CylE: Cylinder Embeddings for Multi-hop Reasoning over Knowledge Graphs

This is the codebase for the paper CylE: Cylinder Embeddings for Multi-hop Reasoning over Knowledge Graphs (EACL 2023).

Getting started

Step 1: Data Preparation

• Download the datasets here, then move KG_data.zip to ./cyle/ directory

• Unzip KG_data.zip to ./cyle/data/:

  cd cyle/
  unzip -d data KG_data.zip
  

Step 2: Installing Requirements

• If you are familiar with pip or conda, please install requirements by your own preference:

  python=3.8
  pytorch=1.9.1
  tqdm
  tensorboardX
  

• [Optional] For those prefer to use Anaconda, you can follow these:

  conda create --name cyle python=3.8.11
  conda activate cyle
  conda install pytorch==1.9.1 cudatoolkit=10.2 -c pytorch
  conda install tqdm
  pip install tensorboardX
  

Step 3: Training Model

• Run script ./run.sh

• Type [1-3] to select which dataset to run model, type [4/q/quit/exit] to exit

Citation

If you find this code useful for your research, please consider citing the following paper:

@inproceedings{nguyen2023cyle,
    title = "{C}yl{E}: Cylinder Embeddings for Multi-hop Reasoning over Knowledge Graphs",
    author = "Nguyen, Chau Duc Minh  and
      French, Tim  and
      Liu, Wei  and
      Stewart, Michael",
    booktitle = "Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics",
    month = may,
    year = "2023",
    address = "Dubrovnik, Croatia",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2023.eacl-main.127",
    pages = "1736--1751",
}

Acknowledgement

We acknowledge the code of KGReasoning for their contributions.