Enhancing Social Recommendation With Adversarial Graph Convolutional Networks

Recent reports from industry show that social recommender systems consistently fail in practice. According to the negative findings, the failure is attributed to: (1) a majority of users only have a very limited number of neighbors in social networks and can hardly benefit from relations; (2) social relations are noisy but they are often indiscriminately used; (3) social relations are assumed to be universally applicable to multiple scenarios while they are actually multi-faceted and show heterogeneous strengths in different scenarios. Most existing social recommendation models only consider the homophily in social networks and neglect these drawbacks. In this paper we propose a deep adversarial framework based on graph convolutional networks (GCN) to address these problems. Concretely, for the relation sparsity and noises problems, a GCN-based autoencoder is developed to augment the relation data by encoding high-order and complex connectivity patterns, and meanwhile is optimized subject to the constraint of reconstructing the original social profile to guarantee the validity of new identified neighborhood. After obtaining enough purified social relations for each user, a GCN-based attentive social recommendation module is designed to capture the heterogeneous strengths of social relations. These designs deal with the three problems faced by social recommender systems respectively. Finally, we adopt adversarial training to unify and intensify all components by playing a minimax game and ensure a coordinated effort to enhance social recommendation. Experimental results on multiple open datasets demonstrate the superiority of our framework and the ablation study confirms the importance and effectiveness of each component.