D2G2

This repository maintains the implementation of D2G2 (SDM 21), a generic framework of deep generative models for interpretable dynamic graph generation. Detailded information about D2G2 is provided in D2G2.

Requirements

• PyTorch 1.4 or higher
• Python 3.7

Instructions

1. Clone/download this repository.
2. Add datasets consisting of adjacency matrix and feature matrix with time dimension to dataset folder.
3. Run the code.
   • model.py: the D2G2 model.
   • trainer.py: train D2G2.

Reference

@inproceedings{zhang2021disentangled,
title={Disentangled Dynamic Graph Deep Generation},
author={Zhang, Wenbin and Zhang, Liming and Pfoser, Dieter and Zhao, Liang},
booktitle={Proceedings of the 2021 SIAM International Conference on Data Mining (SDM)},
pages={738--746},
year={2021},
organization={SIAM}
}