mattrilla - 1:26 am on Sep 21, 2012 (gmt 0)
Thanks Tedster - I didn't know that!
The analysis is based on traffic from the last 3 months. I compare sentences and phrases larger than 20 characters across the whole site. It's not a perfect process, but I created it to be quick.
It had occurred to me that Google could be using "semantic analysis" of sentences - which basically means if a sentence has the same meaning, then statistically they are the same. Semantic checking would mean the end of all sites that have multiple pages targeting similar keywords. Unfortunately, I didn't have time to write that =)
I hope Panda isn't using that!