Flow Matching for Collaborative Filtering

2025-02-11

AI Summary

Existing generative collaborative filtering (CF) models face two key challenges: inaccurate posterior approximation and poor compatibility with the discrete, binary nature of implicit feedback. To address these, we propose FlowCF—the first discrete flow matching generative model specifically designed for CF. Our method tackles the problem by: (1) introducing a behavior-guided prior that explicitly captures user behavior sparsity and heterogeneity; and (2) constructing a discrete probability flow framework that performs flow matching directly in the binary feedback space, thereby avoiding mismatches induced by continuous relaxations. FlowCF combines theoretical rigor with practical efficiency: it achieves state-of-the-art recommendation accuracy across multiple benchmark datasets and attains the fastest inference speed among comparable models, significantly enhancing feasibility for industrial deployment.

Technology Category

• Machine Learning: Learning Preferences or Rankings
• Reasoning under Uncertainty: Relational Probabilistic Models
• Data Mining & Knowledge Management: Recommender Systems

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs

Abstract

Generative models have shown great promise in collaborative filtering by capturing the underlying distribution of user interests and preferences. However, existing approaches struggle with inaccurate posterior approximations and misalignment with the discrete nature of recommendation data, limiting their expressiveness and real-world performance. To address these limitations, we propose FlowCF, a novel flow-based recommendation system leveraging flow matching for collaborative filtering. We tailor flow matching to the unique challenges in recommendation through two key innovations: (1) a behavior-guided prior that aligns with user behavior patterns to handle the sparse and heterogeneous user-item interactions, and (2) a discrete flow framework to preserve the binary nature of implicit feedback while maintaining the benefits of flow matching, such as stable training and efficient inference. Extensive experiments demonstrate that FlowCF achieves state-of-the-art recommendation accuracy across various datasets with the fastest inference speed, making it a compelling approach for real-world recommender systems.

Problem

• Improves posterior approximations in recommendation models
• Aligns with discrete nature of recommendation data
• Handles sparse and heterogeneous user-item interactions

Innovation

• Flow-based recommendation system
• Behavior-guided prior alignment
• Discrete flow framework