mlsanityData to 7 Oct 2026

Graph Few-shot Class-incremental Learning

arXiv 2112.12819 ↗First posted 23 Dec 2021cs.LGcited by 40

Not on the map. Why?

Abstract (arXiv)

The ability to incrementally learn new classes is vital to all real-world artificial intelligence systems. A large portion of high-impact applications like social media, recommendation systems, E-commerce platforms, etc. can be represented by graph models. In this paper, we investigate the challenging yet practical problem, Graph Few-shot Class-incremental (Graph FCL) problem, where the graph model is tasked to classify both newly encountered classes and previously learned classes. Towards that purpose, we put forward a Graph Pseudo Incremental Learning paradigm by sampling tasks recurrently from the base classes, so as to produce an arbitrary number of training episodes for our model to practice the incremental learning skill. Furthermore, we design a Hierarchical-Attention-based Graph Meta-learning framework, HAG-Meta. We present a task-sensitive regularizer calculated from task-level attention and node class prototypes to mitigate overfitting onto either novel or base classes. To employ the topological knowledge, we add a node-level attention module to adjust the prototype representation. Our model not only achieves greater stability of old knowledge consolidation, but also acquires advantageous adaptability to new knowledge with very limited data samples. Extensive experiments on three real-world datasets, including Amazon-clothing, Reddit, and DBLP, show that our framework demonstrates remarkable advantages in comparison with the baseline and other related state-of-the-art methods.

Key facts. Cited by 40 papers in the map (data through 7 Oct 2026). Builds on 33 of its 57 references that are in the map. First posted on arXiv 23 Dec 2021.

Builds on

33 of 57 references.

References · 33

  1. 2017Attention Is All You Needcited by 59,932
  2. 2016Semi-Supervised Classification with Graph Convolutional Networkscited by 11,282
  3. 2015Distilling the Knowledge in a Neural Networkcited by 10,076
  4. 2017Graph Attention Networkscited by 8,854
  5. 2017Inductive Representation Learning on Large Graphscited by 6,284
  6. 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,861
  7. 2016Overcoming catastrophic forgetting in neural networkscited by 4,718
  8. 2018How Powerful are Graph Neural Networks?cited by 4,355
  9. 2017Prototypical Networks for Few-shot Learningcited by 3,001
  10. 2016Learning without Forgettingcited by 2,465
  11. 2016iCaRL: Incremental Classifier and Representation Learningcited by 2,313
  12. 2016Convolutional Neural Networks on Graphs with Fast Localized Spectral Filteringcited by 2,085
21 more
  1. 2017Gradient Episodic Memory for Continual Learningcited by 1,871
  2. 2016Variational Graph Auto-Encoderscited by 1,373
  3. 2018Experience Replay for Continual Learningcited by 1,074
  4. 2018Riemannian Walk for Incremental Learning: Understanding Forgetting and Intransigencecited by 734
  5. 2019Large Scale Incremental Learningcited by 727
  6. 2013An Empirical Investigation of Catastrophic Forgetting in Gradient-Based Neural Networkscited by 711
  7. 2020Long-tail learning via logit adjustmentcited by 605
  8. 2017Measuring Catastrophic Forgetting in Neural Networkscited by 359
  9. 2019Maintaining Discrimination and Fairness in Class Incremental Learningcited by 281
  10. 2020Few-Shot Class-Incremental Learningcited by 249
  11. 2015Deep Convolutional Networks on Graph-Structured Datacited by 246
  12. 2021Few-Shot Incremental Learning with Continually Evolved Classifierscited by 184
  13. 2019Meta-GNN: On Few-shot Node Classification in Graph Meta-learningcited by 105
  14. 2017Encoder Based Lifelong Learningcited by 102
  15. 2015Inferring Networks of Substitutable and Complementary Productscited by 91
  16. 2020Graph Prototypical Networks for Few-shot Learning on Attributed Networkscited by 90
  17. 2020Be More with Less: Hypergraph Attention Networks for Inductive Text Classificationcited by 73
  18. 2018Incremental Few-Shot Learning with Attention Attractor Networkscited by 65
  19. 2020XtarNet: Learning to Extract Task-Adaptive Representation for Incremental Few-Shot Learningcited by 30
  20. 2020Graph-Based Continual Learningcited by 24
  21. 2021Split-and-Bridge: Adaptable Class Incremental Learning within a Single Neural Networkcited by 14

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 8; 2023: 9; 2024: 15; 2025: 6; 2026: 2.

40 papers.

Sort:

  1. 2023A Survey on Few-Shot Class-Incremental Learningcited by 48
  2. 2023Virtual Node Tuning for Few-shot Node Classificationcited by 46
  3. 2022Graph Lifelong Learning: A Surveycited by 30
  4. 2022Graph Few-shot Learning with Task-specific Structurescited by 26
  5. 2022Transductive Linear Probing: A Novel Framework for Few-Shot Node Classificationcited by 24
  6. 2023S3C: Self-Supervised Stochastic Classifiers for Few-Shot Class-Incremental Learningcited by 23
  7. 2022FAITH: Few-Shot Graph Classification with Hierarchical Task Graphscited by 22
  8. 2024Continual Learning on Graphs: Challenges, Solutions, and Opportunitiescited by 20
  9. 2024A Survey of Data-Efficient Graph Learningcited by 18
  10. 2024Towards Robust Graph Incremental Learning on Evolving Graphscited by 18
  11. 2023Continual Graph Learning: A Surveycited by 17
  12. 2024Graph Learning under Distribution Shifts: A Comprehensive Survey on Domain Adaptation, Out-of-distribution, and Continual Learningcited by 17
28 more
  1. 2023A Survey of Imbalanced Learning on Graphs: Problems, Techniques, and Future Directionscited by 15
  2. 2023Few-shot Node Classification with Extremely Weak Supervisioncited by 14
  3. 2024Topology-aware Embedding Memory for Continual Learning on Expanding Networkscited by 14
  4. 2022Few-shot Class-incremental Learning for 3D Point Cloud Objectscited by 13
  5. 2024Rethinking Few-shot Class-incremental Learning: Learning from Yourselfcited by 13
  6. 2022Few-Shot Learning on Graphscited by 11
  7. 2023Contrastive Meta-Learning for Few-shot Node Classificationcited by 11
  8. 2022Supervised Graph Contrastive Learning for Few-shot Node Classificationcited by 9
  9. 2024Continual Learning for Smart City: A Surveycited by 9
  10. 2024An Efficient Memory Module for Graph Few-Shot Class-Incremental Learningcited by 8
  11. 2023On the Limitation and Experience Replay for GNNs in Continual Learningcited by 7
  12. 2024Towards Continuous Reuse of Graph Models via Holistic Memory Diversificationcited by 7
  13. 2023A Survey On Few-shot Knowledge Graph Completion with Structural and Commonsense Knowledgecited by 6
  14. 2025Few-Shot Graph Out-of-Distribution Detection with LLMscited by 6
  15. 2024Catching Chameleons: Detecting Evolving Disinformation Generated using Large Language Modelscited by 5
  16. 2022Few-Shot Semantic Relation Prediction across Heterogeneous Graphscited by 4
  17. 2024NC-NCD: Novel Class Discovery for Node Classificationcited by 4
  18. 2024Graph Memory Learning: Imitating Lifelong Remembering and Forgetting of Brain Networkscited by 4
  19. 2024MLDGG: Meta-Learning for Domain Generalization on Graphscited by 4
  20. 2024Foundation Model-Powered 3D Few-Shot Class Incremental Learning via Training-free Adaptorcited by 3
  21. 2025Latest Advancements Towards Catastrophic Forgetting under Data Scarcity: A Comprehensive Survey on Few-Shot Class Incremental Learningcited by 3
  22. 2024Inductive Graph Few-shot Class Incremental Learningcited by 2
  23. 2025Enhancing Distribution and Label Consistency for Graph Out-of-Distribution Generalizationcited by 2
  24. 2025Generative Risk Minimization for Out-of-Distribution Generalization on Graphscited by 2
  25. 2025Constraint Multi-class Positive and Unlabeled Learning for Distantly Supervised Named Entity Recognitioncited by 0
  26. 2025MEGA: Second-Order Gradient Alignment for Catastrophic Forgetting Mitigation in GFSCILcited by 0
  27. 2026BAED: a New Paradigm for Few-shot Graph Learning with Explanation in the Loopcited by 0
  28. 2026ReDIL-GNN: Resynthesis Domain Incremental Learning for Circuit Graph Neural Networkscited by 0