Graph convolution machine for context-aware recommender system

The latest advance in recommendation shows that better user and item representations can be learned via performing graph convolutions on the user-item interaction graph. However, such finding is mostly restricted to collaborative filtering (CF) scenario, where contexts are not available. In this work, we extend advantages of context-aware recommender system (CARS, which represents a generic type models handle various side information). We propose Graph Convolution Machine (GCM), an end-to-end framework consists three components: encoder, convolution (GC) layers, decoder. encoder projects users, items, into embedding vectors, passed GC layers refine embeddings with decoder digests refined output prediction score by considering interactions among user, item, context embeddings. conduct experiments real-world datasets from Yelp Amazon, validating effectiveness GCM benefits for CARS.