Causal-Invariant Cross-Domain Out-of-Distribution Recommendation

2025-05-22

AI Summary

Cross-domain recommendation (CDR) faces out-of-distribution (OOD) challenges arising jointly from cross-domain distribution shift (CDDS) between source and target domains and single-domain distribution shift (SDDS) within the target domain, severely hindering knowledge transfer and generalization. To address this, we propose a causal invariant preference modeling framework that introduces a novel two-level causal structure modeling approach, integrated with an LLM-guided confounder discovery module to jointly disentangle CDDS and SDDS. Our method synergistically combines causal inference, invariant risk minimization, and LLM-enhanced causal discovery to achieve robust cross-domain preference learning. Evaluated on two real-world datasets, our approach consistently outperforms state-of-the-art methods, achieving an average 12.7% improvement in recommendation accuracy across diverse OOD scenarios. This work establishes an interpretable and generalizable causal paradigm for CDR under OOD conditions.

Technology Category

• Machine Learning: Causal Learning
• Reasoning under Uncertainty: Causality
• Knowledge Representation and Reasoning: Preferences

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
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Cross-Domain Recommendation (CDR) aims to leverage knowledge from a relatively data-richer source domain to address the data sparsity problem in a relatively data-sparser target domain. While CDR methods need to address the distribution shifts between different domains, i.e., cross-domain distribution shifts (CDDS), they typically assume independent and identical distribution (IID) between training and testing data within the target domain. However, this IID assumption rarely holds in real-world scenarios due to single-domain distribution shift (SDDS). The above two co-existing distribution shifts lead to out-of-distribution (OOD) environments that hinder effective knowledge transfer and generalization, ultimately degrading recommendation performance in CDR. To address these co-existing distribution shifts, we propose a novel Causal-Invariant Cross-Domain Out-of-distribution Recommendation framework, called CICDOR. In CICDOR, we first learn dual-level causal structures to infer domain-specific and domain-shared causal-invariant user preferences for tackling both CDDS and SDDS under OOD environments in CDR. Then, we propose an LLM-guided confounder discovery module that seamlessly integrates LLMs with a conventional causal discovery method to extract observed confounders for effective deconfounding, thereby enabling accurate causal-invariant preference inference. Extensive experiments on two real-world datasets demonstrate the superior recommendation accuracy of CICDOR over state-of-the-art methods across various OOD scenarios.

Problem

• Address cross-domain and single-domain distribution shifts in recommendation systems
• Infer causal-invariant user preferences under out-of-distribution environments
• Integrate LLMs with causal discovery for effective deconfounding in recommendations

Innovation

• Dual-level causal structures for invariant preferences
• LLM-guided confounder discovery module
• Effective deconfounding for accurate preference inference