This repository contains the supporting code and data files for our experiment.
The trained KGE Models using the datasets in this repository can be found at: https://doi.org/10.5281/zenodo.10296229
The scripts found in this directory were developed using Python 3.7.9, 3.9.x, and 3.10.x. Researchers may need to adjust the scripts to appropriately point to the correct directories.
test_parse.py is used to generate head, relation, and tail files, which can be used with dglke_predict.
The data files found in this sub-directory contains the exclusive extension we implemented into fb15k-237 to generate fb15k-238 and -239.
The {train, test, valid}.txt files are the original files used in the benchmarking dataset FB15k-237
These files can be generated with the script [code/fb15k237-classifying.py]. The existing files are the files used in the experiment.