Overview
Newsler combines personalized news aggregation with a recommendation engine to help users cut through information overload. Originally developed as a Computer Science IA project, the system aggregates content from multiple sources, filters by user preferences, and applies matrix factorisation-based recommendations to surface the most relevant articles.
Key Features
- Smart Filtering: Prioritises news by user-defined keywords and topic interest.
- Real-time Updates: Pulls data from multiple News APIs for up-to-the-minute coverage.
- Personalised Recommendations: Matrix factorisation-based engine to tailor feeds at scale.
- Minimalist Reader Mode: Distraction-free article view with archiving support.
- Archive & Offline: Save articles for offline reading and long-term reference.
Contributions (Recommender)
- Built the recommendation pipeline to handle user interaction data and deliver personalised feeds.
- Designed the backend for scalable updates and aggregate indexing to support thousands of users.
- Implemented secure data handling and authentication to protect user preferences and archives.