---
pretty_name: "Wikidata5M"
license: "cc0-1.0"
tags:
- "knowledge-graph"
- "wikidata"
- "triples"
- "datasets"
- "tabular"
---

# Wikidata5M

## Overview
- dataset_id: wikidata5m
- license: cc0-1.0 (Wikidata)
- homepage: https://deepgraphlearning.github.io/project/wikidata5m

## Standardization
- split_aliases: {"dev": "validation", "val": "validation", "valid": "validation"}
- rename_map: {"head": "subject", "relation": "predicate", "tail": "object"}
- selected_columns: []

## Exported Artifacts
- train.parquet
- validation.parquet
- test.parquet

## Citation

Wang et al. (2021), KEPLER: A Unified Model for Knowledge Embedding and Pre-trained Language Representation.