Abstract:Graph clustering is one of the most fundamental tasks in graph learning. Recently, numerous graph clustering models based on dual network (Auto-encoder+Graph Neural Network(GNN)) architectures have emerged and achieved promising results. However, we observe several limitations in the literature: 1) simple graph neural networks that fail to capture the intricate relationships between nodes are used for graph clustering tasks; 2) heterogeneous information is inadequately interacted and merged; and 3) the clustering boundaries are fuzzy in the feature space. To address the aforementioned issues, we propose a novel graph clustering model named Dynamic Graph Attention-guided Graph Clustering with Entropy Minimization self-supervision(DGAGC-EM). Specifically, we introduce DGATE, a graph auto-encoder based on dynamic graph attention, to capture the intricate relationships among graph nodes. Additionally, we perform feature enhancement from both global and local perspectives via the proposed Global-Local Feature Enhancement (GLFE) module. Finally, we propose a self-supervised strategy based on entropy minimization theory to guide network training process to achieve better performance and produce sharper clustering boundaries. Extensive experimental results obtained on four datasets demonstrate that our method is highly competitive with the SOTA methods. The figure presents the overall framework of proposed Dynamic Graph Attention-guided Graph Clustering with Entropy Minimization selfsupervision(DGAGC-EM). Specifically, the Dynamic Graph Attetion Auto-Encoder Module is our proposed graph auto-encoder based on dynamic graph attention, to capture the intricate relationships among graph nodes. The Auto-Encoder Module is a basic autoencoder with simple MLPs to extract embeddings from node attributes. Additionally, the proposed Global-Local Feature Enhancement (GLFE) module perform feature enhancement from both global and local perspectives. Finally, the proposed Self-supervised Module guide network training process to achieve better performance and produce sharper clustering boundaries

A self-supervised learning model for graph clustering optimization problems

Homophily-Enhanced Self-Supervision for Graph Structure Learning: Insights and Directions.

Self-Supervised Contrastive Graph Clustering Network via Structural Information Fusion

Self-supervised Contrastive Attributed Graph Clustering

Structured graph learning for clustering and semi-supervised classification

Graph Representation Learning via Contrasting Cluster Assignments

Learning to solve graph metric dimension problem based on graph contrastive learning

Self-supervised Graph Convolutional Clustering by Preserving Latent Distribution

CARL-G: Clustering-Accelerated Representation Learning on Graphs

Supervised contrastive learning for graph representation enhancement

GSSCL: A framework for Graph Self-Supervised Curriculum Learning based on clustering label smoothing

CC-GNN: A Clustering Contrastive Learning Network for Graph Semi-Supervised Learning

GLCC: A General Framework for Graph-Level Clustering

Deep Contrastive Graph Learning with Clustering-Oriented Guidance

Dynamic Graph Attention-Guided Graph Clustering with Entropy Minimization Self-Supervision

Learning with Adaptive Neighbors for Image Clustering

Multi-Task Curriculum Graph Contrastive Learning with Clustering Entropy Guidance

Adaptive Graph Convolutional Clustering Network with Optimal Probabilistic Graph

Semi-Supervised Clustering via Dynamic Graph Structure Learning

Learning to Solve Combinatorial Optimization Problems on Real-World Graphs in Linear Time

Self-Consistent Contrastive Attributed Graph Clustering With Pseudo-Label Prompt