Semantic spam Attack

Today i notice five new comments on my blog that passed spam-filter and waited for approving. They all seemed to be OK. Example is:

Yes we all know about that, but you published this on 2005, now it became very famous, moreover it very difficult to handle , all keys are very different so new people couldn’t work properly, its takes too much of time to take download, xp speed is better than vista in pc but in mac vista is best it works nice and speed too, every operation will done easily. Now lots of people are using, its try to hard learning this os but after knowing all its easy only, good nice information i studied in wikipedia, thanks for your blog.

on my deleted post About Windows Vista form Year 2005

But if you look at this post more precisely you notify that the is no point in this post. Is that possibly any question in this comment? Or maybe it gives us new information? Probably it is just emotional expression about my wonderful post about Windows Vista!?

Of cause NOT! My old short post from 2005 about future features of Vista deserves no attention in a year 2007. It is only refraining of sentences used in original post.

And there were at least five of such comments. I have even already approved some of them days earlier, because they looked like they were written by humans. But having five in one night, made me suspicious. I started my investigation!
I found and marked as spam them all on my blog!

I filtered all comment that just refrained my sentences. In a manner if i e.g. say “Java is cool” in one sentence, the comment was containing corresponding sentence like “I found java is cool too, thank you for you great post!”.

Also the were always Link in a name of Commenting person, here are some of them:

Look! Even design-template of the first an last is the same! And second looks as spammers site as it is.

Please buy nothing from that domains an don’t give them any traffic! Don’t give the evil spammers any chance!

Nevertheless, spammer seems to use very advanced tools! Look, I don’t think they wrote all of that comments using own hands. That to much work for spammer (Spamer don’t like much work, otherwise they would do something more useful). Also the result of the comment could be much better in poor human made case. And that comments have had the same structure.

Cause Englisher language has not really difficult grammar in case of sentence building, i thing that is fully automated generated spam. Also possible is that, that spam was made by humans used an very cool tool that generates much of the work!

It seems for me, we entering new Loop of anti-spam war! I wish you good day and successful battle!

P.S. Does anybody know how we can effective protect our blogs? My first intuitive solution is to restrict comments only to registered users. But it’s only the first one.

If you enjoyed this post, please consider to leave a comment or subscribe to the feed and get future articles delivered to your feed reader.

Comments

This has been going on for ages. It happens in two ways. One, the spammers hire third world workers at low pay to copy and paste and twist thing around to become human spammers, or human spammers that want link juice play just enough with words to look good so the comment spam slips through by the lazy blogger, which you are clearly not.

The other method is the creativity of spam bot developers, and yes, they are getting more and more creative, which requires vigilance.

The best comment spam fighter right now is Akismet. With a full version self-hosted blog, back that up with Spam Karma and Bad Behavior, or at least one of them. That helps filter the filter.

Also, realize that we can’t “get them all”. It’s not possible. They keep changing their methods and introducing new ones. Akismet takes time to learn, as the community teaches it, and things slip through during the learning process, just because of the numbers. Have you looked at the numbers recently? Spam represents over 90% of comments. Some days, it’s over 95%. Some will slip through, that’s the nature of the process. Just mark them as spam and speed up the process of the spam on your blog being removed from mine.

It’s team work. :D

Also, CAPTCHAs don’t work. Nor do torture tests, like the one you use. Sure it’s good for a while, but they know how to get around them, as you’ve proven. So why bother torturing your readers? Many won’t comment if they see these things.

Just be vigilant and let the best tools do their work, and do nothing to interfere with the conversation.

Hi, Thank you for comment! This is definitely not spam (And even so, i like it any way :) ) :) As you mean it could be of cause done by low payed workers, but I’m a bit sceptic about this. May be you can provide some further information, or evidences pointing to it? I think that spam can be generated in automatically, and of cause “math torture” on my blog can scare only very primitive crawlers. There are no much work to do to build such one “refraiser” (imho), any good informatics student should be able to do that. :)

An thank you for the tip about Spam Karma and Bad Behavior I’ll try them out!

Leave a comment

(required)

(required)