Lots of interesting stuff during ALA and I'm still processing it. Ronald Murray presented his work on a "FRBR paper tool" which, as someone said after he presented at CC:DA, was either exhilarating, exhausting, or terrifying, or all of the above. What was exciting for me was that it seemed to move us beyond some of my concerns about FRBR. That is, people seemed to be chasing Work, Expression, Manifestation, and Item to the detriment of moving forward toward a response that built a richer as well as collocated response to a query. Murray's networks of bibliographic relationships allow a system to build trees that can use the building blocks. Not only did it seem promising for texts but he also momentarily had a parallel 4-layer context for archives: fonds, series, folders, items (I think that was it). I wonder if there's a similar hierarchy for visual materials and cultural objects which will make sense. A starting point into the web stuff on Murray's work can be found at http://dltj.org/article/frbr-paper-tool-presentation/ (first hit when you google "frbr paper tool" -- the report on a November 2009 presentation at the Library of Congress).
Oh ... and his charts were beautiful. As I was looking at them, I was thinking of hiring a silversmith to turn one into a necklace. Or maybe a knit version? If it was macramé, it would probably work for philodendron.
Later that morning, I went to an OCLC program entitled "Cataloging alchemy: making your data work harder" which included Rich Greene talking about GLIMIR (Global Library Manifestation Identifier). They've focused so far on parallel-record and reproduction identifiers which will help them pull together editions/printings/records for closely-related resources. Rich indicated that they might be able to do some enhancing of the resulting cluster, e.g., access points at the cluster level. I'm dreaming of enhanced subject access because of this, or more contents analysis because one edition has a contents note. Hmm, enhanced RLIN clusters. The basis for GLIMIR building is the new version of the duplicate detection report which OCLC has started using. This will help match and merge simple vendor records with fuller ones, both in batch-loading and already in the database. They plan on working their way through the 195 million records in the OCLC database.
There were other exciting things at ALA Annual, that was just on Monday. I'll work on my ALA report as soon as I can. Meanwhile, it's still a couple hours to Binghamton and then three hours to Alfred. Buses with wireless internet access are pretty nice but it's a little cramped for spreading out your ALA notes.