Scale-Free, Attributed and Class-Assortative Graph Generation

This repository contains code for the paper "Scale-Free, Attributed and Class-Assortative Graph Generation to Facilitate Introspection of Graph Neural Networks," published at KDD MLG 2020. For clarity:

• CABAM Simulation Examples.ipynb: Several examples of generated graphs with various desiderata, their assortativity properties, illustrations of theoretical and empirical quantities as discussed in the paper, and example class-conditional feature generation given a graph and assigned node labels.
• Dataset Summaries.ipynb: Driver to generate the dataset statistics for existing GNN benchmarks provided in the paper.
• cabam_utils.py: Main graph generation code and helpers.
• graph_preprocessing_utils.py: Misc. helpers to load and process existing graph datasets from data/.
• graph_summary_utils.py: Misc. helpers for driver to generate benchmark dataset statistics and properties.

If you use the model, or graphs generated with the model for evaluation in your own work, please cite

 @inproceedings{cabam2020shah,
     author = {Shah, Neil},
     title = {Scale-Free, Attributed and Class-Assortative Graph Generation to Facilitate Introspection of Graph Neural Networks},
     booktitle = {KDD Mining and Learning with Graphs},
     year = {2020}
   }