Leave No One Behind: Fairness-Aware Cross-Domain Recommender Systems for Non-Overlapping Users

2025-07-23

AI Summary

In cross-domain recommendation (CDR), existing methods rely on overlapping users to transfer knowledge, resulting in significantly inferior performance for non-overlapping users and introducing fairness bias. To address this, we propose FairCDR—a framework that bridges the representation gap for non-overlapping users by generating faithful and personalized synthetic source-domain user embeddings. Its key contributions are: (1) a dual-attention mechanism jointly modeling user–item interactions and cross-domain semantic associations; (2) a distribution-alignment limiter constraining the synthetic user generation process to ensure fidelity and domain consistency; and (3) a model-agnostic design enabling plug-and-play integration with diverse CDR architectures. Extensive experiments on three public benchmarks demonstrate that FairCDR substantially improves recommendation accuracy for non-overlapping users (average +12.7% NDCG@10), effectively eliminating the performance gap without compromising overall recommendation quality.

Technology Category

• Application Domains: Humanities & Computational Social Science
• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Humans and AI: Intelligent User Interfaces

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Economics and fairness of platforms and recommendation systems

Abstract

Cross-domain recommendation (CDR) methods predominantly leverage overlapping users to transfer knowledge from a source domain to a target domain. However, through empirical studies, we uncover a critical bias inherent in these approaches: while overlapping users experience significant enhancements in recommendation quality, non-overlapping users benefit minimally and even face performance degradation. This unfairness may erode user trust, and, consequently, negatively impact business engagement and revenue. To address this issue, we propose a novel solution that generates virtual source-domain users for non-overlapping target-domain users. Our method utilizes a dual attention mechanism to discern similarities between overlapping and non-overlapping users, thereby synthesizing realistic virtual user embeddings. We further introduce a limiter component that ensures the generated virtual users align with real-data distributions while preserving each user's unique characteristics. Notably, our method is model-agnostic and can be seamlessly integrated into any CDR model. Comprehensive experiments conducted on three public datasets with five CDR baselines demonstrate that our method effectively mitigates the CDR non-overlapping user bias, without loss of overall accuracy. Our code is publicly available at https://github.com/WeixinChen98/VUG.

Problem

• Address bias in cross-domain recommendations for non-overlapping users
• Enhance fairness without sacrificing overall recommendation accuracy
• Generate virtual source-domain users to improve non-overlapping user performance

Innovation

• Generates virtual source-domain users for fairness
• Uses dual attention mechanism for user similarity
• Model-agnostic limiter aligns with real-data distributions