Abstract:Nowadays, graph collaborative filtering is the most practical and classical method for recommender systems, which learns user preferences through historical interactions between users and items. The recent combination of contrastive learning and recommender systems has led to a continuous improvement in the performance of graph collaborative filtering. A variety of data augmentation methods have also emerged. However, some graph augmentation methods are proven to lose graph information, such as the way of dropout on the graph structure. Inspired by different adversarial samples and different data augmentation methods, we adopt the method of adding interference patterns to the embedding space, as opposed to other models, we take a partial sampling of the embedding and then add the partially sampled data into the embedding, thus forming different sub-views. The extensive experimental results show that our experiments can improve the performance of the model more compared to other contrastive learning methods with our data enhancement method.

Towards Graph Contrastive Learning for Recommendation with Sampling Embedding Perturbation