Causality Enhancement for Cross-Domain Recommendation

2025-10-16

AI Summary

Cross-domain recommendation suffers from negative transfer and causal confounding due to task inconsistency between source and target domains, particularly when unbiased causal labels are unavailable—hindering faithful modeling of underlying causal mechanisms. To address this, we propose the first causally enhanced cross-domain recommendation framework: (i) it formalizes the recommendation process as a causal graph; (ii) constructs a heuristic causal-aware dataset to approximate counterfactual supervision; and (iii) introduces a model-agnostic partially labeled causal loss, enabling unbiased generalization to unseen cross-domain patterns directly from biased observational data. Extensive experiments demonstrate significant improvements in target-domain recommendation performance across multiple public benchmarks and real-world production systems. The framework exhibits strong generality and has been successfully deployed in industry, delivering substantial business gains.

Technology Category

• Machine Learning: Causal Learning
• Reasoning under Uncertainty: Causality
• Application Domains: Humanities & Computational Social Science

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• 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 forms a crucial component in recommendation systems. It leverages auxiliary information through source domain tasks or features to enhance target domain recommendations. However, incorporating inconsistent source domain tasks may result in insufficient cross-domain modeling or negative transfer. While incorporating source domain features without considering the underlying causal relationships may limit their contribution to final predictions. Thus, a natural idea is to directly train a cross-domain representation on a causality-labeled dataset from the source to target domain. Yet this direction has been rarely explored, as identifying unbiased real causal labels is highly challenging in real-world scenarios. In this work, we attempt to take a first step in this direction by proposing a causality-enhanced framework, named CE-CDR. Specifically, we first reformulate the cross-domain recommendation as a causal graph for principled guidance. We then construct a causality-aware dataset heuristically. Subsequently, we derive a theoretically unbiased Partial Label Causal Loss to generalize beyond the biased causality-aware dataset to unseen cross-domain patterns, yielding an enriched cross-domain representation, which is then fed into the target model to enhance target-domain recommendations. Theoretical and empirical analyses, as well as extensive experiments, demonstrate the rationality and effectiveness of CE-CDR and its general applicability as a model-agnostic plugin. Moreover, it has been deployed in production since April 2025, showing its practical value in real-world applications.

Problem

• Enhancing cross-domain recommendations using causal relationships between domains
• Addressing negative transfer from inconsistent source domain tasks in recommendations
• Developing unbiased causal loss for cross-domain pattern generalization

Innovation

• Proposes a causality-enhanced framework for cross-domain recommendation
• Constructs a causality-aware dataset using heuristic methods
• Derives an unbiased Partial Label Causal Loss for generalization