Decentralized Graph Neural Network for Privacy-Preserving Recommendation
This addresses privacy concerns in recommender systems for users, though it appears incremental as it builds on existing decentralized and federated GNN methods.
The paper tackles the challenge of building a graph neural network-based recommender system without violating user privacy by proposing DGREC, a decentralized GNN that includes a local differential privacy mechanism, and it demonstrates consistent superiority in experiments on three public datasets.
Building a graph neural network (GNN)-based recommender system without violating user privacy proves challenging. Existing methods can be divided into federated GNNs and decentralized GNNs. But both methods have undesirable effects, i.e., low communication efficiency and privacy leakage. This paper proposes DGREC, a novel decentralized GNN for privacy-preserving recommendations, where users can choose to publicize their interactions. It includes three stages, i.e., graph construction, local gradient calculation, and global gradient passing. The first stage builds a local inner-item hypergraph for each user and a global inter-user graph. The second stage models user preference and calculates gradients on each local device. The third stage designs a local differential privacy mechanism named secure gradient-sharing, which proves strong privacy-preserving of users' private data. We conduct extensive experiments on three public datasets to validate the consistent superiority of our framework.