My impression is that all personal data is kept client side and all data sent to them is anonymous, though we all recall how well that tends to work in practice.
whoa this is pretty good - the idea is actually quite novel. Since most people consume content from select websites, they are bound to go to those sites more and essentially that is the internet to them. So why not help them try to find the most relevant content from that very site. I will give it a shot.
So SiteSimon will never turn over my information to law enforcement or use it to target advertisements? That seems highly unlikely to me.