v0.3.0

Performance & Personalization Upgrade

This release introduces major enhancements to the RecSys-Lite API, focusing on improved performance and more personalized recommendation
experiences.

API Response Caching

• Implemented TTL-based caching system for all recommendation endpoints
• Added configurable cache expiration (0-3600 seconds)
• Improved response times by up to 5x for frequently requested recommendations
• Added cache metrics and statistics for monitoring performance
• New endpoints for managing and configuring caches

Personalized Reranking

• Four powerful reranking strategies to enhance recommendation quality:
• Diversity: Ensures recommendations span multiple categories
• Novelty: Surfaces unexpected or less obvious items
• Popularity: Boosts trending or popular items
• Hybrid: Combines all strategies with configurable weights
• Recommendation explanations based on the dominant ranking factor
• Customizable weights for fine-tuning the balance between strategies
• Category boosting for business-driven recommendation priorities

Documentation

• Comprehensive API documentation for all new parameters
• Example responses showing the complete capabilities
• Updated CHANGELOG detailing all improvements