Vertical Federated Graph Neural Network for Recommender System

TL;DR

This study proposes the first vertical federated GNN-based recommender system, called VerFedGNN, and designs a framework to transmit the summation of neighbor embeddings using random projection, and gradients of public parameter perturbed by ternary quantization mechanism.

Abstract

Conventional recommender systems are required to train the recommendation model using a centralized database. However, due to data privacy concerns, this is often impractical when multi-parties are involved in recommender system training. Federated learning appears as an excellent solution to the data isolation and privacy problem. Recently, Graph neural network (GNN) is becoming a promising approach for federated recommender systems. However, a key challenge is to conduct embedding propagation while preserving the privacy of the graph structure. Few studies have been conducted on the federated GNN-based recommender system. Our study proposes the first vertical federated GNN-based recommender system, called VerFedGNN. We design a framework to transmit: (i) the summation of neighbor embeddings using random projection, and (ii) gradients of public parameter perturbed by ternary quantization mechanism. Empirical studies show that VerFedGNN has competitive prediction accuracy with existing privacy preserving GNN frameworks while enhanced privacy protection for users' interaction information.

Figures (5)

• Figure 1: Overall framework of VerFedGNN
• Figure 2: RMSE with varying participation rate $\alpha$.
• Figure 3: RMSE with varying privacy budget $\frac{1}{r}$.
• Figure 4: RMSE by inverse dimension reduction ratio $N_u/q$.
• Figure 5: Communication cost by user size and dimension for GCN

Theorems & Definitions (24)

• Definition 3.1
• Definition 4.1
• Definition 4.2
• Definition 5.1
• Lemma 5.2
• Theorem 5.3
• Theorem 5.4
• proof
• Remark 5.5
• Theorem 5.6