Graph Federated Learning for Personalized Privacy Recommendation

2025-08-08

AI Summary

Existing federated recommendation systems (FedRecs) assume uniform privacy requirements across all users, overlooking the potential performance gains from leveraging publicly shared user interaction data. Method: We propose GFed-PP, the first FedRec framework incorporating user-level privacy grading—distinguishing *private users* (whose interaction data remains local and non-shared) from *public users* (whose interaction data is voluntarily shared). Based on this distinction, we construct two heterogeneous graphs: a user-item interaction graph and a user-relational graph. A lightweight graph convolutional network (GCN) is deployed locally on clients to learn personalized embeddings, while server-side graph aggregation enables collaborative optimization. Contribution/Results: GFed-PP jointly addresses privacy heterogeneity and recommendation accuracy. Extensive experiments on five real-world datasets demonstrate significant improvements over state-of-the-art methods, with substantial gains in recommendation accuracy—validating both privacy adaptivity and performance enhancement.

Technology Category

• Machine Learning: Privacy
• Search and Optimization: Learning to Search
• Data Mining & Knowledge Management: Recommender Systems

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Security and Privacy: Privacy-enhancing technologies

Abstract

Federated recommendation systems (FedRecs) have gained significant attention for providing privacy-preserving recommendation services. However, existing FedRecs assume that all users have the same requirements for privacy protection, i.e., they do not upload any data to the server. The approaches overlook the potential to enhance the recommendation service by utilizing publicly available user data. In real-world applications, users can choose to be private or public. Private users' interaction data is not shared, while public users' interaction data can be shared. Inspired by the issue, this paper proposes a novel Graph Federated Learning for Personalized Privacy Recommendation (GFed-PP) that adapts to different privacy requirements while improving recommendation performance. GFed-PP incorporates the interaction data of public users to build a user-item interaction graph, which is then used to form a user relationship graph. A lightweight graph convolutional network (GCN) is employed to learn each user's user-specific personalized item embedding. To protect user privacy, each client learns the user embedding and the scoring function locally. Additionally, GFed-PP achieves optimization of the federated recommendation framework through the initialization of item embedding on clients and the aggregation of the user relationship graph on the server. Experimental results demonstrate that GFed-PP significantly outperforms existing methods for five datasets, offering superior recommendation accuracy without compromising privacy. This framework provides a practical solution for accommodating varying privacy preferences in federated recommendation systems.

Problem

• Addresses varying user privacy requirements in federated recommendation systems
• Enhances recommendation accuracy while maintaining privacy protection
• Utilizes public user data to improve personalized item embeddings

Innovation

• Builds user-item and user relationship graphs
• Uses lightweight GCN for personalized embeddings
• Optimizes federated framework with local and server components