Intrinsic and Extrinsic Factor Disentanglement for Recommendation in Various Context Scenarios

2025-03-05

AI Summary

In recommender systems, user behavior arises from both stable intrinsic preferences and extrinsic motivations jointly influenced by multiple contextual factors (e.g., time, location); however, existing methods decouple these solely based on a single pre-defined context, ignoring contextual interactions. This work proposes the first general disentanglement framework that jointly models multi-context interactions. It enhances the robustness of intrinsic factors via context-invariant contrastive learning and achieves fine-grained modeling of extrinsic factors through a multi-factor disentanglement neural network coupled with joint context embedding and attention mechanisms. Evaluated on real-world datasets, the method achieves up to a 4% improvement in NDCG, significantly enhancing cross-context recommendation accuracy and the interpretability of learned disentangled factors.

Technology Category

• Machine Learning: Learning Preferences or Rankings
• Data Mining & Knowledge Management: Recommender Systems
• Multiagent Systems: Adversarial Agents

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• 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

In recommender systems, the patterns of user behaviors (e.g., purchase, click) may vary greatly in different contexts (e.g., time and location). This is because user behavior is jointly determined by two types of factors: intrinsic factors, which reflect consistent user preference, and extrinsic factors, which reflect external incentives that may vary in different contexts. Differentiating between intrinsic and extrinsic factors helps learn user behaviors better. However, existing studies have only considered differentiating them from a single, pre-defined context (e.g., time or location), ignoring the fact that a user's extrinsic factors may be influenced by the interplay of various contexts at the same time. In this paper, we propose the Intrinsic-Extrinsic Disentangled Recommendation (IEDR) model, a generic framework that differentiates intrinsic from extrinsic factors considering various contexts simultaneously, enabling more accurate differentiation of factors and hence the improvement of recommendation accuracy. IEDR contains a context-invariant contrastive learning component to capture intrinsic factors, and a disentanglement component to extract extrinsic factors under the interplay of various contexts. The two components work together to achieve effective factor learning. Extensive experiments on real-world datasets demonstrate IEDR's effectiveness in learning disentangled factors and significantly improving recommendation accuracy by up to 4% in NDCG.

Problem

• Differentiates intrinsic and extrinsic factors in user behavior.
• Considers multiple contexts simultaneously for accurate factor disentanglement.
• Improves recommendation accuracy by up to 4% in NDCG.

Innovation

• Disentangles intrinsic and extrinsic user factors
• Uses context-invariant contrastive learning
• Improves recommendation accuracy by 4%