Link to USPTO filing dated October 26 (thanks Amit, via kottke)
Reader comment: Peter Rothman says,
I read the Flickr patent this morning and FWIW I don't think Flickr should be able to get a broad patent on "interestingness". There's a very large number of papers in the image processing and collaborative filtering areas that all define various notions of relevance, interestingness, salience, or novelty. A specific innovative technique might be patentable, but not the general idea of computing how interesting an image or media object is to a person or set of people.
These papers are not obscure, there are many of them, and they are well known to experts in the respective fields. I'd be happy to provide a lengthy prior art file to the patent librarian if they can't find it themselves. I am aware of papers that date back more than a decade in this area. I suspect that the Flickr folks are well aware of this actually and are just trying to get the broadest patent they can get.
FYI I am the chief scientist of a company applying image processing and facial recognition software in the field of bank fraud prevention. My views do not necessarily represent those of my employer however. Previously I was the CTO of Live365.com and the Director of Research and Development at MetaTools, two companies you might have heard of at least.