neu) per article
- Live WebSocket ranking via Django Channels
We can build something that lasts.
Tell me what you are building. I will come back with an honest take on approach, scope and timeline, usually within 24 hours.