JCDL2006 Monday Brain Dump

JCDL is ranking high one of the most useful (and when not useful one of the most interesting) meetings I've attended. This is my first time here, and although I'd heard about it before I did not realize what a treat it would be to attend. I'm a practice-oriented guy with an eye towards useful (implementable) theoretical stuff, and so far JCDL has hit that sweet-spot very nicely. The down side is this -- here at the start of the second day of the meeting I just need to dump my brain into this and a few other postings and hope that I'll have a chance to get back to these thoughts some day.

Potentially Useful Stuff

ClaimID: A system for Personal Identity Management (Poster #14) — A meta-layer on top of various identity systems that allows you to pull all of your web pages and identity tokens together in one place. A way for you to claim that web pages are written by you or are about you or formally disavow web site content or claim that you hold certain identity tokens. It is focused on the broad, end-user market so leverages something called "OpenID" (new to me); it looks like it could, though, also be a place to publish a PGP key and/or serve as a crude Shib IdP. Fred Stutzman (← that is a ClaimID URL) patiently walked me through it -- very interesting stuff.

Metadata Data Dictionary for Analog Sound Recordings (Poster #17) — OhioLINK is getting more into sound recordings with an acquisition of about 30,000 on the drawing board, so I found this poster a useful jumpstart in learning about the various data elements that the authors had found for capturing descriptive, administrative and preservation metadata for analog recordings. (It focused on 78-rpm and LP phonograph artifacts, but it is still a very useful overview.)

Pathways Core: A Data Model for Cross-Repository Services (Poster #6) and Augmenting Interoperability Acrosss Scholarly Repositories (Panel Discussion) — This is truly a placeholder because I'm still digesting what this all means. In short, the former proposes a data model for a package of identifier content that can serve as a pointing surrogate to a content element in a repository. No, this isn't simply an identifier; rather it is a package of the provider's identity and provider's object identifier, a statement (contract?) from the provider about the persistance of the object in its repository, semantic information, and pointers to other identifier packages that make up the whole. Confusing? Yeah. Like I said -- this is a placeholder for a hopeful future posting. So the panel talked about how these identifier packages/surrogates can be used to promote interoperability between repositories. Not really in an OKI Repository OSID kid of way -- rather a way one repository can hold a placeholder surrogate of a content that is actaully in another repository (e.g. "Overlay" journals). Herbert's slides are online; hopefully more will follow.

Named Entities 1 — a group of papers that had some very interesting ideas about how to disambiguate, collate, and dedupe personal names, place names, and the like. Good stuff here -- thanks to Ed Summers for IRC back-channeling me out of where I was to hear this session.

I've got stuff to say about the Open Content Alliance part of the Monday opening session and more stuff on the Pathways Core and Named Entities. But, like I said, this'll have to serve as a placeholder for hopefully future time to write about it.

The text was modified to update a link from http://eprints.rclis.org/archive/00001001/ to http://eprints.rclis.org/handle/10760/4601 on December 31st, 2010.