GNN-based Recommendation System

Published:

An advanced recommendation system that leverages Graph Neural Networks to tackle the cold-start problem. This project enhances traditional collaborative filtering methods with graph-based learning to improve recommendations for new users and items.

Key Features

  • Dynamic graph layers for user-item interactions
  • 20-D user/item embeddings for efficient representation
  • 4.21% improvement in recommendation accuracy
  • Recall@10 improved by 5.38% for sparse user histories

Technologies

  • PyTorch Geometric
  • Graph Neural Networks
  • Collaborative filtering algorithms
  • Python and deep learning frameworks