README.md

metadata

language:
  - en
pretty_name: YourDatasetName
tags:
  - common-crawl
  - web-crawl
  - news
  - corpus
  - article
license: other
size_categories:
  - 10M<n<100M

YourDatasetName

Summary

This dataset is a 10% English-language random sample derived from Common Crawl, provided for research and reproducibility. This sample was created of this dataset

Data origin & rights

Derived from Common Crawl. Copyright in the underlying content remains with the original authors/publishers. Common Crawl Terms of Use apply. oai_citation:4‡Common Crawl

What's inside

• Format: Parquet shards (one file per year)
• Columns: year, source_domain, title, text

Stats:

• Size: ~23GB
• ~ 13’399’600 articles

year	number of articles
2016	122'760
2017	1'750'263
2018	1'700'226
2019	2'239'029
2020	3'690'418
2021	3'896'904

Intended use

research / benchmarking / reproducibility

Privacy / PII notice

This dataset may contain personal data and sensitive content as it originates from web crawl data.