Abstract
Owing to the strength in learning representation of the high-order connectivity of graph neural networks (GNN), GNN-based collaborative filtering has been widely adopted in recommender systems. Furthermore, to overcome the data sparsity problem, some recent GNN-based models attempt to incorporate social information and to design contrastive learning as an auxiliary task to assist the primary recommendation task. Existing GNN and contrastive-learning-based recommendation models learn user and item representations in a symmetrical way and utilize social information and contrastive learning in a complex manner. The above two strategies lead to these models being either ineffective for datasets with a serious imbalance between users and items or inefficient for datasets with too many users and items. In this work, we propose a contrastive graph learning (CGL) model, which combines social information and contrastive learning in a simple and powerful way. CGL consists of three modules: diffusion, readout, and prediction. The diffusion module recursively aggregates and integrates social information and interest information to learn representations of users and items. The readout module takes the average value of user embeddings from all diffusion layers and item embeddings at the last diffusion layer as readouts of users and items, respectively. The prediction module calculates prediction rating scores with an interest graph to emphasize interest information. Three different losses are designed to ensure the function of each module. Extensive experiments on three benchmark datasets are implemented to validate the effectiveness of our model.
Highlights
This paper explores how to overcome the above limitations in existing recommendation models based on graph neural networks (GNN) and contrastive learning and proposes a contrastive graph learning (CGL) model for social recommendation
Matrix factorization (MF)-Bayesian personalized ranking (BPR) [42]: It exploits how to represent users and items in a low-dimensional latent space, which is optimized by the BPR loss
LightGCN [22]: It is a light version of the GNN-based model and only performs aggregation operations
Summary
With the rapid development of networks, it is becoming harder and harder for a user to extract useful information from a mass of redundant information. To achieve better recommendation performance, many existing recommendation models [19–21] encode node embeddings by a GNN framework and simultaneously resort to contrastive learning in the learning process. Existing GNN-based models [11, 22] learn representations of users and items in the same way and do not consider the different sparsities of users and items. On the one hand, existing social recommendation models [19, 25] utilize contrastive learning in a quite complex manner such as encoding hypergraph and data augmentation. It is reasonable to characterize each user’s preference by item aggregation and friend aggregation separately and to require user representations learned from the two views (user–item graph or user–user graph) to have consistent agreement [26] This argument motivates us to simplify the contrastive learning task between social user embeddings and interest user embeddings.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
