Learning Binarized Representations with Pseudo-positive Sample Enhancement for Efficient Graph Collaborative Filtering

2025-06-03

AI Summary

To address information loss inherent in binary embeddings for graph collaborative filtering, this paper proposes a high-discriminative binarization learning framework. Methodologically, it unifies graph neural networks, binary embedding learning, pseudo-label generation, and multi-granularity supervision modeling. Key contributions include: (1) a novel pseudo-positive sample augmentation mechanism that jointly leverages observed user-item interactions and synthetically generated implicit embedding samples; (2) a fine-grained inference distillation paradigm integrating knowledge distillation with contrastive learning to construct strong supervisory signals; and (3) an embedding sample synthesis strategy to enhance matching accuracy of binary representations. Extensive experiments on five public benchmarks demonstrate state-of-the-art performance: the proposed method achieves 1–10% absolute improvement in Recall@20 over BiGeaR, the previous best binary graph collaborative filtering approach.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Reasoning under Uncertainty: Graphical Models
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Learning vectorized embeddings is fundamental to many recommender systems for user-item matching. To enable efficient online inference, representation binarization, which embeds latent features into compact binary sequences, has recently shown significant promise in optimizing both memory usage and computational overhead. However, existing approaches primarily focus on numerical quantization, neglecting the associated information loss, which often results in noticeable performance degradation. To address these issues, we study the problem of graph representation binarization for efficient collaborative filtering. Our findings indicate that explicitly mitigating information loss at various stages of embedding binarization has a significant positive impact on performance. Building on these insights, we propose an enhanced framework, BiGeaR++, which specifically leverages supervisory signals from pseudo-positive samples, incorporating both real item data and latent embedding samples. Compared to its predecessor BiGeaR, BiGeaR++ introduces a fine-grained inference distillation mechanism and an effective embedding sample synthesis approach. Empirical evaluations across five real-world datasets demonstrate that the new designs in BiGeaR++ work seamlessly well with other modules, delivering substantial improvements of around 1%-10% over BiGeaR and thus achieving state-of-the-art performance compared to the competing methods. Our implementation is available at https://github.com/QueYork/BiGeaR-SS.

Problem

• Addressing information loss in graph representation binarization for collaborative filtering
• Enhancing binary embeddings using pseudo-positive samples and supervision
• Improving efficiency and performance in recommender systems via binarized representations

Innovation

• Binarized graph representations for efficient filtering
• Pseudo-positive samples enhance supervisory signals
• Fine-grained inference distillation improves performance