Search papers, labs, and topics across Lattice.
The paper introduces FedRecGEL, a federated recommendation framework designed to learn generalized item embeddings in the presence of heterogeneous and sparse local data. FedRecGEL reformulates federated recommendation as a multi-task learning problem centered on items and employs sharpness-aware minimization (SAM) to improve generalization and stabilize training. Experiments on four datasets demonstrate that FedRecGEL significantly improves federated recommendation performance compared to existing methods.
Federated recommendation systems can learn more robust item embeddings, and thus perform better, by using sharpness-aware minimization to combat data heterogeneity and sparsity.
Federated recommender systems enable collaborative model training while keeping user interaction data local and sharing only essential model parameters, thereby mitigating privacy risks. However, existing methods overlook a critical issue, i.e., the stable learning of a generalized item embedding throughout the federated recommender system training process. Item embedding plays a central role in facilitating knowledge sharing across clients. Yet, under the cross-device setting, local data distributions exhibit significant heterogeneity and sparsity, exacerbating the difficulty of learning generalized embeddings. These factors make the stable learning of generalized item embeddings both indispensable for effective federated recommendation and inherently difficult to achieve. To fill this gap, we propose a new federated recommendation framework, named Federated Recommendation with Generalized Embedding Learning (FedRecGEL). We reformulate the federated recommendation problem from an item-centered perspective and cast it as a multi-task learning problem, aiming to learn generalized embeddings throughout the training procedure. Based on theoretical analysis, we employ sharpness-aware minimization to address the generalization problem, thereby stabilizing the training process and enhancing recommendation performance. Extensive experiments on four datasets demonstrate the effectiveness of FedRecGEL in significantly improving federated recommendation performance. Our code is available at https://github.com/anonymifish/FedRecGEL.
