Rotate3D: Representing Relations as Rotations in Three-Dimensional Space for Knowledge Graph Embedding

This is the code of Rotate3D.

Dependencies

• Python 3.7+
• PyTorch 1.0+

The implementation of Rotate3D

Explanation of the implementation

Link Prediction

To reproduce the results of Rotate3D, run the following commands.

cd LP

# WN18
bash runs.sh train Rotate3D wn18 0 0 512 256 1000 12.0 1.0 0.0001 80000 8 0 2 --disable_adv 

# FB15k
bash runs.sh train Rotate3D FB15k 0 0 1024 256 1000 24.0 0.5 0.00005 150000 8 0 2

# WN18RR
bash runs.sh train Rotate3D wn18rr 0 0 512 256 500 6.0 1.0 0.00005 80000 8 0.1 1 --disable_adv

# FB15k-237
bash runs.sh train Rotate3D FB15k-237 0 0 1024 256 1000 12.0 1.0 0.00005 100000 8 0 2

Citation

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

@inproceedings{
  gao2020rotate3d,
  title={Rotate3D: Representing Relations as Rotations in Three-Dimensional Space for Knowledge Graph Embedding},
  author={Gao, Chang and Sun, Chengjie and Shan, Lili and Lin, Lei and Wang, Mingjiang},
  booktitle={Proceedings of the 29th ACM International Conference on Information \& Knowledge Management},
  pages={385--394},
  year={2020}
}

Acknowledgement

We refer to the code of RotatE, HAKE. Thanks for their contributions.