How it works: javascript:

   1. fetches latest 100 tweets by json api
   2. removes urls and @mentions
   3. removes stop words
   4. count frequency of all words
   5. post list to backend

   6. sort words entropy looking up in a frequency database
   7. search highest entropy words at amazon.co.uk
   8. send back data as json

What do you guys think? Ideas?

(Apologies for stealing the color theme and css from here!)