G

 

Semantics

 

<< Previous    1  [2]  3    Next >>

Unfortunately, Google are in a battle against the finest computer the world has known: the human brain. So it wasn't long before these strategies were reverse engineered and keyword stuffing and link farms made their appearance.

In addition, Google were using some less refined techniques, such as the age of a web site and the length of time a domain name has been registered for and how quickly a web site grows. These criteria are meant to flag sites created specifically to make income from Google's AdSense program, by spotting web sites that apparently arrive overnight, having been created by content generating software for the sole purpose of providing a hollow shell of a web site, upon which to place AdSense ads.

Unfortunately, this is sometimes a blunt weapon and penalizes quite legitimate web sites, that may have merely changed servers or have suddenly published a sudden increase of content, not via some nefarious method, but simply because the webmaster has been burning the proverbial midnight oil.

Think about it: how logical is it to deny the searcher really good fresh, original content, seamlessly matching their requirement, which the webmaster worked on for many months, prior to the launch, simply because the web site is only a few weeks old, the domain name may only be registered for a year or so and because the whole web site -- despite being many months in preparation -- was all loaded in a day or so?

There had to be a better way.

There is. It's here. And it's name is . . . yes, you've guessed it: latent semantic indexing. This is a smarter way of judging the content of web pages and looking at the pages, in the context of the entire web site, to determine a common theme which the web site covers in depth.

This means that a web site with excellent content isn't penalized simply because it hasn't been around for long and doesn't have many links, because it hasn't sought a truckload of artificial ones, and adds fresh content on a very regular basis.

<< Previous    1  [2]  3    Next >>