Graph Signal Processing for Cross-Domain Recommendation

2024-07-17

arXiv.org

AI Summary

Cross-domain recommendation faces key challenges including data sparsity, cold-start issues, overreliance on overlapping users, and neglect of domain-specific characteristics. Method: This paper pioneers the integration of graph signal processing (GSP) into cross-domain recommendation, proposing a novel similarity graph construction method that does not require strong assumptions about user overlap. It constructs a dual-source tunable similarity graph by jointly modeling target-domain-exclusive structural information and source-domain-bridging similarities, and introduces personalized graph signal filtering alongside a source-target joint embedding mechanism to unify intra- and cross-domain recommendation. Contribution/Results: Extensive experiments on multiple benchmark datasets demonstrate that our approach consistently outperforms state-of-the-art encoder-based models—especially under severe overlap scarcity (e.g., <10% overlapping users)—validating its superior generalization capability and practical deployability.

Technology Category

• Application Domains: Humanities & Computational Social Science
• Machine Learning: Graph-based Machine Learning
• 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: Graph embeddings and representation learning for Web-related graphs
• Search and Retrieval-Augmented AI: Vertical and domain-specific search

Abstract

Cross-domain recommendation (CDR) extends conventional recommender systems by leveraging user-item interactions from dense domains to mitigate data sparsity and the cold start problem. While CDR offers substantial potential for enhancing recommendation performance, most existing CDR methods suffer from sensitivity to the ratio of overlapping users and intrinsic discrepancy between source and target domains. To overcome these limitations, in this work, we explore the application of graph signal processing (GSP) in CDR scenarios. We propose CGSP, a unified CDR framework based on GSP, which employs a cross-domain similarity graph constructed by flexibly combining target-only similarity and source-bridged similarity. By processing personalized graph signals computed for users from either the source or target domain, our framework effectively supports both inter-domain and intra-domain recommendations. Our empirical evaluation demonstrates that CGSP consistently outperforms various encoder-based CDR approaches in both intra-domain and inter-domain recommendation scenarios, especially when the ratio of overlapping users is low, highlighting its significant practical implication in real-world applications.

Problem

• Overcoming data sparsity and cold-start in cross-domain recommendation
• Addressing domain sensitivity and lack of overlapping user limitations
• Developing unified adaptive framework for intra-domain and inter-domain recommendations

Innovation

• Graph signal processing for unified recommendations
• Adaptive cross-domain similarity graph integration
• Personalized graph signals for source and target domains