metadata

annotations_creators:
  - found
language:
  - en
  - de
  - fr
  - es
  - it
  - ru
  - ja
  - zh
  - pt
  - nl
  - pl
  - ar
  - tr
  - sv
  - cs
  - fi
  - ko
  - uk
  - hu
  - da
  - el
  - 'no'
  - ro
  - sk
  - vi
  - bg
  - et
  - lt
  - lv
  - sl
  - sr
  - hr
  - ca
  - id
  - ms
  - th
  - he
  - fa
  - hi
language_bcp47:
  - en-US
  - en-GB
  - zh-CN
  - zh-TW
  - pt-BR
license:
  - cc0-1.0
multilinguality:
  - multilingual
pretty_name: Public MediaWiki Collection
size_categories:
  - 1M<n<2M
source_datasets:
  - original
task_categories:
  - text-generation
  - text-classification
task_ids:
  - language-modeling
  - topic-classification

Dataset Card for Public MediaWiki Collection

Dataset Summary

This dataset contains 1,662,448 articles harvested from 930 random public MediaWiki instances found across the Internet. The collection was created by extracting current page content from these wikis, preserving article text, metadata, and structural information. The dataset represents a diverse cross-section of public wiki content spanning multiple domains, topics, and languages.

Languages

The dataset is multilingual, covering 35+ languages found across the collected wiki instances.

Dataset Structure

Data Fields

This dataset includes the following fields:

id: Unique identifier for the article (string)
title: Title of the article (string)
text: Main content of the article (string)
metadata: Dictionary containing:
- templates: List of templates used in the article
- categories: List of categories the article belongs to
- wikilinks: List of internal wiki links and their text
- external_links: List of external links
- sections: List of section titles and their levels

Data Splits

All examples are in a single split.

Additional Information

License

This dataset is released under Creative Commons CC0 1.0 Universal (CC0 1.0) Public Domain Dedication.

To learn more about CC0 1.0, visit: https://creativecommons.org/publicdomain/zero/1.0/

Dataset Curators

nyuuzyou