In T318092: [M] Exclude certain sections from having topics in the section topics pipeline, we made the first iteration on section exclusion.
Now we should exclude further titles as per T279519#7368694.
Tasks
- compile the list of titles via section alignment
- add them to the current blacklist
- some wikis have a community-configured list of sections that should be excluded from link recommendations. We ought to use these - for more information see T311730#8421980
Outcome
All manually curated denylists are propagated to all Wikipedias where section alignment is available.
This resulted in:
- a more extensive coverage of obvious denylisted sections, such as references, references and external links, references and footnotes, references and notes
- some list sections, such as awards and nominations, discography
- total rows with current denylist = 1.2 B (1,198,537,554)
- total rows without denylist = 1.4 B (1,400,403,749)
- difference = -200 M rows
NOTE: other potentially meaningful ones slipped in, such as biography, criticisms, family relations, geography, history.