> Potentially bad keyword in body, potentially bad keyword in title, potentially bad keyword in username ---------- Potentially bad keyword in body - Position 0-14: roblocodes.com Potentially bad keyword in title - Position 0-14: roblocodes.com Potentially bad keyword in username - Position 0-14: roblocodes.com
For the past year or so, MathOverflow at any rate has had a flood of spammy new users. Typically, the username is the name of a company or product; occasionally it is offensive. The user is created, and fills out its autobiography with advertising information. They get the "autobiographer" badge and are visible on our main page for several hours in the list of "recently awarded badges". Presumably the goal is search engine manipulation.
@TimCampion It's possible, but Charcoal doesn't have anything in place to do so at this time. I'm also not aware of any existing projects which do so. Unfortunately, there isn't any notification when a new user is created, as far as I'm aware. Having such notification would make it quite a bit easier to accomplish something along the lines of what you're wanting. It would be useful, at least for some persistent trolls/spammers, to be able to identify the new accounts early.
Effectively, that leaves it up to the individual site moderators to decide if they want to spend time cleaning out such profiles.
In addition, destroying the profile, even with also applying a 365 day suspension (what we typically do on SO for spammers and trolls), doesn't prevent the user from immediately recreating their profile from the main SE account, along with any spam in their profile which gets copied to the new site profile from the main SE account.
Note: if the spammer/troll is still active and interested, it's not uncommon to have the profile recreated quite quickly after being destroyed. In fact, recreating the profile is nearly a no-brainer for the spammer/troll, only requiring a couple/few clicks. When destroying such profiles, I've generally found it more effective to wait a couple/few days prior to destroying the profile to the point where the spammer/troll isn't really paying attention to the account/profile.
@TimCampion If you're wanting those cleaned out on the site(s) you frequent, I'd suggest posting a Meta and getting buy-in from the moderators, as it's their time which will be expended doing it. I'd note that userscripts can get the work required for actually suspending/destroying the user down to only a few clicks, but the moderator still needs to take the time to evaluate the profile to see if it really should be destroyed.
That evaluation can be made easier through automation (still would need someone to write the code), but the process really does need to have a human in the loop, because there are legitimate profiles which will match most/all metrics which might be used to auto-select them.
@Makyen technically there is the /event endpoint on the API api.stackexchange.com/docs/… but you have to call that a lot for each site ... and you can't filter on user_created only.
@Makyen I don't think we have a userscript that adds a report user profile as spam button, right? We could maybe make something that finds the fiirst post of the site and then mod flag from the API with a customtext and link to the userprofile.
@rene I hadn't looked at that endpoint. It could be useful, but, ouch, that's a lot of polling if you want to cover all the sites. That would require >30k API requests to cover all sites for 24 hours, really ouch. Assuming all that is wanted is newly created users, it would probably be better to poll using /users with the results sorted to give most recent creation date first.
@rene I'm not aware of one which has a "report user as spam" function, although I do have a couple which have a "destroy spammer" button. :)
It would be possible to create a userscript which allows them to be reported, but given that doing this really requires buy-in from the site mods, it might be better to have the profiles reported into a (private?) chat room, rather than as flags. Using chat would allow whichever mods ended up wanting to do it to go through the notifications (preferably with some amount of automation), without requiring the other mods to skip/ignore/wade through a bunch of custom flags.
I'd expect that unless all the moderators on a site agreed, you'd get some which were annoyed by the influx of a large number of custom flags asking for the destruction of a bunch of spam profiles.
If chat was used, there would be some additional logistics to work out, but it's probably do-able.
