Source code for pykeen.datasets.aristo

"""The Aristo-v4 dataset."""

import pathlib

import click
from docdata import parse_docdata
from more_click import verbose_option

from .base import PackedZipRemoteDataset


[docs] @parse_docdata class AristoV4(PackedZipRemoteDataset): """The Aristo-v4 dataset from [chen2021]. .. note:: The dataset is based on the Aristo tuple KG from https://aclanthology.org/Q17-1017/. .. warning:: While the original dataset is described as having 44,950 entities, after removing the entities not present in train, only 42,016 remain. Similarly, only 1,593 relations occur in training (from the original 1,605 ones). Consequently, some validation and testing triples are removed (originally: 20,000). Finally, only 242,567 of 242,594 original training triples are unique. --- name: Aristo-v4 statistics: entities: 42016 relations: 1593 training: 242567 testing: 18414 validation: 18444 triples: 279425 citation: author: Chen year: 2021 link: https://openreview.net/pdf?id=Qa3uS3H7-Le """ def __init__(self, **kwargs): """Initialize the Aristo-v4 dataset. :param kwargs: keyword arguments passed to :class:`pykeen.datasets.base.ZipFileRemoteDataset`. """ super().__init__( url="https://zenodo.org/record/5942560/files/aristo-v4.zip", relative_training_path=pathlib.PurePath("train"), relative_testing_path=pathlib.PurePath("test"), relative_validation_path=pathlib.PurePath("valid"), **kwargs, )
@click.command() @verbose_option def _main(): for cls in [AristoV4]: cls().summarize() if __name__ == "__main__": _main()