facebookresearch / tkbc

A knowledge base completion method which handles temporal metadata

Knowledge Base Completion (kbc)

This code reproduces results in Tensor Decompositions for Temporal Knowledge Base Completion (ICLR 2020).

Installation

Create a conda environment with pytorch and scikit-learn :

conda create --name tkbc_env python=3.7
source activate tkbc_env
conda install --file requirements.txt -c pytorch

Then install the kbc package to this environment

python setup.py install

Datasets

To download the datasets, go to the tkbc/scripts folder and run:

chmod +x download_data.sh
./download_data.sh

Once the datasets are downloaded, add them to the package data folder by running :

python tkbc/process_icews.py
python tkbc/process_yago.py
python tkbc/process_wikidata.py  # about 3 minutes.

This will create the files required to compute the filtered metrics.

Reproducing results

In order to reproduce the results on the smaller datasets in the paper, run the following commands

python tkbc/learner.py --dataset ICEWS14 --model TNTComplEx --rank 156 --emb_reg 1e-2 --time_reg 1e-2

python tkbc/learner.py --dataset ICEWS05-15 --model TNTComplEx --rank 128 --emb_reg 1e-3 --time_reg 1

python tkbc/learner.py --dataset yago15k --model TNTComplEx --rank 189 --no_time_emb --emb_reg 1e-2 --time_reg 1

License

tkbc is CC-BY-NC licensed, as found in the LICENSE file.