Unveiling Contrastive Learning's Capability of Neighborhood Aggregation for Collaborative Filtering

2025-04-14

AI Summary

This work uncovers the intrinsic mechanism by which graph contrastive learning (GCL) implicitly performs neighborhood aggregation in collaborative filtering: we theoretically prove that its optimization objective is equivalent to gradient updates of graph convolutional layers. Addressing the limitation of existing GCL methods—whose neighborhood aggregation capability is hindered by suboptimal positive-pair construction (e.g., neglecting high-order neighbors or injecting noise)—we propose LightCCF, a lightweight framework that explicitly steers high-quality neighborhood information aggregation via semantic-consistent positive-pair reconstruction. LightCCF features low computational complexity and robustness to over-smoothing. Empirical evaluation on three highly sparse public benchmark datasets demonstrates significant improvements in both recommendation accuracy and training efficiency. The implementation is publicly available.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Search and Optimization: Learning to Search
• Computer Vision: Learning & Optimization for CV

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Personalized recommendation is widely used in the web applications, and graph contrastive learning (GCL) has gradually become a dominant approach in recommender systems, primarily due to its ability to extract self-supervised signals from raw interaction data, effectively alleviating the problem of data sparsity. A classic GCL-based method typically uses data augmentation during graph convolution to generates more contrastive views, and performs contrast on these new views to obtain rich self-supervised signals. Despite this paradigm is effective, the reasons behind the performance gains remain a mystery. In this paper, we first reveal via theoretical derivation that the gradient descent process of the CL objective is formally equivalent to graph convolution, which implies that CL objective inherently supports neighborhood aggregation on interaction graphs. We further substantiate this capability through experimental validation and identify common misconceptions in the selection of positive samples in previous methods, which limit the potential of CL objective. Based on this discovery, we propose the Light Contrastive Collaborative Filtering (LightCCF) method, which introduces a novel neighborhood aggregation objective to bring users closer to all interacted items while pushing them away from other positive pairs, thus achieving high-quality neighborhood aggregation with very low time complexity. On three highly sparse public datasets, the proposed method effectively aggregate neighborhood information while preventing graph over-smoothing, demonstrating significant improvements over existing GCL-based counterparts in both training efficiency and recommendation accuracy. Our implementations are publicly accessible.

Problem

• Reveals CL's inherent neighborhood aggregation in graph convolution
• Identifies misconceptions in positive sample selection limiting CL potential
• Proposes LightCCF for efficient neighborhood aggregation in sparse datasets

Innovation

• Graph contrastive learning for neighborhood aggregation
• LightCCF method with novel aggregation objective
• Prevents over-smoothing with low time complexity