You can not select more than 25 topics Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.
Hongyi Zhang 925b86798b update format_dgl edge_size and node_size 3 weeks ago
dgl_example update format_dgl edge_size and node_size 3 weeks ago
openhgnn_example ICDM RGCN examples 4 weeks ago
pyg_example PyG update 3 weeks ago
.gitignore ICDM RGCN examples 4 weeks ago
README.md docs 4 weeks ago

README.md

icdm 2022 baseline

We implement RCCN baseline on ICDM 2022 : Risk Commodities Detection on Large-Scale E-commerce Graphs. Our implementation include three popular GNN platforms: DGL, PyG and OpenHGNN. You can use it as a reference for data processing, model training and inference.

Environment Setup

  • Option1: Run on OpenI Cloud platform.
    • We provide an image with pytorch, dgl and pyg installed. You can directly create an instance from that. Also, you can use dataset provided on OpenI.
  • Option2: Run locally.
    • Download the dataset and config the environment locally.

Experiment Result

AP session1 session2
pyg 0.9174 0.8931
dgl 0.8751 0.8417
openhgnn 0.8658 0.8387

简介

No Description

Python Shell

贡献者 (1)