More FRBR Workshop, identifiers, etc.

Although the content of a conference is its justification, meeting people for the first time is always one of a conference’s pleasures.  The FRBR Workshop had a very nice dinner in the OCLC Atrium and, among the other fascinating people at our table, was the Inquiring Librarian (Jenn  Riley of Indiana U.) who writes a blog worth looking at.

Diane Vizine-Goetz gave an excellent presentation about the use of subjects in fiction, although I suppose I may be prejudiced as we are working on some projects together.

Ketil Albertsen gave a very thoughtful talk about identifiers.  For the Paradigma project, they needed to develop identifiers for harvested objects.  Anyone who has worked with identifiers will recognize that identifiers aren’t as simple as they seem.  Ketil’s presentation lists more than 30 issues they considered while designing their identifiers, along with pros and cons, and an estimate of how sure they are of their decision.  Here are the first few rules (slightly paraphrased):

• ID value carries no information about identified object
• ID values should use a restricted character/symbol set
• Check digits are include in ID display
• Check digits are not stored internally
• IDs have a fixed length

• An external readable format is rigidly defined
• Primary resolution service optionally identified in external ID format
• Binary IDs are displayed as decimal digits
• One ID scheme covers all object classes & information types
• An ID may be assigned to both physical and abstract objects
• One ID scheme handles both static and dynamic resources
• An object is either exactly specified or explicitly defined as an aggregate
• An ID may identify a rule to determine object components
• An ID identifies a unique object
• Identified objects are not assumed to be atomic
• Each component of a composite must be identified
• IDs are independent of the location of the object
• Location IDs are independent of the stored object
• References to an object do not require contents interpretation

Here's one that might be controversial:

• One ID refers to several object aspects: value, metadata, converted versions...

Anyone interested in identifiers should look at Ketil's paper which has much more information about each of these and additional points.

--Th

Comments

[Simon Pockley]

I've looked at Ketil's ppt but I'm struggling to understand current views on the relationship between identifiers up and down the various layers of FRBR.

What is your understanding of how an item level id would relate to a work level id?