Datasets
========

We provide some common Knowledge Graphs and their train, valid, test splits under this link: `https://web.informatik.uni-mannheim.de/pyclause/ <https://web.informatik.uni-mannheim.de/pyclause/>`_ .

The additional files used by our examples section, such as rules or external entity/relation names are also provided with the respective dataset.
A KG file contains **tab** separated string triples **subject  predicate   object** in every line.

The datasets are taken from `libKGE <https://github.com/uma-pi1/kge>`_ which is a library for training embedding models.