Image via WikipediaTwitter is getting ready to start testing the markup of the messages. I can't see this becoming wide-spread unless there are apps to make this easy for the user. They do offer some suggestions on types of data to markup, for books it is the 10 or 13 digit ISBN and the year of publication.
This might make for richer data mining for the Library of Congress's archive of Twitter.BTW the my library is on Twitter. It seems to be working well for us.