Unit 5 Readings

Witten
2.2
I like Witten's emphasis on presenting an image of stability and continuity to the user, despite the impermanence and fluidity of digital documents. I hadn't previously considered the comparison between different editions of a physical work and different modifications of a digital document. The discussion on authority control is not new information, but I wonder if it is more or less relevant/necessary in the context of digital libraries with sophisticated search functions. The LCSH information is also not new, though it's a good overview of basic cataloging practices. It is interesting to contrast the linear arrangement of books on shelves with the endless rearrangement possible in a digital collection.
5.4-5.7
Good overview of MARC and DC. BibTeX was new to me; I like that it seems compatible with XML. Refer seems, on the surface, a little more comprehensible to me, but maybe I'm just drawn to the alphabetical ordering of categories. The section on TIFF was enlightening; I had no idea the format carried so much metadata. I'd like to explore that capability further. MPEG-7 seems extremely useful for digital libraries; the projected capabilities mentioned are pretty amazing, although the technical stuff was a little over my head.
Automatic extraction of metadata makes me a little nervous, perhaps because I am a firm believer in authority control. Many of the methods discussed seemed iffy, though key phrase assignment/extraction seems useful. With the volume of information out there it's useful to have some automated methods, but there will always be a need for manual assignment of metadata. Phrase hierarchies actually seem pretty interesting, though all the talk about complex algorithms is beyond me.

Gilliland
The table on typology of data standards is a really useful way of thinking about this stuff (e.g. LCSH vs. AACR, MARC fields vs. MARC21). Tables 2 and 3 are useful as well. She makes an important point that metadata for digital objects needs to exist independent of the current storage/retrieval system in order to survive migration.

Weibel
It's encouraging to read the reflections of someone who was involved with Dublin Core from its very first development. I laughed a little when he mentioned that "participants spent an hour of scarce plenary time talking about Type before realizing that the librarians and the computer scientists had been talking about completely different concepts" -- but the important thing is that they produced a functional standard in the end. Weibel questions the possible future effect of "folksonomies" on the metadata landscape, and I am both interested and discomfited by the notion. I liked his anecdote about the China-Mongolia railroad gages.