Adaptive Graph Encoder for Attributed Graph Embedding

By Ganqu Cui, Jie Zhou, Cheng Yang, Zhiyuan Liu. Knowledge Discovery and Data Mining, 2020

297 citations recorded by the source corpus.

Attributed graph embedding, which learns vector representations from graph topology and node features, is a challenging task for graph analysis. Recently, methods based on graph convolutional networks (GCNs) have made great progress on this task. However,existing GCN-based methods have three major drawbacks. Firstly,our experiments indicate that the entanglement of graph convolutional filters and weight matrices will harm both the performance and robustness. Secondly, we show that graph convolutional filters in these methods reveal to be special cases of generalized Laplacian smoothing filters, but they do not preserve optimal low-pass characteristics. Finally, the training objectives of existing algorithms are usually recovering the adjacency matrix or feature matrix, which are not always consistent with real-world applications. To address these issues, we propose Adaptive Graph Encoder (AGE), a novel attributed graph embedding framework. AGE consists of two modules: (1) To better alleviate the high-frequency noises in the node features, AGE first applies a carefully-designed Laplacian smoothing filter. (2) AGE employs an adaptive encoder that iteratively strengthens the filtered features for better node embeddings. We conduct experiments using four public benchmark datasets to validate AGE on node clustering and link prediction tasks. Experimental results show that AGE consistently outperforms state-of-the-artgraph embedding methods considerably on these tasks.

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

View Paper: https://www.semanticscholar.org/paper/7d4dcee4628745f79f5e352859172817fece3e83

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. Zhiyuan Liu
2. Maosong Sun
3. Philip S. Yu
4. Jie Tang
5. Chuan Shi
6. Xiao Liu
7. Yuxiao Dong
8. Jie Zhou

In Information Systems

1. Xiao Wang
2. Maosong Sun
3. Chuan Shi
4. Philip S. Yu
5. Zhiyuan Liu
6. Jia Wu
7. Yuxiao Dong
8. Jie Tang

In Statistical and Nonlinear Physics

1. Chuan Shi
2. Xiao Wang
3. Jie Tang
4. Yuxiao Dong
5. Philip S. Yu
6. Hongxia Yang
7. Cécile Paris
8. Deyu Bo

Citations in the corpus

• Beyond Low-frequency Information in Graph Convolutional Networks (2021, background): “Some recent studies (Wu et al. 2019; Cui et al. 2020) emphasize that the entanglement of filters and weight matrices may be harmful to the performance and robustness of the model.”
• Graph Neural Networks: A Review of Methods and Applications (2018, background): “[1] proposed the GraphSAGE, a general inductive framework.”
• GraphMAE: Self-Supervised Masked Graph Autoencoders (2022, background): “For GAEs [3, 20, 26, 27, 31] that leverage feature reconstruction, most of them still employ the vanilla architecture that risks learning trivial solutions.”
• A Comprehensive Survey on Community Detection With Deep Learning (2021, methodology): “Adaptive graph encoder (AGE) [91] is another smoothing filter model scalable to community detection.”