'Web Noise' - the search engine killer

8 May 2002 | flamefront

[flamefront]

I have discovered (perhaps I am not the first) a huge vulnerability in our use of the Web and for the sake of future discussion I will coin a term and call it 'web noise'. While concerns about computer virus's and email scripts dominate the news about computer attacks, a subtle version has been introduced by what appears to be folks inspired by rap lyrics. Imagine one day you want to search for some information and you get the garbled text results like the following large sites illustrate:

• freakystylz.com
• yuppychat.com

• web.iwasateenagewolfman.com

As you drill down the links int these large trees of random text web pages the leaf nodes end up with references to odd places on commercial sites like amazon.com.

Certainly one can put screens on a firewall or server to limit access to certain internet domains supporting this web noise but search engines in their ever greater need for comprehensiveness will get tripped up by the exstensive web noise from sites like these that could clog the searches entirely at some point. The techniques of screening out domains that apply to overcoming 'denial of service' attacks would not be effective for search engines.

[flamefront]

Discovered after finding this showing up in a search I did yesterday on politically sensitive information.

For your information and discussion.

[flamefront]

ping

[claptrap]

Dear flame can you wrap your mind around this word "GEEK"?

[kezekiel]

Interesting... just more crud to sift through. Researching on the web has always been something of an art, though. You have to play with word order, phrasing, ignore the paid sites, etc. This stuff just adds a layer of junk to it all.

[billbears]

interesting

[VoiceOfBruck]

I don't think this is a big problem. Yes, there's oodles of junk out there, but with experience in searching you can instantly skip right over it.

[N00dleN0gg1n]

Spoofing search engines has been a common activity since the early days of the web. I'm sure the clever folks at Google wil find a way to filter this stuff out.

[flamefront]

Geeky is exactly why I tired to keep this discreet. But search engines have become the mainstay of web activity for everyone looking for anything these days. Nevermind, I just wanted to point out the phenomenon.

[flamefront]

Yes, Google and others can select sites. It is the larger priciple at hand that I find significant - even more sophisticated forms of noise can overload the searches at some point.

Today Google could content check for consistency too. But the scripts that generated the web noise can also become undetectably similar to say, a Sentor's speech or a book review.

[rs79bm]

Those sites are not web noise. They're actually called "read it once and suffer a headache".

[flamefront]

Your search - "read it once and suffer a headache" - did not match any documents.

Suggestions:
• Make sure all words are spelled correctly.
• Try different keywords.
• Try more general keywords.

[KateUTWS]

Headache is RIGHT! What is the tie-in with Amazon?

[flamefront]

I guess it is a way to fool the search engines to show that the web noise is just a relay along the way. Search engines prefer intermediary nodes that are not just dead ends.

It's not jsut amazon that is used though. Sometimes the noisy sites link each other.

[flamefront]

Thanks. I follow what you say and it is a clear sense of the link operation, but the end result of the links is nothing of significance, just more noise. So, I don't think the seeding promotes anything of advantage to anyone.

[N00dleN0gg1n]

They're already more coherent than some Senators.

This arms race has been going on since Alta Vista was a pup, and it's been interesting to watch. Based on past performance, I really don't expect the search sites to lose any time soon.