Predicting Dynamic Embedding Trajectory in Temporal Interaction Networks

Modeling sequential interactions between users and items/products is crucial in domains such as e-commerce, social networking, and education. Representation learning presents an attractive opportunity