What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 134 Stack Exchange communities.

As it is, we have to moderate edits by new users, so it seams reasonable as a stop-spam mechanism to require all external links from <50 rep users moderated.

A lot of the spam I see works like this:

  1. User joins
  2. User searches for all relevant posts.
  3. User pastes links to their product.

And 99% of the time, its the same bad link.

Once we get the spammer, they come back with a different name, and repeat from #1 with the same link.

However, if either links themselves required moderation to be posted, or all articles by new users with <50 rep with edits were automatically put into the moderation queue ( so they wouldn't appear till approved ) it would limit the amount of user-visible link-spam.

Even a nicer thing to have would be heurisitics on the users links, ie:

  1. This link has appeared more than once in the last $period
  2. This user has pasted >1 external link in the last $period
  3. This user has pasted this link >1 times in the last $period

etc, to quickly put a gag on spammers.

share|improve this question
    
There are already limits to the number of links a new user can insert into their posts. –  Adam Davis Feb 15 '11 at 22:08
3  
The limits are fine and dandy, but it doesn't stop the most common spam case: links scattered across a dozen answers. –  Kent Fredric Feb 15 '11 at 22:11
1  
and even worse, links scattered across a handful of different users. –  Kent Fredric Feb 15 '11 at 22:13
    
What if the link is to new (new article or software), relevant, valid content and relates to multiple posts. Should we punish them? –  Lucifer Mar 25 '13 at 14:09

2 Answers 2

...they come back with a different name, and repeat from #1 with the same link.

Not if you make sure that the spam link gets blacklisted.

Especially now that it's easy to notice these cases via the /review route (they're almost always low-scoring answers), I don't see this as a problem in need of a solution. Spam seems to be dealt with in a timely fashion, so as long as the community has the tools necessary to take care of it when it does pop up, adding any sort of barrier to potentially useful answers seems unnecessary.

I'd agree that anyone going around posting the same link repeatedly is likely spamming, but I also don't see that as being a prevalent enough problem that the system needs to try and address it with automatic functionality.

share|improve this answer
    
I thought of blacklisting, but I believe if blacklisting was the ONLY protection, that would simply encourage spammers to seek varying URL masking tools ( ie: hide it behind a new bit.ly url every time ). –  Kent Fredric Feb 15 '11 at 22:29
    
Also, its not being targeted effectively, look at this user I found today: stackoverflow.com/users/496300/datasanj , has been a member for 3 months, and in that time has posted 6 answers, all of which are identical in nature. This sort of spam behaviour can only be detected by long terrm trending. –  Kent Fredric Feb 15 '11 at 22:31
    
@KentFredric If they started masking the URLs, then that might be an issue. I haven't seen any indication of that, though. As far as that user goes...I guess I'm not seeing a huge fault with the system. Six answers across the whole of Stack Overflow is really insignificant, so you coming across one of the posts, feeling that there was an issue, and flagging it seems to resolve that problem in an effective enough way. –  Tim Stone Feb 15 '11 at 22:42
    
Interface change influences behaviour, and if spammers discover they can't re-paste the same URL any-more because its "blacklisted", then they will do one of 2 things, comply with the system and stop posting bad urls, or try to subvert it by masking the URL. My idea was to try something that targeted both of these behaviours. –  Kent Fredric Feb 15 '11 at 22:45
    
Also, I'm a perfectionist, your argument about "6 urls across the whole of stack overflow being insignificant" can be really scaled to any undesirable behaviour on stackoverflow which degrades quality. Such as: users editing other users posts to be their own : stackoverflow.com/edit-suggestions/5866 . Sure, maybe nobody will notice, but its still awful, and should be stopped if possible. –  Kent Fredric Feb 15 '11 at 22:49
1  
@KentFredric It's not about people not noticing at all, it's about the problem being small enough that a few people can notice it and take corrective action before it gets out of hand. I feel like this works well enough that it doesn't warrant additional moderation burden (as a queue would) or development time. I do agree that it'd be nice if the system was able to ensure that everything was done correctly and malice-free, but I don't see that as a practical design goal. –  Tim Stone Feb 15 '11 at 23:17

Perhaps what is required is to limit the overall number of links they can post, in addition to one per post? So while they remain under 50 points, they can post a maximum of three links concurrently?

I realize this is one more point for the team to moderate, but the idea here being trying to help find some middle ground. At least by limiting to a hard number, the system is fairly self regulating.

share|improve this answer
    
"they can post a maximum of three links concurrently" ? eh? I don't get what you mean by that, "oh, I need another link, lets go back, find where I last used one, and delete that!" ? –  Kent Fredric Feb 15 '11 at 22:12
1  
If it causes them to quit spamming quite as often, or causes them a lot of headache to be spammy, sure. Spam is effective when it's automated. When it's a hassle it's called advertisement. –  jcolebrand Feb 15 '11 at 22:16
    
I just see that limitation being more detrimental to new users, and it still wont solve the problem of them scaling out by just having more users. –  Kent Fredric Feb 15 '11 at 22:18
    
@Kent ~ So you're id'ing two problems: multiple posts by the same user or group of users (aka sockpuppeting) and one user spamming multiple times. I'm proposing a hassle for the second one, in addition to the other hassles. For the sockpuppeting, there's already lots of tools for that one, such as /review (which you should be using already) –  jcolebrand Feb 15 '11 at 22:23
    
But they're both really the same behaviour pattern and can be targeted by the same mechanisim. And /review doesn't help target sock-puppeting very well, you see the occasional evidence of spam in there, and when you see it, you suspect the user of being a spammer, and then visit their profile, and THEN discover their spam history. –  Kent Fredric Feb 15 '11 at 22:26
    
And then you work out what they were spamming, and then search for it, and discover several other users were spamming that too. –  Kent Fredric Feb 15 '11 at 22:27
    
So I think I see what you're saying but I feel like this is a different way to achieve the goal you're asking for. I'm just one guy, doesn't mean I know all the answers. –  jcolebrand Feb 15 '11 at 22:28

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .