Advanced search engine optimization demands that the SEO go above and beyond the basic optimization/ranking process and look deeper into the methods the search engines use to understand the structure of the web. One of these methods is concept discovery through the use of putting together related terms into document concepts and eventually a full-fledged categorization methodology.Luckily, if the search engines can do it with their index, so can we, albeit to a lesser extent. Therefore, I've decided to begin work on our next tool at SEOmoz - one which will use a variation on a time-tested model for the discovery of related terms and the analysis of how on-topic a page is.
The research and legwork involved in this project involves a lot of help through the community forums at SEW. Specifically two threads on the subject: An on-topic analysis is already built and has been used by orion of SEW, but we hope to make this tool, as our others, free and available to the community as a whole. Wish us luck!
I think the related terms thing does work, but the 20% unique searches that Google gets daily has turned this related terminology word into the long tail keywords and this a very big portion of traffic now comes from the long tails keywords...
Nice post. For such thing I'm using Russian software called "KeyCollector"...
This is a great consideration which is my main concern right now. I want to come up with as many terms related to my page as possible so that queries on these terms will return my page. I want to target as many queries as possible. Including ones with missellings. But the real question is how do you know if you have too many keywords involved in your text and are getting penalized? How do you know if the SE deems some of your keywords to be NOT on topic because the SE does not know how some of the words relate even though the average user WOULD see how they relate?