Multi-view Contrastive Learning for Entity Typing over Knowledge Graph

This repo provides the source code & data of our paper: Multi-view Contrastive Learning for Entity Typing over Knowledge Graph.

Dependencies

• conda create -n mclet python=3.7 -y
• PyTorch 1.8.1
• dgl 0.6.1
• scipy 1.7.3

Running the code

Dataset

• Download the datasets from Here.
• Create the root directory ./data and put the dataset in.

Training model

sh FB15kET.sh
sh YAGO43kET.sh

• Note: Before running, you need to create the ./logs folder first.

Citation

If you find this code useful, please consider citing the following paper.

@article{
  author={Zhiwei Hu and Víctor Gutiérrez-Basulto and Zhiliang Xiang and and Ru Li and Jeff Z. Pan},
  title={Multi-view Contrastive Learning for Entity Typing over Knowledge Graph},
  publisher="The Conference on Empirical Methods in Natural Language Processing",
  year={2023}
}

Acknowledgement

We refer to the code of CET, TET, and MiNer. Thanks for their contributions.