PGRA

The code for the paper "PGRA: Projected Graph Relation-Feature Attention Network for Heterogeneous Information Network Embedding", Information Sciences.

Overview

Requirements

We tested the code on:

• python 3.6
• pytorch 1.5.1
• networkx 2.3

other requirements:

• numpy
• pandas

Usage

Run the code using the command:

python src/main.py [Dataset_Name] [options]

We provide datasets in "data.zip" file.

Options

For PGRA-DistMult, use default hyperparameters

For PGRA-TransH, use

--score [l1/l2] --pre transh

For the best neighbor regularization settings (lambda) on DBLP/Yelp/DM/Aminer, use

--best_lambda

or to manually set, use

--nb_reg [value]

Reference

If you find the code helpful, please cite our work:

@article{chairatanakul2021pgra,
  title = {PGRA: Projected graph relation-feature attention network for heterogeneous information network embedding},
  journal = {Information Sciences},
  volume = {570},
  pages = {769-794},
  year = {2021},
  issn = {0020-0255},
  doi = {https://doi.org/10.1016/j.ins.2021.04.070},
  url = {https://www.sciencedirect.com/science/article/pii/S0020025521004059},
  author = {Nuttapong Chairatanakul and Xin Liu and Tsuyoshi Murata}
}