- DC2007MeetingMinutes

Identifiers Special Session Minutes, DC-2007, Singapore, 30 August 2007

Meeting Summary

33 people attended
Five minute presentations:

NISO digital identifiers roundtable
ISO Identifiers
PIDs at National Library Board
Identifier principles and compromise
Identifiers at National Library of New Zealand

Discussions on identifier issues and best practices. Key points raised:

Confusion over identification vs. resolution/location
Importance of being able to resolve identifiers
A best practice might be: semantic/intelligence is OK for access ids, but not for abstract work ids

Establishment of Identifiers Community:

DCMI not the right place to do identifiers work, but is a good place to disseminate
New community to be announced in October 2007

Presentations

NISO digital identifiers roundtable

Presented by John Kunze, California Digital Library.

The issues aren’t identifiers themselves but the services supporting identifiers, ie. infrastructure (we've been looking at the whole in the donut so we think it's OK)
Established info registry and 'n2t' (name to thing) resolver
Outreach and education is needed

Roundtable agenda, discussion documents, and final report

ISO Identifiers

Presented by Juha Hakala, National Library of Finland.

ISBN now used for book chapters
ISSN about to offer linking between manefestations
ISNI name identifier proposal
DOIs have a problem of scope since they can applied to anything (including things with ISO identifiers), also concern not very open and may not always remain free
URN being pushed in Europe with resolution services.

PIDs at National Library Board

Presented by Ganesh Yanamandra, National Library Board, Singapore.

Chose DOIs over URNs

Resolvable
Supports rights information

Made application profile and mapped to DC.

PIDs at National Library Board presentation (PDF, 281 KB)

Identifier principles and compromise

Presented by Stu Weibel, OCLC.

The issues conflict, no solution can solve them all, so must compromise
Canonical identifier - not the only id, but choose a best one. Multiple identifiers for the same thing dilute social bibliography, though make it easier to get to a given resource
Branding - Branding is critical to value recognition, provenance, trust. Every URI is a mini-billboard,

http://

WorldCat

Identifier principles presentation (PDF, 79 KB)

Identifiers at National Library of New Zealand

Presented by Douglas Campbell, National Library of New Zealand.

Assigning a new identifier series for each collection item
Resolver to map this identifier to all other identifiers used internally, and present on a single page

Identifiers at National Library of NZ presentation (PDF, 299 KB)

General discussion

Issues raised

Multiple identifier systems

Any convergence possible?
All are purpose-built to support communities (though that's not wrong)

Branding
Protocol for identifiers not always included or apparent
Embeddability and resolvability are critical (browser plugins don't work)
Expectations identifiers are hackable - eg. predictable, truncatable (chopped off end gives information, or appendability)
One identifier may have multiple resolutions
Confusion of location and identity

Can one identifier fulfill both?
Is there any meaningful distinction?

Why not just give location (not an identifier)?

Not canonical, location may change, location may be reused over time
Library community has a conceptual model of use - FRBR (Functional Requirements of Bibliographic Records)

Embed semantic information in identifiers?

No in theory, but in practice people don't use numbers

Best practice suggestions

Opaque identifiers

Good for abstract work
But for service points include semantics

Persistent and dynamic linking

Need to change current cataloguing practices (OpenURL/URN not URL), especially for copied records, this is automatable

Identifiers Community

DCMI not the right place to do identifiers work, but is a good place to disseminate
Support for establishing a new DCMI Identifiers community

> DC2007MeetingMinutes