hybrid-graph-benchmark

This is a benchmark dataset for evaluating hybrid graph (a unified definition for higher-order graphs, including hypergraphs and hierarchical graphs) learning algorithms. It contains:

Get started by installing from GitHub

Modules

Hybrid Graph Datasets

Hybrid Graph Evaluation Framework

Summary

Click the dataset name to see more details:

Name #Graphs #Nodes #Edges #Hyperedges Avg. Node Degree Avg. Hyperedge Degree Clustering Coef. Task Type
MUSAE-GitHub 1 37,700 578,006 223,672 30.7 4.6 0.168 Node Classification
MUSAE-Facebook 1 22,470 342,004 236,663 30.4 9.9 0.360 Node Classification
MUSAE-Twitch 6 5,686 143,038 110,142 50.6 6.0 0.210 Node Classification
MUSAE-Wiki 3 6,370 266,998 118,920 88.8 14.4 0.413 Node Regression
GRAND-Tissues 6 5,931 5,926 11,472 2.0 1.3 0.000 Node Classification
GRAND-Diseases 4 4,596 6,252 7,743 2.7 1.3 0.000 Node Classification
Amazon-Computers 1 10,226 55,324 10,226 10.8 4.0 0.249 Node Classification
Amazon-Photos 1 6,777 45,306 6,777 13.4 4.8 0.290 Node Classification

License

Source code: MIT license
MUSAE & GRAND datasets: GPLv3 license
Amazon datasets: Amazon Service license

Cite This Project

@article{Li2023HybridGraph,
    title={Hybrid Graph: A Unified Graph Representation with Datasets and Benchmarks for Complex Graphs},
    author={Zehui Li and 
            Xiangyu Zhao and 
            Mingzhu Shen and
            Guy-Bart Stan and
            Pietro Li{\`o} and
            Yiren Zhao},
    journal={arXiv preprint arXiv:2306.05108},
    year={2023}
}