User talk:Doc James/COI
Note about Researchoncommand[edit]
see this. The original article he created was deleted for being an advert. Jytdog (talk) 05:37, 17 February 2015 (UTC)
COI detection at AfC and nnnn999 users[edit]
Hi, Doc James. I spend a bunch of time at AfC. My gut estimate is that the majority of the articles involve COI. I tag user pages for many of these. Because AfC gets newbies who are writing articles, they are logically ignorant of the COI policy. Most folks respond either by complying, or stopping editing. (The latter usually for commercial promo articles.) What I have seen is what looks to me like a paid editing pattern, which I'm pretty sure has been noted before. In these cases, a new user name is minted with a name+number pattern for each article. Once the article goes into main space, that user no longer edits. Most likely the contract is to "get an article into Wikipedia." Since these are new users, they have to go through AfC. A new name is used for each new article, as a way to avoid creating a pattern. Of course, the anti-pattern is also a pattern -- one that could probably be detected algorithmically.
Now, what to do with this knowledge? I keep a running list of the folks I hit with a COI template on their user page. I've got about 4 score. For those the obvious COI indicator is the username, which is close to or identical to the subject of the article. These name+number usernames are opaque so I don't tag them. They only smack of COI/paid as a group, not individually. I could keep track of the name+number users that look like paid editing, but it seems to me that some code could do that more efficiently.
Thanks for listening. If this triggers your paid-editing-spidey-sense at all and there is something useful I can do, just let me know. LaMona (talk) 16:48, 26 February 2016 (UTC)