Abstract:Community detection aims to partition network nodes into a set of clusters, such that nodes are more densely connected to each other within the same cluster than other clusters. For attributed networks, apart from the denseness requirement of topology structure, the attributes of nodes in the same community should also be homogeneous. Network embedding has been proved extremely useful in a variety of tasks, such as node classification, link prediction, and graph visualization, but few works dedicated to unsupervised embedding of node features specified for clustering task, which is vital for community detection and graph clustering. By post-processing with clustering algorithms like k -means, most existing network embedding methods can be applied to clustering tasks. However, the learned embeddings are not designed for clustering task, they only learn topological and attributed information of networks, and no clustering-oriented information is explored. In this article, we propose an algorithm named Network Embedding for node Clustering (NEC) to learn network embedding for node clustering in attributed graphs. Specifically, the presented work introduces a framework that simultaneously learns graph structure-based representations and clustering-oriented representations together. The framework consists of the following three modules: graph convolutional autoencoder module, soft modularity maximization module, and self-clustering module. Graph convolutional autoencoder module learns node embeddings based on topological structure and node attributes. We introduce soft modularity, which can be easily optimized using gradient descent algorithms, to exploit the community structure of networks. By integrating clustering loss and embedding loss, NEC can jointly optimize node cluster labels assignment and learn representations that keep local structure of network. This model can be effectively optimized using stochastic gradient algorithm. Empirical experiments on real-world networks and synthetic networks validate the feasibility and effectiveness of our algorithm on community detection task compared with network embedding based methods and traditional community detection methods.

Neighbor Aggregative Network Embedding

Gahne: Graph-Aggregated Heterogeneous Network Embedding

Adaptive Neighbor Graph Aggregated Graph Attention Network for Heterogeneous Graph Embedding

Deep Kernel Network Embedding

EPINE: Enhanced Proximity Information Network Embedding

Learning Network Embedding with Community Structural Information

Sparse Attributed Network Embedding Via Adaptively Aggregating Neighborhood Information

Improving Knowledge Graph Embedding Using Dynamic Aggregation of Neighbor Information.

Attributed Network Embedding with Community Preservation

Inductive Network Representation based on Neighbor Aggregation

Network Embedding Algorithm Taking in Variational Graph AutoEncoder

Enhancing Graph Collaborative Filtering Via Neighborhood Structure Embedding

Hierarchical Community Structure Preserving Network Embedding: A Subspace Approach

Community-aware Dynamic Network Embedding by Using Deep Autoencoder.

Toward Noise-Resistant Graph Embedding with Subspace Clustering Information.

GraphHINGE: Learning Interaction Models of Structured Neighborhood on Heterogeneous Information Network

Network Embedding for Community Detection in Attributed Networks

Neighborhood Graph Embedding for Nodes Clustering of Social Network

Deep Network Embedding with Aggregated Proximity Preserving

Neoknn: A Network Embedding Method Only Knowing Neighbor Nodes

Large-Scale Nodes Classification with Deep Aggregation Network