Attributed Graph Clustering via Adaptive Graph Convolution

By Xiaotong Zhang, 刘晗, Qimai Li, Xiao-Ming Wu. International Joint Conference on Artificial Intelligence, 2019

349 citations recorded by the source corpus.

Attributed graph clustering is challenging as it requires joint modelling of graph structures and node attributes. Recent progress on graph convolutional networks has proved that graph convolution is effective in combining structural and content information, and several recent methods based on it have achieved promising clustering performance on some real attributed networks. However, there is limited understanding of how graph convolution affects clustering performance and how to properly use it to optimize performance for different graphs. Existing methods essentially use graph convolution of a fixed and low order that only takes into account neighbours within a few hops of each node, which underutilizes node relations and ignores the diversity of graphs. In this paper, we propose an adaptive graph convolution method for attributed graph clustering that exploits high-order graph convolution to capture global cluster structure and adaptively selects the appropriate order for different graphs. We establish the validity of our method by theoretical analysis and extensive experiments on benchmark datasets. Empirical results show that our method compares favourably with state-of-the-art methods.

Subfields: Artificial Intelligence, Information Systems, Statistical and Nonlinear Physics

View Paper: https://www.semanticscholar.org/paper/012e01b18192c2b825aa62daf93a9a2340db710a

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Philip S. Yu
2. Chuan Zhou
3. Jia Wu
4. Quan Z. Sheng
5. Shan Xue
6. Jian Yang
7. Emmanuel Müller
8. Cécile Paris

In Information Systems

1. Jia Wu
2. Philip S. Yu
3. Cécile Paris
4. Surya Nepal
5. Quan Z. Sheng
6. Chuan Zhou
7. Bryan Perozzi
8. Yongfeng Zhang

In Statistical and Nonlinear Physics

1. Philip S. Yu
2. Cécile Paris
3. Jia Wu
4. Surya Nepal
5. Quan Z. Sheng
6. Emmanuel Müller
7. John Palowitch
8. Chuan Zhou

Citations in the corpus

• Adaptive Graph Encoder for Attributed Graph Embedding (2020, neutral): “AGC [38] exploits high-order graph convolution to filter node features.”
• Graph Clustering with Graph Neural Networks (2020, background): “• AGC(graph, features) (Zhang et al., 2019) applied spectral clustering over features smoothed by the graph.”
• A Comprehensive Survey on Community Detection With Deep Learning (2021, background): “To this end, a low-pass filter is embedded in adaptive graph convolution (AGC) [69] with spectral clustering through where the frequency response function of G denoted as p ((cid:6)) = diag ( p (λ 1 ), . . . , p (λ n )) is decreasing and nonnega-tive on all eigenvalues.”