8/18    I have a buddy who figures he's gonna put togther the next big thing
        in Web 2.0.  He's particularly interested in rating algorithms based
        on user input/voting.  In other words, if users 1 and 2 both rate
        a giving movie highly, does user 1's opinion on a different movie
        mean to user 2?  (Just an example, I don't think he's interested in
        movies.) Can anyone point to any literature on the subject, or would
        anyone like to chat with him about it?  (Or even work with him on it,
        if you have time. I don't.) Thanks. -jrleek
        \- some of this turns on whether you are doing "revealed
           preference" [say what you actually buy from AMAZONG or rent
           from Nflix] or if it is a voting system, in which case one
           must look at the great arrow impossibility theorem. the first
           is more of a pure stat problem, and the second depends on a
           lot of details about incentives, the exact design of the
           system etc. you know, somebody sends me a question like this
           every 2-3yrs for the past 10yrs at least. another approach is
           not no use generic statistical correlation techniques but to
           do a domain-speific analysis/modelling ... which is what i
           believe those pandora people do.