Are Graph Augmentations Necessary?: Simple Graph Contrastive Learning for Recommendation

By Junliang Yu, Hongzhi Yin, Xin Xia, Tong Chen, Li-zhen Cui, Quoc Viet Hung Nguyen. Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, 2021

990 citations recorded by the source corpus.

Contrastive learning (CL) recently has spurred a fruitful line of research in the field of recommendation, since its ability to extract self-supervised signals from the raw data is well-aligned with recommender systems' needs for tackling the data sparsity issue. A typical pipeline of CL-based recommendation models is first augmenting the user-item bipartite graph with structure perturbations, and then maximizing the node representation consistency between different graph augmentations. Although this paradigm turns out to be effective, what underlies the performance gains is still a mystery. In this paper, we first experimentally disclose that, in CL-based recommendation models, CL operates by learning more uniform user/item representations that can implicitly mitigate the popularity bias. Meanwhile, we reveal that the graph augmentations, which used to be considered necessary, just play a trivial role. Based on this finding, we propose a simple CL method which discards the graph augmentations and instead adds uniform noises to the embedding space for creating contrastive views. A comprehensive experimental study on three benchmark datasets demonstrates that, though it appears strikingly simple, the proposed method can smoothly adjust the uniformity of learned representations and has distinct advantages over its graph augmentation-based counterparts in terms of recommendation accuracy and training efficiency. The code is released at https://github.com/Coder-Yu/QRec.

Subfields: Artificial Intelligence, Computer Networks and Communications, Information Systems

View Paper: https://www.semanticscholar.org/paper/06c9cbf943185d1d1554e760279412598c6d81a6

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Hongzhi Yin
2. Tong Chen
3. Xin Xia
4. Junliang Yu
5. Zi-Liang Huang
6. Li-zhen Cui
7. Chao Huang
8. Lianghao Xia

In Computer Networks and Communications

1. Jundong Li
2. Hongzhi Yin
3. Dawei Yin

In Information Systems

1. Hongzhi Yin
2. Tong Chen
3. Xin Xia
4. Junliang Yu
5. Zi-Liang Huang
6. Li-zhen Cui
7. Chao Huang
8. Lianghao Xia

Citations in the corpus

• Representation Learning with Large Language Models for Recommendation (2023, methodology): “The evaluation results reveal several interesting observations, as outlined below: • Overall, we consistently observe that integrating RLMRec with the backbone recommenders leads to improved performance compared to the original versions.”
• XSimGCL: Towards Extremely Simple Graph Contrastive Learning for Recommendation (2022, methodology): “Though it seems a convention to separate the pipelines of the recommendation task and the contrastive task in CL-based recommender systems [12], [24], [25], [27], we question the necessity of this architecture.”
• Self-Supervised Learning for Recommender Systems: A Survey (2022, methodology): “FeaturePerturbation [82],[83]addsrandomnoisestooriginal user/item representations.”
• LightGCL: Simple Yet Effective Graph Contrastive Learning for Recommendation (2023, methodology): “To improve the graph contrastive learning in recommendation, SimGCL (Yu et al., 2022a) offers embedding augmentation with random noise perturbation.”
• LLMRec: Large Language Models with Graph Augmentation for Recommendation (2023, background): “Some studies [27] used contrastive learning to add self-supervised signals ( e . g ., SGL [42], SimGCL [45]).”
• Hard Negative Sampling via Large Language Models for Recommendation (2025, unknown): “Following the approach in (Yu et al., 2022), we categorize the test set into three subsets based on item popularity: Unpopular (the 80% of items with the fewest clicks), Popular (the top 5% of items), and Normal (the remaining items).”