Beyond Negative Transfer: Disentangled Preference-Guided Diffusion for Cross-Domain Sequential Recommendation

2025-08-30

AI Summary

Cross-domain sequential recommendation (CDSR) suffers from two key challenges: negative transfer due to domain heterogeneity—causing preference conflicts—and behavioral noise—such as spurious clicks—that further degrades modeling fidelity. To address these, we propose the Decoupled Preference-guided Diffusion Model (DPDM), the first diffusion-based framework for CDSR. DPDM introduces a preference-decoupling encoder to separate domain-invariant and domain-specific preferences, and explicitly models both components during the reverse diffusion process to enable precise cross-domain knowledge transfer and sequence denoising. It integrates a temporal-aware reverse diffusion network with a generative denoising framework, effectively mitigating negative transfer while enhancing dynamic preference modeling. Extensive experiments on multiple real-world datasets demonstrate that DPDM consistently outperforms state-of-the-art baselines, validating its effectiveness, robustness, and superiority in handling noisy, heterogeneous sequential behaviors across domains.

Technology Category

• Machine Learning: Learning Preferences or Rankings
• Knowledge Representation and Reasoning: Preferences
• Computer Vision: Diffusion Models for Vision

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
• Web Mining and Content Analysis: Content-based information diffusion

Abstract

Cross-Domain Sequential Recommendation (CDSR) leverages user behaviors across domains to enhance recommendation quality. However, naive aggregation of sequential signals can introduce conflicting domain-specific preferences, leading to negative transfer. While Sequential Recommendation (SR) already suffers from noisy behaviors such as misclicks and impulsive actions, CDSR further amplifies this issue due to domain heterogeneity arising from diverse item types and user intents. The core challenge is disentangling three intertwined signals: domain-invariant preferences, domain-specific preferences, and noise. Diffusion Models (DMs) offer a generative denoising framework well-suited for disentangling complex user preferences and enhancing robustness to noise. Their iterative refinement process enables gradual denoising, making them effective at capturing subtle preference signals. However, existing applications in recommendation face notable limitations: sequential DMs often conflate shared and domain-specific preferences, while cross-domain collaborative filtering DMs neglect temporal dynamics, limiting their ability to model evolving user preferences. To bridge these gaps, we propose extbf{DPG-Diff}, a novel Disentangled Preference-Guided Diffusion Model, the first diffusion-based approach tailored for CDSR, to or best knowledge. DPG-Diff decomposes user preferences into domain-invariant and domain-specific components, which jointly guide the reverse diffusion process. This disentangled guidance enables robust cross-domain knowledge transfer, mitigates negative transfer, and filters sequential noise. Extensive experiments on real-world datasets demonstrate that DPG-Diff consistently outperforms state-of-the-art baselines across multiple metrics.

Problem

• Disentangling domain-invariant and domain-specific user preferences
• Mitigating negative transfer from conflicting cross-domain signals
• Filtering sequential noise amplified by domain heterogeneity

Innovation

• Decomposes user preferences into domain-invariant and domain-specific components
• Uses disentangled guidance to jointly steer reverse diffusion process
• Filters sequential noise while enabling robust cross-domain knowledge transfer