Recent Revision History of Featured Articles on Wikipedia

The dataset is composed of recent revisions (from July 21st 2020 to Dec 10th 2021) of 6,039 featured articles on Wikipedia. Each page is stored in a .json file with all associated revisions contained within, sorted from most recent to oldest. Each revision contains metadata as well as the full wikitext content.

Extra fields were added to revisions:

content-plaintext: Parsed wikitext content (using python's mwparserfromhell package)

comment-plaintext: Parsed plaintext of comment

content-diff: Sentence-level diffs of current revision's content and the previous revision's content, produced by the difflib python package (unified diff)

content-plaintext-diff: Similar to content-diff, but with content-plaintext as input

reverted: Indicates whether the revision was reverted by a later revision

Total size after decompression: ~ 59.2 GB