SAMGC

This is the code of paper: Shared-Attribute Multi-graph Clustering with Global Self-Attention.

Cite this paper

@InProceedings{10.1007/978-3-031-30105-6_5,
author="Chen, Jianpeng and Yang, Zhimeng and Pu, Jingyu and Ren, Yazhou and Pu, Xiaorong and Gao, Li and He, Lifang"
title="Shared-Attribute Multi-Graph Clustering with Global Self-Attention",
booktitle="Neural Information Processing",
year="2023",
publisher="Springer International Publishing",
address="Cham",
pages="51--63",
isbn="978-3-031-30105-6"
}

Chen, J. et al. (2023). Shared-Attribute Multi-Graph Clustering with Global Self-Attention. In: Tanveer, M., Agarwal, S., Ozawa, S., Ekbal, A., Jatowt, A. (eds) Neural Information Processing. ICONIP 2022. Lecture Notes in Computer Science, vol 13623. Springer, Cham. https://doi.org/10.1007/978-3-031-30105-6_5

Requirements

• Python 3.7.11

• PyTorch 1.9.1

• munkres 1.1.4

• scikit-learn 0.24.2

• scipy 1.6.2

Datasets

• ACM and Cora are included in .\data\cora and .\data\ respectively.
• Large datasets will uploaded after review.

Test SAMGC

• Test SAMGC on ACM: sh test_acm.sh
• Test SAMGC on Cora: sh test_cora.sh

Train SAMGC

• Train SAMGC on ACM: sh train_acm.sh

• Train SAMGC on Cora: sh train_cora.sh

Results of SAMGC

	NMI	ARI	ACC	F1
ACM	77.2	82.8	94.0	94.0
Cora	58.2	51.1	73.5	72.7