wiki spam/1

A case of wiki spam

At first the front page of the wiki was changed like this:

http://sandbox.merjis.com/index/diff?version=326

Notice the added links. They've added them using single square brackets [...] which on other wikis turns them into links. Not, however, on this wiki.

Shortly afterwards the spam was removed:

http://sandbox.merjis.com/index/diff?version=327

This was all done in the space of a few minutes from an AOL dial-up address (172.190.194.104), and it appears from the profile in the logs to have been done manually using MSIE.

The spam is for some well-known spammy domains, which are already included in one of the Movable Type blacklists I checked.

Adding, then removing?

Why should someone want to add and then remove a link? I have two theories for this:

  1. They're just testing out the wiki to see how it works, perhaps in preparation for using a script to attack the site.
  2. The "recent versions" link shows old pages, so even after the spam has been reverted (either by the spammer or by us), the link to their site remains, still visible to Google and still stealing page rank.

Note that I get an email whenever someone changes any page on any of my Wiki sites, so I can revert changes pretty quickly too even if the spammer doesn't.

What do I deduce from this?

There are a few things to deduce from all this:

  1. Either someone is paying roomfuls of monkeys to edit wikis manually, or else adding a single link to a wiki is so valuable (in terms of the stolen page rank and consequent artificial boost to search engine rankings) that it's actually worth an "SEO" person doing the work manually.
  2. Reverting a page does not revert the damage. If your wiki can show old versions of a page, then you're still allowing people to steal page rank. This must be a considerable problem for Wikipedia.
  3. Spammers are stupid (using the [...] single link form and not even reading the help page).

Solving wiki spam

This sort of low-level wiki spam is actually quite simple to resolve. I've now changed the Wiki to add the following <meta> header to old versions of pages, the recent changes lists, and a few other places:

<meta name="robots" content="noindex,nofollow" />

This tells Google not to index the page, and not to follow any links in the page. Of course, I don't want this on every page - because I like my Google page rank and I like to share it with worthy sites. I only want it on the old versions, which may contain unreverted spammy links.

It seems to work.

This Google search doesn't return any links for the spam-containing page, which hopefully means that Google is neither indexing that page, nor following the link to the spammer. (This would assume that the spammer was smart enough to have given a link, instead of using the wrong syntax).

More about the spammer

After a bit of investigation, I found the name and address of the antisocial twunt doing the spamming. Please consider not doing any business with this person.

Bob Harris 27 Old Tatham Holme On Spalding Moor YO43 4BN York GB

mailto:BH861839@aol.com, mailto:bob.harris@bslh.net

Phone: 01430 861 839

Same spammer hit this site too:

http://www.skynet.ie/~dach/cgi-bin/wiki.pl?Spam