Personalized Federated Knowledge Graph Embedding with Client-Wise Relation Graph
Xiaoxiong Zhang, Zhiwei Zeng, Xin Zhou, Dusit Niyato, Zhiqi Shen
TL;DR
The paper addresses semantic disparity in federated knowledge graph embedding by introducing PFedEG, which builds a client-wise relation graph to produce personalized supplementary knowledge for each client. By combining server-side affinity-weighted aggregation with local KGE learning, PFedEG aligns global information with each KG's semantics, improving link-prediction performance across multiple datasets and KGE models. Key contributions include two affinity strategies, a regularized client update, and comprehensive experiments plus ablations demonstrating improvements over state-of-the-art FKGE methods. The approach enhances privacy-preserving knowledge sharing in heterogeneous multi-source KGs and opens avenues for optimizing inter-client relation weights.
Abstract
Federated Knowledge Graph Embedding (FKGE) has recently garnered considerable interest due to its capacity to extract expressive representations from distributed knowledge graphs, while concurrently safeguarding the privacy of individual clients. Existing FKGE methods typically harness the arithmetic mean of entity embeddings from all clients as the global supplementary knowledge, and learn a replica of global consensus entities embeddings for each client. However, these methods usually neglect the inherent semantic disparities among distinct clients. This oversight not only results in the globally shared complementary knowledge being inundated with too much noise when tailored to a specific client, but also instigates a discrepancy between local and global optimization objectives. Consequently, the quality of the learned embeddings is compromised. To address this, we propose Personalized Federated knowledge graph Embedding with client-wise relation Graph (PFedEG), a novel approach that employs a client-wise relation graph to learn personalized embeddings by discerning the semantic relevance of embeddings from other clients. Specifically, PFedEG learns personalized supplementary knowledge for each client by amalgamating entity embedding from its neighboring clients based on their "affinity" on the client-wise relation graph. Each client then conducts personalized embedding learning based on its local triples and personalized supplementary knowledge. We conduct extensive experiments on four benchmark datasets to evaluate our method against state-of-the-art models and results demonstrate the superiority of our method.