I have been meaning to write for a while to suggest that Bookmarks develop some kind of algorithm for recommending books. Pandora has an algorithm for recommending music and Netflix has one for recommending movies, but in the book world, most recommendations are laughable. (Amazon.com suggests other books by the same author, or just something that's generally popular.) My theory used to be that books are so complex that developing an algorithm for recommending them can't be done. But now I see that Goodreads.com has developed such a tool. Apparently, it will learn from a user's reviews and wish lists. I was trying it out, and it's really addictive. I'd be interested to know what other readers think.
One thing the tool doesn't do is cross genres, since it lists books according to categories--such as Contemporary Fiction, Fantasy, and Mystery. I think the most exciting aspect of books is to find connections across genres. For instance, all those fans of A Game of Thrones might like Dorothy Dunnett's Lymond or Niccolo series. They have huge casts, complicated plotting, plenty of action, and some mystical elements. Developing such an algorithm would be an interesting idea for an article: if you liked "x," try "y," which is in a different genre.
It will be...