Well, this is interesting… some thoughts on a new tool in the google webmaster’s toolbox…. Just goes to show that mining content doesn’t always tell you relevance. Sometimes controlled vocabularies (subject headings) and taxonomies (categories) are a GOOD thing… Sigh…

Here is what Google says the keyword significance list is for:
Below are the most common keywords Google found when crawling your site. These should reflect the subject matter of your site.
Google is basically looking at your website, sorting its content by the number of times each word was used, and revealing to you what keywords you use the most. It’s similar to a tag cloud, but instead of being all crazy looking like a cloud it just nicely organized.

…but then….

Using this site as an example I found that the sixth most used word was “Post”. This website isn’t about a post or posts. Doing some quick on site research revealed, I often referenced other posts on the site with terms like next post, previous post, or popular post. Eliminating the unnecessary uses of the word will improve the way Google sees my site. Keyword density for more important words will now also rise up more. Now I just need to delete all those articles I wrote about on how to paint a white fence post.

So, you should just use words in your posts that you want to reflect on the relevance of your website to a particular subject in google, am I reading that right? So, if I want this site to rank high for “semantic web” I should just type the term semantic web a million times and that would do it? Sounds like another opportunity for spam hits in google…