[TAC] Stop Human Spam

[TAC] Stop Human Spam [Paid] 1.4.8

No permission to buy ($19.00)
hmmm, I've not really seen this behaviour from humans, but bots often try anything (humans often less so, human try to put real links down, and try to get around the detection by being sneaky, which they often give up after 3-6 attempts)
So these are newly registered members then (do you use fbhp?)

I can extend the functionality to look at all account details fields, and stop links / sneaky URLs
But the custom ones you use on registering will need to be solved with regex
StopHumanSpam only looks at users content after the registration process has been complete
 
Last edited:
Yes, newly registered members (humans). Yes, I use fbhp.
I don't mind keeping it like this. If you would like to extend your add-on I would definitely try that. But you don't have to do that specially for me.
The regex works good, But if they just write their business name in the location or occupation field and there is no way to prevent that. Only manual moderation works and you have to keep an eye on all profiles the whole time.
 
If it's not a url (not even a sneaky one), it's pretty hard to discriminate / detect it
I'm only trying to prevent the link layers and common spam words, business names could be just about anything
 
If possible they post urls or sneaky urls but because of the regex in the custom user fields they use their business names now.
 
Humans are often innovative, if they are not making a mess of your forum, not leaking out links, not offending and not annoying other users, putting their business name some where they're not supposed to is probably going to be more work to try to stop them than it's worth (and may end up putting off your "good" forum users).

I would just let it go (people only look at profiles when they the user has something interesting to say anyway, spammers benefit almost nothing from this)

Country: Xenzine Articles
DOB: TAC Anti-Spam

:)

Although you did mention citation, hmmm
The only way that wouldn't bother your other users I can think of is to make your fields drop down boxes, rather than string fields, limiting the users choices
That too might get a bit cumbersome for non-spam users
 
Last edited:
I know, It's a difficult situation.
Yes, I was thinking about drop down boxes. I don't need the occupation field but I do need the location field.
I can live with it as long if it's less than 50% that does this. But yesterday in my tests it was already more than 50% that spammed their profiles.

What I could do is hide profiles for unregistered users so only registered users can see profiles.
Then I take 90% of their power away.

*edit*

I made the profiles Members only and enabled the update profile option again, now they can spam in profiles. It's only visible to members so they don't have much benefit from it.
 
Last edited:
yep, similar to what @Fred. mentioned

They never use to do this, because they are not real links, they get no "link juice", but it seems like they've stopped caring and just post anywhere they can.
So if this is common, I'll update the plugin and protect all account areas from links / sneaky urls

btw, I know you hate me saying this, but this really does smell like bot again to me (the trend change for many forums ... that's not usual for humans, and the not caring where / why /how ), at very least semi automation
They have also doen this more than 1000 times
https://www.google.co.uk/webhp?sourceid=chrome-instant&rlz=1C1CHBF_en-GBGB730GB730&ion=1&espv=2&ie=UTF-8#q=usaonlineshop/testo-ultra
and this is probably after half the sites have cleaned the spam mess up, it probably had more than a few thousand of those links (would you sit down and do that 3000 times, only to get most of them deleted??)

Wow, of those > 1000 spam posts, most of them were done within the last 10 - 20 hours
I am absolutely certain that's a bot
Could you register on more than a thousand forums, set up accounts, activate your email, and sit there and find the fields where you can post links ---- in 24 hours

It's fairly convincing right, and it's not just one person / forum where we have seen this behaviour change, using recaptcha is not evidence that it is not a bot.
Trend changes like these interest me (mainly because trend changes like these are due to crowd behaviour, and that crowd is usually bot users, humans don't usually all change behaviour all at once, but an update in software means many bot users do)

- That said, it doesn't mean I wont update it, I'll try to cover those account areas next version, just have some tweaking to do for semi-automation detection with fbhp first
 
Last edited:
Yer, but I only got the one instance of spam... which I wouldn't get from bots doing this. Not for the daily traffic received. It was that I got the one instance, human based, even getting their URL in the profile, which I reported to you. Like you said, they don't care any longer and just want scraping recognition.
 
Lots of different types of botters, and lots of botting software. Some botters even call their selves white hat, because they build the links / content slowly against forums with relevant content
.. if you use a bot, as far as I'm concerned, it's black hat

Some botters just do profile things, the reason for this is to avoid detection. Most forum admins only look at forum posts, they don't bother going through every profile to check for links / spam there.

Not all bots are leaking through theses days, googles no captcha is still quite strong and APIs detect quite a few. But neither of those are full proof
 
Last edited:
Yes
True, That one must he a bot.
I made the profiles members only, Problem solved. By doing that I also solved another problem with members that want their profile removed from Google. Every week I get emails about that.
They can still spam in profiles but they won't benefit from it because it's members only.
 
@tenants we got the following report from a client of ours:

If you go to this thread and make a quoted post (as an anonymous user) with a word that is in the TAC stop human spam list, you will see that both the post you made and the post that you quoted are pulled into the moderation queue. Since the word "krause" is in the list, you can post the sentence "Was Krause there?", to see it in action.

It seems to happen when a forum is on full moderation (posts and threads have to be reviewed before going live) and a new post is made on a thread in such a forum that also contains a word in the TAC stop human spam list.

As well as:
We have discovered another way that posts get pulled back into the moderation queue. If someone quotes a post and does not complete the captcha correctly then the quoted post gets pulled into the mod queue. If no post is quoted, then the last post on the thread will get pulled into the queue. This only happens if the post includes a word in the TAC stop human spam list and the captcha is completed incorrectly.
 
okay... I'll try to reproduce it and find a fix, I think this is more important that extending the accounts areas, so I'll get this one out first, thanks for letting me know

There must be something strange going on with quoting that makes it pick up both ids
 
If you go to this thread and make a quoted post (as an anonymous user) with a word that is in the TAC stop human spam list, you will see that both the post you made and the post that you quoted are pulled into the moderation queue. Since the word "krause" is in the list, you can post the sentence "Was Krause there?", to see it in action.

It seems to happen when a forum is on full moderation (posts and threads have to be reviewed before going live) and a new post is made on a thread in such a forum that also contains a word in the TAC stop human spam list.

Just trying to reprduce this, to get this clear

The forum has to be under moderation
The user must be a logged out user
The user must use a banned key word
(I take it you have ticked the option to moderate banned key words in shs)
The user must quote the first thread post in that thread

Have I missed anything?

- It's quite tricky to reproduce (A bit edge case, but will fix once I can reproduce)
 
Last edited:
Since in both these cases, the forum is under moderation, I believe this simple change will fix it:

Code:
        if($forum['moderate_replies'] == 1) // new, prevents edge case bug

        {

            return parent::actionAddReply();

        }

I'm just testing it now, seems to work. We just allow the core to moderate and do nothing with shs (since the reply will always be moderated by the core anyway), in this case we skip the shs checks
 
Since in both these cases, the forum is under moderation, I believe this simple change will fix it:

Code:
        if($forum['moderate_replies'] == 1) // new, prevents edge case bug

        {

            return parent::actionAddReply();

        }

I'm just testing it now, seems to work. We just allow the core to moderate and do nothing with shs (since the reply will always be moderated by the core anyway), in this case we skip the shs checks

I'll go ahead and make that change on his site and see if he's able to reproduce the error still :)
 
Top Bottom