RSHN

RSHN[ICDM2019]

• paper: Relation Structure-Aware Heterogeneous Graph Neural Network
• Code from author: RSHN

How to run

• Clone the Openhgnn-DGL

  python main.py -m RSHN -t node_classification -d aifb -g 0

  If you do not have gpu, set -gpu -1.

  Supported Dataset: RDFDataset

Performance: Node classification

Method	AIFB	MUTAG	BGS	AM
RSHN	97.22	82.35	93.10	90.40
RSHN(openhgnn)	97.22	85.29	93.10	89.39

The experiments in RSHN have serious problems according to the official code. First, it does not use validation set, and just tune hyperparameters on test set. Second, it reports the accuracy at the epoch with best accuracy on test set in the paper. And in this setting, we give our performance.

TrainerFlow: [node classification flow](../../trainerflow/#Node classification flow)

Model

• 1) Coarsened Line Graph Neural Network (CL-GNN):
  • We implement the API coarsened_line_graph
• 2) Heterogeneous Graph Neural Network (H-GNN):

Hyper-parameter specific to the model

You can modify the parameters in openhgnn/config.ini

Description

# The next two hyper-parameters are used in building the coarsened-line graph.
rw_len = 5
batch_size = 1000
#	edga_layer means number of CL-GNN layers, node_layer means number of H-GNN layers
num_node_layer = 2
num_edge_layer = 1

Best config can be found in best_config

More

Contirbutor

Tianyu Zhao[GAMMA LAB]

If you have any questions,

Submit an issue or email to tyzhao@bupt.edu.cn.