Grants talk:IdeaLab/Percentage of abandoned articles metric

At what threshold of time with no edits would an article be considered abandoned?

This very much depends on the topic and the quality of any given article. In some areas a solidly researched and written article could sometimes go for years or even decades without needing substantial reworking. Some examples and rough calculations are at de:Benutzer:Karsten11/Die Grenzen der Wartung. --HHill (talk) 17:20, 11 July 2018 (UTC)

I wonder if the threshold for an abandoned new short article can be approximated by

$k*{\frac {\text{text in article}}{\text{total text in wiki}}}*({\text{edit frequency}})$

where k is just a tuning constant. An old large article would not necessarily have the same threshold, but for an percentage of abandoned articles this would be sufficient. (It would be somewhat better to track normalized changerate in articles, instead of tracking whether the articles are actually abandoned.) — Jeblad 22:09, 14 July 2018 (UTC)

This should probably not be an edit frequency, but a edit distance. An edit distance is far less sensitive to random fixes, like spelling corrections. — Jeblad 04:52, 27 July 2018 (UTC)

"abandoned" may not be a helpful or appropriate word

I find this interesting, but I think this loses sight of the logical goal for an encyclopedia. The goal is to produce a great reference tool, not merely to edit endlessly. (This is more likely for certain articles than for others.) An article may have reached a point where it is quite adequate on its specific topic, so editors may stop editing. This is not necessarily a problem, but may simply be a sign of good editing by experts. I am not concerned about "abandoned articles". A mathematical formula can tell us how many articles have not been edited in a long time. But such a formula does not tell us if the articles need more editing. Pete unseth (talk) 16:48, 20 July 2018 (UTC)

"Abandoned" is seen from the point of view of edits. But if an article is excellent, it may not have any recent edits, but the article may be read by many users. That is not "abandoned". Pete unseth (talk) 16:05, 11 March 2022 (UTC)

Similar existing projects

Don't we already have something similar at w:en:Wikipedia:Dusty articles, and w:en:Wikipedia:Database reports/Forgotten articles? At least for English Wikipedia anyway. -- OlEnglish ^(Talk) 19:55, 29 July 2018 (UTC)

Questions about your idea

Hi FDans (WMF) and Linus Godwin korah - thanks for submitting your idea for the Inspire Campaign (and for expressing interest)! I just have some questions I'd like to ask about your idea:

1. Do you have any ideas on how to develop this idea further?
2. How would you respond to concerns raised here and on the Idea itself, like:
   1. How would you differentiate good versus bad faith edits to "abandoned" articles?
   2. How would you filter out articles that are simply "done" (that is, there is not much more that can be reasonably added to them) from articles that are genuinely being abandoned?
3. How long do you think a project like this might take?
4. Is this something you would propose that your team works on, or a side-project?
5. Where do you think the research might be hosted in the end?

Potentially you may also want to launch a consultation to get the larger community's opinions on this too; I'm curious if this is an idea you're thinking of raising at the Foundation or one you'll take on as a side project? Thanks for submitting! Joe Sutherland (Wikimedia Foundation) (talk) 00:36, 10 August 2018 (UTC)

@FDans (WMF): Hey Francisco, I wanted to check in to see what kind of thinking or planning you may have been doing around this idea. Feel free to respond here and if you'd like to chat further, feel free to reach me by e-mail (cschillingwikimedia.org). Thanks, I JethroBT (WMF) (talk) 21:33, 13 November 2018 (UTC)