Project Description

A real-time dashboard that automatically collects news headlines from 100+ sources, analyzes their sentiment using a custom RoBERTa-based API, and visualizes the results across topics. Headlines are grouped into Positive, Neutral, and Negative columns — each card is clickable and opens a popup with the article description and a link to the full story.

• Fetches headlines across 4 topics: technology, science, business, health
• Sentiment analysis powered by the custom Sentiment API
• Stores articles in PostgreSQL with automatic deduplication
• Auto-refreshes every 6 hours via a background scheduler
• Interactive filters by topic and sentiment
• Pie chart and bar chart for sentiment distribution at a glance
• 3-column headline feed (Positive / Neutral / Negative) with confidence scores

Technologies