Graph meets probabilistic generation model: A new perspective for graph disentanglement

Zouzhang Peng, Shuai Zheng, Zhenfeng Zhu, Zhizhe Liu, Jian Cheng, Honghui Dong, Yao Zhao

Abstract: Highlights•Present a new probabilistic generative model for graph representation learning.•Propose an innovative approach for parameter estimation instead of general EM.•Propose a twin-branch structure to facilitate factor-level and feature-level node embedding.