Can subjects be relevancy ranked?

Over at LibraryThing Time Spallding wondered Can subjects be relevancy ranked?
Some ideas he considered:

Treating subjects as links, and running some sort of “page-rank” style connection algorithm against them. Maybe this would bring out coincidences that simple statistics misses.
Using other library data, such as LCC and Dewey. This would be reminiscent of how I made LibraryThing’s LCSH/LCC/Dewey recommendations.
Doing statistics on other fields, such as the title. So, for example, there’s probably a statistical correlation between “Man-woman relationships” and books with “dating,” “men and women” and “proposal” in the title.