OK, that is good to know. I had talked to somebody about this problem several years ago and I had thought that you could start with some kind of metric space and see which songs were 'close' to each other. The guy I was talking to seemed to view the problem as a kind of Bayes Theorem problem. The example being that if you had some obscure Beatles album then you most likely had 'Abbey Road' but the inverse probably wasn't true and that is where a software service could step in ... to recommend the obscure album as one you would probably like.
One thing that I really like about the service is finding artists that seem to lead to lots of other great songs. It isn't always who I think it is. For example on a lark I put in "New Radicals" because I liked one of their songs and I was so impressed with how well that defined a particular sound. When I put in big names like the Beatles or Rolling Stones I don't think I left those stations defined for very long. I wonder is there is some kind of structure to the what Pandora has calculated that could show which bands or songs seem to be most influential.
One thing that I really like about the service is finding artists that seem to lead to lots of other great songs. It isn't always who I think it is. For example on a lark I put in "New Radicals" because I liked one of their songs and I was so impressed with how well that defined a particular sound. When I put in big names like the Beatles or Rolling Stones I don't think I left those stations defined for very long. I wonder is there is some kind of structure to the what Pandora has calculated that could show which bands or songs seem to be most influential.