Graph Contrastive Learning for Optimizing Sparse Data in Recommender Systems with LightGCL

TL;DR

This paper introduces LightGCL, a graph contrastive learning model utilizing SVD for robust augmentation, significantly improving recommendation accuracy and fairness in sparse, noisy data environments.

Contribution

LightGCL is a novel graph contrastive learning approach that leverages SVD for effective graph augmentation without stochastic perturbations, enhancing recommendation performance.

Findings

Significant performance improvements over state-of-the-art models.

Enhanced fairness and reduced popularity bias.

Robustness to data sparsity and noise.

Abstract

Graph Neural Networks (GNNs) are powerful tools for recommendation systems, but they often struggle under data sparsity and noise. To address these issues, we implemented LightGCL, a graph contrastive learning model that uses Singular Value Decomposition (SVD) for robust graph augmentation, preserving semantic integrity without relying on stochastic or heuristic perturbations. LightGCL enables structural refinement and captures global collaborative signals, achieving significant gains over state-of-the-art models across benchmark datasets. Our experiments also demonstrate improved fairness and resilience to popularity bias, making it well-suited for real-world recommender systems.