Recommended System: Attentive Neural Collaborative Filtering

In recent years, neural networks have yielded immense success on speech recognition, computer vision and natural language processing. However, the exploration of neural networks on recommender systems has received relatively less scrutiny. In this work, we strive to develop techniques based on neural networks to tackle the key problem in recommendation - collaborative filtering - on the basis of implicit feedback. Although some recent work has employed deep learning for recommendation, they primarily used it to model auxiliary information, such as textual descriptions of items and acoustic features of music. When it comes to model the key factor in collaborative filtering - the interaction between users and item features, they still resorted to matrix factorization and applied an inner product on the latent features of users and items. And the collaboration signal hidden in the user-item interaction is not encoded during the embedding process. Therefore, the resulting embedding may not be sufficient to capture the collaborative filtering effect. By replacing the inner product with a neural architecture that can learn an arbitrary function from data, we present a general method named ANCF(Attention Neural network Collaborative Filtering). ANCF captures collaborative filtering signals and refines the embedding of users and items according to the structure of the graph. By introducing an attention mechanism, the user vector and the item vector are learned on the user-item interaction graph, neighbor interaction information is aggregated to encode, and the embedding is propagated on the user-item interaction graph. This makes it possible to explicitly inject user-item collaboration signals into the embedding process. Extensive experiments conducted on two real world datasets show that ANCF's recall and ndcg have increased by 30% and 35%, so our proposed ANCF method has been significantly improved over the state-of-the-art method. Empirical evidence shows that using deeper layers of neural networks offers better recommendation performance.