User:Dc987/Dataset for the Rocket article

Labels

 * 'bad' - automatically identified edits that generally require a revert (some human verified)
 * 'good' - automatically identified good faith edits (some human verified)

Additional extra information (revisions in these lists are already in the 'good' or 'bad'):
 * 'good (corrected by user)' - good edits (automatically identified, corrected by a human)
 * 'constructive (corrected by user)' - good edits, questionably good edits, partial vandalism reverts, etc. (human corrections)
 * 'bad (corrected by user)' - bad edits (automatically identified, corrected by a human)
 * 'bad (verified by user)' - bad edits (automatically identified, verified by a human)
 * 'good (verified by user)' - good edits (automatically identified, verified by a human)

Bot source code
See: User_talk:Dc987test

Usage
Add following URL before the revision ID to see the diff:

Example: http://en.wikipedia.org/w/index.php?diff=prev&oldid=7267547

In python you can execute this data set as a code and get a dictionary of sorted revisions lists. Here is an example that prints html refs to the Wikipedia diffs for the 'bad' revisions:

There is also a helper python module available: that provides is_good_or_bad helper routines.

Latest dataset version
See