Advances in Graph Neural Networks for Node Classification and Recommendation Systems

The field of graph neural networks (GNNs) is rapidly advancing, with a focus on improving node classification and recommendation systems. Recent developments have explored the use of ensemble learning, contrastive learning, and graph attention networks to enhance the performance and robustness of GNNs. Additionally, there is a growing interest in semi-supervised and unsupervised learning methods, which can effectively leverage large amounts of unlabeled data. Noteworthy papers in this area include:

AugWard, which proposes a novel graph representation learning framework that carefully considers the diversity introduced by graph augmentation, and
CombiGCN, which introduces a user-user weighted connection graph to improve the accuracy of GNN models for recommendation systems.

Sources

Adapt, Agree, Aggregate: Semi-Supervised Ensemble Labeling for Graph Convolutional Networks

Does GCL Need a Large Number of Negative Samples? Enhancing Graph Contrastive Learning with Effective and Efficient Negative Sampling

Shapley-Guided Utility Learning for Effective Graph Inference Data Valuation

Social Network User Profiling for Anomaly Detection Based on Graph Neural Networks

Towards Efficient Training of Graph Neural Networks: A Multiscale Approach

Peer Disambiguation in Self-Reported Surveys using Graph Attention Networks

BeLightRec: A lightweight recommender system enhanced with BERT

Semi-supervised Node Importance Estimation with Informative Distribution Modeling for Uncertainty Regularization

SE-GNN: Seed Expanded-Aware Graph Neural Network with Iterative Optimization for Semi-supervised Entity Alignment

AugWard: Augmentation-Aware Representation Learning for Accurate Graph Classification

Improvement Graph Convolution Collaborative Filtering with Weighted addition input

CombiGCN: An effective GCN model for Recommender System

Fusion of Graph Neural Networks via Optimal Transport