Why mark spam accounts as spam?

Spam patrol page:

Is it worth spending time doing this?


my understanding is all new profiles are hidden but new obs are not. So if the spammer is making posts they are worth banning or marking because once they are marked as spam the obs don’t show either. If it’s the sort of account that has a spam profile but doesn’t post, maybe it isn’t valuable to bother

1 Like

By definition, none of the accounts on the “Probably spam” filter have made observations.


is there value in ‘rescuing’ the real accounts that fall in there then? Or is that rare or they can get out anyway?


One reason I can think of is that many scientists would use iNat for data collection etc, and its difficult to filter through all the observations of preference and doing due diligence on the spam accounts with spam observations.


Debatable. These profile descriptions are being hidden, so they are not serving the spammer’s purpose, but they are still visible to curators, which is annoying to curators, and as Charlie pointed out, these are still active user accounts that spammers could use to make more spam on iNat. You could also argue that by marking these accounts as spam you are helping train Akismet to recognize the kind of spam we receive at iNat, but given how bad a job Akismet seems to do on our spam and the enormous corpus of training data they presumably have already, you could just as easily argue that Akismet doesn’t need more data. Akismet’s a bit of a black box. You could also argue that by marking spammers you are helping iNat build it’s own corpus of spam that we could use to make our own classifier at some point in the future, but I have to admit there’s little enthusiasm for that on the development team at present.

So, up to you. I personally mark a few spammers a day out of a compulsive desire to get rid of them, but I don’t think my motivations for doing so are all that rational.


That being the case, the following is probably not worth a feature request. I had been thinking though, two things would make spam patrol (for what it’s worth) much more efficient:

  1. ability to search user description contents. ?q=“To learn more, go to website:” would find about 95% of the stuff I have been flagging.
  2. ability to select and flag batches of said results – knowing that one would still want to do a final visual scan for any false-positives first.

The Recent Users page is a great tool for all kinds of other queries, though, so definitely keep it around, either way.

And know that I also share some of that irrational compulsion! :stuck_out_tongue_winking_eye:


I sometimes rescue an account when it’s crystal clear that it’s legit, but it’s visually lost in a sea of spam accounts, and I want to make sure I (and others) don’t accidentally flag it as spam. Otherwise I leave them alone and let them post (to get out of the list) or not.

1 Like

I don’t know how Akismet works and how often it runs on the same recent users, but some spammers create an account and fill the profile with spam later. I’ve gone through the same recent users several times and I always found new spam. The usernames are always like this name_number_name.

1 Like

Yes, I notice the same thing. Every so often I’ll go back through the past few weeks, and find an occasional spam account that I’m sure wasn’t there before. The descriptions almost always end with “To learn more, go to website:” – which I wish was searchable in the filters.

Even the ability to mark a bunch of things (think a check box) and then hit the button once to mark as spam or not spam, then confirm. Would save quite a lot of scrolling down and up.

1 Like

Yes, exactly my wish too!