Overcoming Pitfalls in Graph Contrastive Learning Evaluation: Toward Comprehensive Benchmarks
Not on the map. Why?
Abstract (arXiv)
The rise of self-supervised learning, which operates without the need for labeled data, has garnered significant interest within the graph learning community. This enthusiasm has led to the development of numerous Graph Contrastive Learning (GCL) techniques, all aiming to create a versatile graph encoder that leverages the wealth of unlabeled data for various downstream tasks. However, the current evaluation standards for GCL approaches are flawed due to the need for extensive hyper-parameter tuning during pre-training and the reliance on a single downstream task for assessment. These flaws can skew the evaluation away from the intended goals, potentially leading to misleading conclusions. In our paper, we thoroughly examine these shortcomings and offer fresh perspectives on how GCL methods are affected by hyper-parameter choices and the choice of downstream tasks for their evaluation. Additionally, we introduce an enhanced evaluation framework designed to more accurately gauge the effectiveness, consistency, and overall capability of GCL methods.
Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 39 of its 43 references that are in the map. First posted on arXiv 24 Feb 2024.
Builds on
39 of 43 references.
References · 39
- 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 11,979
- 2016Semi-Supervised Classification with Graph Convolutional Networkscited by 11,282
- 2017Graph Attention Networkscited by 8,854
- 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,482
- 2018How Powerful are Graph Neural Networks?cited by 4,355
- 2017Neural Message Passing for Quantum Chemistrycited by 3,460
- 2021SimCSE: Simple Contrastive Learning of Sentence Embeddingscited by 2,168
- 2020Graph Contrastive Learning with Augmentationscited by 1,203
- 2018Graph Convolutional Neural Networks for Web-Scale Recommender Systemscited by 1,133
- 2018Deep Graph Infomaxcited by 1,101
- 2018Pitfalls of Graph Neural Network Evaluationcited by 994
- 2016Revisiting Semi-Supervised Learning with Graph Embeddingscited by 970
27 more
- 2019Graph WaveNet for Deep Spatial-Temporal Graph Modelingcited by 820
- 2015Gated Graph Sequence Neural Networkscited by 792
- 2015Convolutional Networks on Graphs for Learning Molecular Fingerprintscited by 751
- 2020Contrastive Multi-View Representation Learning on Graphscited by 651
- 2020Deep Graph Contrastive Representation Learningcited by 535
- 2020Graph Contrastive Learning with Adaptive Augmentationcited by 519
- 2018Graph Convolutional Policy Network for Goal-Directed Molecular Graph Generationcited by 363
- 2021Graph Self-Supervised Learning: A Surveycited by 289
- 2016Molecular Graph Convolutions: Moving Beyond Fingerprintscited by 238
- 2021Large-Scale Representation Learning on Graphs via Bootstrappingcited by 184
- 2022SimGRACE: A Simple Framework for Graph Contrastive Learning without Data Augmentationcited by 171
- 2023GraphPrompt: Unifying Pre-Training and Downstream Tasks for Graph Neural Networkscited by 166
- 2023All in One: Multi-task Prompting for Graph Neural Networkscited by 147
- 2017Predicting multicellular function through multi-layer tissue networkscited by 138
- 2021From Canonical Correlation Analysis to Self-supervised Graph Neural Networkscited by 138
- 2021An Empirical Study of Graph Contrastive Learningcited by 113
- 2019Learning Disentangled Representations for Recommendationcited by 110
- 2022Universal Prompt Tuning for Graph Neural Networkscited by 100
- 2017Geometric Matrix Completion with Recurrent Multi-Graph Neural Networkscited by 88
- 2021Graph Barlow Twins: A self-supervised representation learning framework for graphscited by 85
- 2019Graph Transfer Learning via Adversarial Domain Adaptation with Graph Convolutioncited by 67
- 2022Spectral Feature Augmentation for Graph Contrastive Learning and Beyondcited by 49
- 2022COSTA: Covariance-Preserving Feature Augmentation for Graph Contrastive Learningcited by 38
- 2019Deep Social Collaborative Filteringcited by 32
- 2023Rethinking Sensors Modeling: Hierarchical Information Enhanced Traffic Forecastingcited by 11
- 2023Multi-label Node Classification On Graph-Structured Datacited by 10
- 2021Social Network Analysis: From Graph Theory to Applications with Pythoncited by 4
Built on by
Citing papers per year: 2024: 2; 2025: 1.
3 papers.