I just realized that Thomson Reuter's impressive service, Web of Science, has now expanded to include a book citation index. Previously, a book would be in the database if it were cited by an article, but the citations of books and articles FROM books was left out. Collecting data from books--with so many publishers, genres, and formats--is no easy task, but the folks at Thomson Reuters figured out a way to do it.
From what I can tell, they have a moving wall of including books--only books published within the past 5 years (or 7 years for social sciences) are included. One reason for this may be to limit the scope of an expansive universe of books. But the only reason I can think of for making it a moving wall (so that books are dropped from the database over time) would be to favor recent book citations. Is this motivated by an assumption that science is cumulative, and only recent publications matter? Here's the description of the methodology: http://wokinfo.com/media/pdf/BKCI-SelectionEssay_web.pdf
My questions is: does anyone have any experience working with the Book Citation Index, and if so, what have your experiences been like? What kinds of books are included, and what books have been left out? Is this tool only valuable for showing who are "hot" authors right now, or can you use the data to analyze patterns over time?