MDAP: A Multi-view Disentangled and Adaptive Preference Learning Framework for Cross-Domain Recommendation

TL;DR

The paper introduces MDAP, a novel framework for cross-domain recommendation that captures diverse user preferences through multi-view disentangled learning and adaptive feature combination, improving recommendation accuracy especially in sparse data scenarios.

Contribution

The paper proposes a multi-view disentangled and adaptive preference learning framework that enhances cross-domain recommendation by effectively capturing and combining user preferences.

Findings

Outperforms state-of-the-art CDR models on benchmark datasets

Improves recommendation accuracy in sparse data scenarios

Provides deeper insights into user behavior across domains

Abstract

Cross-domain Recommendation systems leverage multi-domain user interactions to improve performance, especially in sparse data or new user scenarios. However, CDR faces challenges such as effectively capturing user preferences and avoiding negative transfer. To address these issues, we propose the Multi-view Disentangled and Adaptive Preference Learning (MDAP) framework. Our MDAP framework uses a multiview encoder to capture diverse user preferences. The framework includes a gated decoder that adaptively combines embeddings from different views to generate a comprehensive user representation. By disentangling representations and allowing adaptive feature selection, our model enhances adaptability and effectiveness. Extensive experiments on benchmark datasets demonstrate that our method significantly outperforms state-of-the-art CDR and single-domain models, providing more accurate recommendations and deeper insights into user behavior across different domains.