Abstract:Graph embedding methods aim to represent nodes in the network into a low-dimensional and continuous vector space while preserving the topological structure and varieties of relational information maximally. Nowadays the structural connections of networks and the attribute information about each node are more easily available than before. As a result, many community detection algorithms for attributed networks have been proposed. However, the majority of these methods cannot deal with the overlapping community detection problem, which is one of the most significant issues in the real-world complex network study. In addition, it is quite challenging to make full use of both structural and attribute information instead of only focusing on one part. To this end, in this paper we innovatively combine the graph embedding with multiobjective evolutionary algorithms (MOEAs) for overlapping community detection problems in attributed networks. As far as I am concerned, MOEA is first used to integrate with graph embedding methods for overlapping community detection. We term our method as MOEA-GE OV , which can automatically determine the number of communities without any prior knowledge and consider topological structure and vertex properties synchronously. In MOEA-GE OV , two objective functions concerning community structure and attribute similarity are carefully designed. Moreover, a heuristic initialization method is proposed to get a relatively good initial population. Then a novel encoding and decoding strategy is designed to efficiently represent the overlapping communities and corresponding embedded representation. In the experiments, the performance of MOEA-GE OV is validated on both single and multiple attribute real-world networks. The experimental results of community detection tasks demonstrate our method can effectively obtain overlapping community structures with practical significance.

Semi-supervised Overlapping Community Detection in Attributed Graph with Graph Convolutional Autoencoder

Detecting communities with multiple topics in attributed networks via self-supervised adaptive graph convolutional network

Community Detection Via Joint Graph Convolutional Network Embedding In Attribute Network

Masked Dual Graph Autoencoder for Attributed Graph Community Detection

A Graph Convolutional Network-Based Approach for Community Detection in Attributed Networks

Unsupervised Learning for Community Detection in Attributed Networks Based on Graph Convolutional Network

Community-Centric Graph Convolutional Network for Unsupervised Community Detection

Deep Dual Graph Attention Auto-Encoder for Community Detection

Semi-supervised community detection method based on generative adversarial networks

Graph Convolutional Networks Meet Markov Random Fields: Semi-Supervised Community Detection in Attribute Networks

Community Detection Fusing Graph Attention Network

Multiple Topics Community Detection in Attributed Networks

Community Detection Based on Deep Dual Graph Autoencoder

Community Detection in Attributed Graphs: an Embedding Approach

Weakly-supervised learning for community detection based on graph convolution in attributed networks

GeoAI-Enhanced Community Detection on Spatial Networks with Graph Deep Learning

Overlapping Community Detection Based on Attribute Augmented Graph

QD-GCN: Query-Driven Graph Convolutional Networks for Attributed Community Search

Atrributed Graph Embedding Based on Multiobjective Evolutionary Algorithm for Overlapping Community Detection

CommunityGAN: Community Detection with Generative Adversarial Nets

VGAER: Graph Neural Network Reconstruction based Community Detection