XSimGCL: Towards Extremely Simple Graph Contrastive Learning for Recommendation

2023/01/01 by Junliang Yu, Xin Xia, Tong Chen +4 · 278 citations
Computer Science · Psychology · #Advanced Graph Neural Networks #Artificial intelligence #Bipartite graph #Caching and Content Delivery #Computer science #Consistency (knowledge bases) #Graph #Information retrieval #Natural language processing #Popularity #Psychology #Recommender Systems and Techniques #Theoretical computer science

paper · doi:10.1109/tkde.2023.3288135

published in IEEE Transactions on Knowledge and Data Engineering, 1-14 (IEEE Computer Society)

openalex publication_date 2023/01/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/01

Abstract

Contrastive learning (CL) has recently been demonstrated critical in improving recommendation performance. The underlying principle of CL-based recommendation models is to ensure the consistency between representations derived from different graph augmentations of the user-item bipartite graph. This self-supervised approach allows for the extraction of general features from raw data, thereby mitigating the issue of data sparsity. Despite the effectiveness of this paradigm, the factors contributing to its performance gains have yet to be fully understood. This paper provides novel insights into the impact of CL on recommendation. Our findings indicate that CL enables the model to learn more evenly distributed user and item representations, which alleviates the prevalent popularity bias and promoting long-tail items. Our analysis also suggests that the graph augmentations, previously considered essential, are relatively unreliable and of limited significance in CL-based recommendation. Based on these findings, we put forward an eXtremelySimpleGraphContrastiveLearning method (XSimGCL) for recommendation, which discards the ineffective graph augmentations and instead employs a simple yet effective noise-based embedding augmentation to generate views for CL. A comprehensive experimental study on four large and highly sparse benchmark datasets demonstrates that, though the proposed method is extremely simple, it can smoothly adjust the uniformity of learned representations and outperforms its graph augmentation-based counterparts by a large margin in both recommendation accuracy and training efficiency. The code and used datasets are released athttps://github.com/Coder-Yu/SELFRec.

Citations

• Momentum Contrast for Unsupervised Visual Representation Learning
• Self-supervised Learning: Generative or Contrastive
• LightGCN
• Bias and Debias in Recommender System: A Survey and Future Directions
• Graph Contrastive Learning with Augmentations
• SimCSE: Simple Contrastive Learning of Sentence Embeddings
• Representation Learning with Contrastive Predictive Coding

Cited by

• Prompt-Based LLMs for Position Bias-Aware Reranking in Personalized Recommendations
• Unifying Inductive, Cross-Domain, and Multimodal Learning for Robust and Generalizable Recommendation
• Post-hoc Popularity Bias Correction in GNN-based Collaborative Filtering
• Leveraging Language Semantics for Collaborative Filtering with TextGCN and TextGCN-MLP: Zero-Shot vs In-Domain Performance
• Enhanced Pre-training of Graph Neural Networks for Million-Scale Heterogeneous Graphs
• Generative Data Augmentation in Graph Contrastive Learning for Recommendation
• Conv4Rec: A 1-by-1 Convolutional AutoEncoder for User Profiling through Joint Analysis of Implicit and Explicit Feedbacks
• RecMind: LLM-Enhanced Graph Neural Networks for Personalized Consumer Recommendations
• Hybrid Matrix Factorization Based Graph Contrastive Learning for Recommendation System
• Breaking the Top-K Barrier: Advancing Top-K Ranking Metrics Optimization in Recommender Systems
• VoteGCL: Enhancing Graph-based Recommendations with Majority-Voting LLM-Rerank Augmentation
• NLGCL: Naturally Existing Neighbor Layers Graph Contrastive Learning for Recommendation
• RecRankerEval: A Flexible and Extensible Framework for Top-k LLM-based Recommendation
• Advancing Loss Functions in Recommender Systems: A Comparative Study with a Rényi Divergence-Based Solution
• AgentRecBench: Benchmarking LLM Agent-based Personalized Recommender Systems
• HGCL: Hierarchical Graph Contrastive Learning for User-Item Recommendation
• Lightweight Embeddings with Graph Rewiring for Collaborative Filtering
• Hyperbolic Contrastive Learning with Model-augmentation for Knowledge-aware Recommendation
• MMHCL: Multi-Modal Hypergraph Contrastive Learning for Recommendation
• Squeeze and Excitation: A Weighted Graph Contrastive Learning for Collaborative Filtering
• Why is Normalization Necessary for Linear Recommenders?