funsec mailing list archives

Re: Comment Spam: new trends, failing counter-measures and why it's a big deal


From: Gadi Evron <ge () linuxbox org>
Date: Tue, 14 Feb 2006 01:34:10 +0200

Stephen J. Smoogen wrote:
On 2/12/06, Gadi Evron <ge () linuxbox org> wrote:

Recently, new bots rendered current anti spam techniques for blogs
almost useless. Here is a short write-up on the subject of comment spam,
referrer spam and what's currently happening in that area.



Nice summary of what is going on with blog attacks. I havent done much
with blogs due to job reasons.. so hadnt really kept up with what the
latest attacks were. They seem to be parallels with website defacement
for profit, SMTP spam, and other crimes.

I was wondering about that smart "bot".. at what point does it become
cheaper to "employ" 1,000 phillipine children with English skills and
have them run through a bot-net to hide their origins.. than develop
an auto-bot that posts spam for you.

Doing automated searches through whois for obviously fake entries and
going from there to  search and verify messages to confirming that
email addresses are correct. Greylisting/whitelisting software might
also have some affect (if one can legally share that data). Say in
this way:

Being goes to blog.
 Being decides to post to blog.
 Being is given a EULA which basically says "Here are our posting
guidelines. You give up your right to anonymity if you wish to post
here."
 Being is sent a cookie with certain data in it, and is put in greylist.
 Server stores data on IP address, post data, and IP addresses in post.
 Greylisted items are posted on delayed time (or after moderation).
 Server sends greylisted aggregated data to central server (for
pattern matching AI that hey this same URL/IP address block was
embedded into 200 blogs today).
 Posting X amount of times moves one up/down from greylist to
whitelist blacklist using a bayesian scoring technique based partially
on keywords, and partially on non-whitelisted URLs.
  Client servers poll central server regularly for data to be added to
black/grey/white keyword-URL lists.

The central server is mainly to help multiple private blogs clear out
bad nets in a short order.. it would not be needed on a large central
blog aggregator that could act as the central server itself.

I believe you have the right of it. Still, there is no "magic bullet" with spam or pretty much anything else in this world.
:/
_______________________________________________
Fun and Misc security discussion for OT posts.
https://linuxbox.org/cgi-bin/mailman/listinfo/funsec
Note: funsec is a public and open mailing list.


Current thread: