Personal Recommendation Using Deep Recurrent Neural Networks in NetEase (ICDE 2016 Paper)

The 2016 paper Personal Recommendation Using Deep Recurrent Neural Networks in NetEase proposes a session-based recommender system for e-commerce based on a deep neural network combining a feed-forward neural network (FNN) and a recurrent neural network (RNN). The FNN part represents a historical-data-based collaborative filtering, and the RNN part captures the user’s purchase intent from the webpage visit sequence in the current session.

Resources

My presentation at my ML study group.