Skip to content

yunchengwang/KGBoost-KGC

Repository files navigation

KGBoost: A Classification-Based Knowledge Base Completion Method with Negative Sampling

This is the Python implementation of KGBoost

Requirements

You will need numpy and xgboost to run KGBoost.

 pip install -r requirements.txt

Download Pre-trained Entity Embedding

To download the pre-trained embedding for entities required in KGBoost. Simply run:

 source download_pretrained.sh

Pre-trained TransE and RotatE are generated using this code.

Pre-trained word embeddings for FB15k-237 can be obtained here.

Run KGBoost

The best configurations for KGBoost-R and KGBoost-T in each dataset are specified in best_config.sh. Uncomment the models and the datasets you want to run.

 source best_config.sh

Citation

If you find the source codes useful, please consider citing our paper:

@article{wang2022kgboost,
  title={KGBoost: A Classification-Based Knowledge Base Completion Method with Negative Sampling},
  author={Wang, Yun-Cheng and Ge, Xiou and Wang, Bin and Kuo, C-C Jay},
  journal={Pattern Recognition Letters},
  year={2022},
  publisher={Elsevier}
}

About

A python implementation of KGBoost

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published