Co-clustering for Federated Recommender System

TL;DR

This paper delves into the inefficiencies of the K-Means method in client grouping, attributing failures due to the high dimensionality as well as data sparsity occurring in FRS, and proposes CoFedRec, a novel Co-clustering Federated Recommendation mechanism to address clients heterogeneity and enhance the collaborative filtering within the federated framework.

Abstract

As data privacy and security attract increasing attention, Federated Recommender System (FRS) offers a solution that strikes a balance between providing high-quality recommendations and preserving user privacy. However, the presence of statistical heterogeneity in FRS, commonly observed due to personalized decision-making patterns, can pose challenges. To address this issue and maximize the benefit of collaborative filtering (CF) in FRS, it is intuitive to consider clustering clients (users) as well as items into different groups and learning group-specific models. Existing methods either resort to client clustering via user representations-risking privacy leakage, or employ classical clustering strategies on item embeddings or gradients, which we found are plagued by the curse of dimensionality. In this paper, we delve into the inefficiencies of the K-Means method in client grouping, attributing failures due to the high dimensionality as well as data sparsity occurring in FRS, and propose CoFedRec, a novel Co-clustering Federated Recommendation mechanism, to address clients heterogeneity and enhance the collaborative filtering within the federated framework. Specifically, the server initially formulates an item membership from the client-provided item networks. Subsequently, clients are grouped regarding a specific item category picked from the item membership during each communication round, resulting in an intelligently aggregated group model. Meanwhile, to comprehensively capture the global inter-relationships among items, we incorporate an additional supervised contrastive learning term based on the server-side generated item membership into the local training phase for each client. Extensive experiments on four datasets are provided, which verify the effectiveness of the proposed CoFedRec.

Figures (6)

• Figure 1: The overall framework of CoFedRec. The pink dots represent the individual models uploaded by participant clients and the green dots are the item embedding vectors of the global aggregation results. Two key parts in CoFedRec are (i) Co-clustering mechanism to cluster participant clients into similar group and dissimilar group and an intelligent group model is generated within the similar group; (ii) Supervised contrastive term upon the global item membership is integrated into the loss function of the local training phase to include the global item insights.
• Figure 2: Distribution of Clients' Participation Rounds on MovieLens-100k and MovieLens-1M Datasets
• Figure 3: Visulization of the clustering results on items.
• Figure 4: Effect of the number of the item clusters.
• Figure 5: Visualization for clustering results on MovieLens-100K via K-Means.