FreeHGC: Training-free Heterogeneous Graph Condensation via Data Selection

Requirements

Our environment configuration

• python 3.8
• torch 1.12.1+cu113
• torch_geometric 2.1.0
• torch_sparse 0.6.15
• torch_scatter 2.0.9
• numpy 1.21.5

Data preparation

For experiments in Motivation section and on four medium-scale datasets, please download datasets DBLP.zip, ACM.zip, IMDB.zip, Freebase.zip from the source of HGB benchmark, and extract content from these compresesed files under the folder './data/'.

For experiments on the large dataset AMiner, The dataset will be downloaded automatically. If the download fails, you can view the source code of torch_geometric.datasets and update the url.

Run

For medium-scale datasets:

python train_hgb.py --dataset ACM --method FreeHGC --reduction-rate 0.1 --pr 0.95 --gpu 0 --num-hops 3 --num-hidden 128 --lr 0.001 --dropout 0.5 --ff-layer-2 2 --ACM-keep-F

For large-scale dataset:

python train_ogbn_pr.py --dataset aminer --method FreeHGC --reduction-rate 0.05 --num-hops 2