Regarding question #2 below, I would tend to agree with Priscilla, and
further suggest that the restriction on harvesting the metadata might be
better described in the Rights schema as a "restriction" on the "act" of
"disseminating metadata" for instance. That could be another way to
approach the issue.
Nancy
Nancy J. Hoebelheinrich
Metadata Coordinator
Digital Services Group
Stanford University Libraries / Academic InfoResources
Stanford, CA 94305-8408
[log in to unmask]
voice: 650-725-6843
fax: 650-725-0547
-----Original Message-----
From: PREMIS Implementors Group Forum [mailto:[log in to unmask]] On Behalf Of
Priscilla Caplan
Sent: Friday, January 27, 2006 8:20 AM
To: [log in to unmask]
Subject: Re: [PIG] Mapping Persistent Identifiers in PREMIS
This is a reply to Dana Schlesinger's note copied below. I sent an
answer to Dana but then realized I should also send it to this list.
1. How to handle multiple identifiers
This shouldn't be a problem -- objectIdentifier is repeatable. I
believe we had the expectation that if there were multiple identifiers,
they would have different objectIdentifierType values, but this is not
stated in the Data Dictionary.
2. Is it correct to map the flag "suppress OAI harvesting" to the
"inhibitors" element?
No, I don't think so. An inhibitor is supposed to be "a characteristic
of the object," that is, a property of the object itself. Suppressing
harvesting of the metadata for the object in the repository application
WOULD inhibit access, but is not a property of the object. The
inhibitors element was intended to alert the repository itself that it
may not be able to migrate or emulate the object unless the inhibitor(s)
could be overcome.
I'm willing to be corrected -- does anyone else have an opinion?
p
> Dear P.I.G members,
>
>
>
> We intend to develop a module to handle the creation and management of
> Persistent Identifiers. Our thought was to use the PREMIS Object Entity in
> order to preserve the object identifier(s). Following is an example of the
> metadata:
>
>
>
> <object
> xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
> xmlns:pre="http://www.loc.gov/standards/premis/Object-v1-0.xsd">
>
> <pre:objectIdentifier>
>
>
> <pre:objectIdentifierType>urn:nbn:de:hbz</pre:objectIdentifierType>
>
> <pre:objectIdentifierValue>
> urn:nbn:de:hbz:1234-123</pre:objectIdentifierValue>
>
> </pre:objectIdentifier>
>
> <pre:objectIdentifier>
>
> <pre:objectIdentifierType>doi</pre:objectIdentifierType>
>
> <pre:objectIdentifierValue>
> doi:10.1392/BC1</pre:objectIdentifierValue>
>
> </pre:objectIdentifier>
>
> <pre:storage>
>
> <pre:contentLocation>
>
> <pre:contentLocationType>URL</pre:contentLocationType>
>
>
>
<pre:contentLocationValue>http://il-dtldev02:8881/R/900272?func=results-jump
> -full&set_entry=000001</pre:contentLocationValue>
>
> </pre:contentLocation>
>
> </pre:storage>
>
> <pre:objectCharacteristics>
>
> <pre:objectCharacteristics>Suppress OAI
> Harvesting</pre:objectCharacteristics>
>
> </pre:objectCharacteristics>
>
> </object>
>
>
>
> In this example, the object has two persistent identifiers (each belongs
to
> a different identifier system, which is unusual case but possible), a deep
> linking URL address and a note that the object identifiers are suppressed
> from being harvested by OAI. The "Identifier Type" used here is simply the
> identifier namespace.
>
>
>
> At this stage we do not handle the object relations and/or versions on the
> PREMIS metadata, though it can be mapped during export.
>
>
>
> I'd appreciate your feedback on the above mapping, especially on how to
> handle multiple identifiers and how to "flag" that a certain identifier is
> not available for harvesting (is the "objectCharacteristics" element
correct
> place for that?)
>
>
>
> Thanks and best regards,
>
> Dana
>
>
>
> -----------------------------
>
> Dana Schlesinger
> Product Analyst
> DigiTool Team - Ex-Libris Israel
> [log in to unmask]
|