lewoniewski's picture
Update README.md
0e1ab4a verified
metadata
license: apache-2.0
language:
  - en
  - ar
  - de
  - es
  - fa
  - fr
  - it
  - ja
  - nl
  - pl
  - pt
  - ru
  - sv
  - vi
  - uk
  - zh
pretty_name: Grokipedia v0.1 and corresponding Wikipedia articles in 16 languages
thumbnail: >-
  https://huggingface.co/datasets/lewoniewski/grokipedia-wikipedia-16-languages/blob/main/grokipedia-wikipedia.png
size_categories:
  - 100K<n<1M

Dataset description

This dataset contains a mapping between Grokipedia v0.1 article pages and the corresponding Wikipedia article titles across 16 language editions (based on Wikipedia and Wikidata dumps from 1 November 2025). Each record includes:

  • The URL of the Grokipedia page: grokipedia_url
  • Wikipedia titles in the following languages (if they exist): ar (Arabic), de (German), en (English), es (Spanish), fa (Persian), fr (French), it (Italian) , ja (Japanese), nl (Dutch), pl (Polish), pt (Portuguese), ru (Russian), sv (Swedish), vi (Vietnamese), uk (Ukrainian), zh (Chinese)

Data construction

To align Grokipedia articles with Wikipedia:

  • Each Grokipedia title was first matched to an English Wikipedia title.
  • If the English title pointed to a redirect, the redirect was automatically followed to its canonical target page.
  • That canonical page served as the reference for retrieving titles in the other 15 Wikipedia languages (using Wikidata).

Unmatched articles

Not all Grokipedia pages could be paired with a Wikipedia article. The dataset includes also list of 3,536 Grokipedia URLs (in a separate file) for which no corresponding articles were found directly in Wikipedia.

Use cases

This dataset can be used for (but not limited to):

  • Cross-platform knowledge base comparison
  • Topic coverage analysis
  • Multilingual alignment studies
  • Benchmarking AI-generated encyclopedic content against human-edited sources

See also