StoryCode is a database of story characteristics used to recommend similar books you might like. Readers rate a book on a number of points including genre & characters, plot, setting, atmosphere and personal views. Each trait includes a left/right slider allowing you to record your own impression. Then, the engine add your results to the collective and recommend a list of similar books that have similar values on those traits. Or, if you are shy about rating the book, you can simply get a recommendation by searching for a book you like and clicking the recommendations link.
This is a cool idea. I’ve seen other collaborative filtering engines that do similar things, but this seems more deterministic. In the matching up, anyway – the ratings in each characteristic are all subjective, natch. And there are a buttload of sliders to rate the book theoretically resulting in a better match. One of the random recommendations on their home page said:
[Iain Banks’]The Algebraist shows an 83% StoryCode Match with [Frank Herbert’s] Dune.
Not having read the Banks book, I’m not sure how accurate that is. But, I guess like any database, the quality of the results all depends on the information that gets put there – garbage in, garbage out.