This is the implementation of the model proposed in Knowledge Hypergraphs: Extending Knowledge Graphs Beyond Binary Relations for knowledge hypergraph embedding and also all the baselines for this task. It can be also used to learn HypE models for any input model. The software can be also used as a framework to implement new knowledge hypergraph embedding models.
Pythonversion 3.6.6Numpyversion 1.16.2PyTorchversion 0.4.0
To run HypE or any of the baselines you should define the following parameters:
model: name of the model
dataset: The dataset you want to run this model on
lr: learning rate
nr: number of negative examples per positive example per arity
out_channels: number of out channels for convolution filters in HypE
filt_w: width of convolutional weight filters in HypE
stride: stride of convolutional weight filters in HypE
emb_dim: embedding dimension
input_drop: drop out rate for input layer of all models
hidden_drop: drop out rate for hidden layer of all models
- Run
python main.py -model model -dataset dataset -lr lr -nr nr -out_channels out_channels -filt_w filt_w -stride stride -emb_dim emb_dim -hidden_drop hidden_drop -input_drop input_drop
The baselines implemented in this package are m-DistMult, m-CP, m-SimplE, Shift1Left, and m-TransH.
If you use this package for published work, please cite the following: Knowledge Hypergraphs: Extending Knowledge Graphs Beyond Binary Relations
Bahare Fatemi
Computer Science Department
The University of British Columbia
201-2366 Main Mall, Vancouver, BC, Canada (V6T 1Z4)
Licensed under the GNU General Public License Version 3.0. https://www.gnu.org/licenses/gpl-3.0.en.html