Web Spam Taxonomy

Paper on current spamming techniques

[stevenmusumeche]

The following paper was written by 2 professors from Stanford University. It contains a very interesting summary of current search engine spam techniques.

Web Spam Taxonomy [dbpubs.stanford.edu] (.pdf 310k)

[dbpubs.stanford.edu:8090...]

[edited by: pageoneresults at 4:57 pm (utc) on April 12, 2004]

[skipfactor]

Their example of meta keyword spamming is funny--geez, glad I quit using the keyword tag:

<meta neame="keywords" content="buy,cheap,cameras,lens,accessories,nikon,canon">

[SlowMove]

I don't know if the professors are aware of it, but a lot of webmasters that are legitimately targeting traffic are using the same techniques that they describe. They have to play by the same set of rules just to stay in the game. And if the rules change, the spammers and the legitimate webmasters will make the necessary changes.

[xbase234]

The definition of spam is flawed in this paper.

Putting titles that accurately describe page content is about good design and usability principles, not spamming. And just because I know that this title is the first thing a user will see when it pops up in a SERP doesn't make me a spammer. It makes me a good designer.

[SlowMove]

I think we're okay unless they convince everyone not to do business on the Web. As long as there's business, we will figure out how to make money.

[your_store]

The definition of spam is flawed in this paper

Agreed. They start the paper by saying spamming is..

any deliberate human action meant to trigger an unjustifiably favorable relevance

But at the end of the paper they say..

..sites serve to search engines a version of their page that is free from navigational links, advertisements, and other visual elements... This kind of activity is welcome by the search engines

Well you can't have it both ways. Either spam is changing your page to rank higher or it isn't. I've always loved academia ;)

[Teknorat]

The link is dead already- anyone got a copy of the article?

[deimosaffair]

It worked fine for me, i just downloaded it now (19.25 GMT on my clock)

Check your link, or make a copy-paste on the link to see if you got it botched.

[SlowMove]

More people would probably read the article if they used html instead of a pdf format. I'm guessing that optimally this file should be broken down into about 12 to 15 static html pages. However, keyword density seems to be about where it should be.

[photonstudios]

There is no such this as spam. Game Over.

[xbase234]

If they broke this paper out into HTML it would be spam, as they defined it! A dichotomy in and of itself.

I'm surprised these guys haven't read this forum, and aren't concerned that their research is in question. If I wrote an argument like this, I would be here trying to further prove my point. But I wouldn't write an argument like this, because it doesn't make sense.

Was this some kind of April Fool's prank?

[stevenmusumeche]

No Joke - this is really how they think. Acadamia, in general, is out of touch with reality.

[gazza_tt]

question- dont search engines automaticaly penalize sites that spam? if so then whats the point in doing it!

[photonstudios]

question- dont search engines automaticaly penalize sites that spam? if so then whats the point in doing it!

Yes they automatically penalize your site, don't spam. Only write high quality content articles and you will be REWARDED by google and other se's ;) Also your users will appreciate the great content and your site will be successful, guaranteed!

[Teknorat]

Oh God that was funny. :D

[bakedjake]

dont search engines automaticaly penalize sites that spam?

Not generally, no.

[aeve]

I'm no SEO, but I thought it was a pretty interesting article. This is more or less a new field for academia and I think there should be more studies like this. Sure, it's simplistic and they take a pretty broad definition of spam, but you've got to start somewhere.

I thought the suggestion of search engines spoofing user-agents was interesting (combat spam by acting like a spam-bot). It seems like a slippery slope but it's an easy fix for cloaking.

Adam