Sam,
In the working group’s discussions,
it was decided to limit the scope of our treatment of rights metadata to
permissions directly related to the repository’s efforts to preserve an archived
object. So we designed the Rights entity in a way that was oriented toward this
purpose: specifically, the granting of a permission by a rights holder to the
repository to take some action: e.g., migration, normalization, duplication, etc.,
documented according to the expression “Agent X grants Right Y to the
repository in regard to Object Z”.
But this does not mean that you can’t
use the Rights entity to document access rights. You could construct
expressions like “Agent X grants to the repository the right to provide
access to Object Z”. But it’s not clear that this is the optimal
way to record rights metadata; other expressions might make more sense.
Brian
From:
Sent: Tuesday, May 09, 2006 11:47
AM
To: [log in to unmask]
Subject: [PIG] PREMIS questions
from LAC
Hello,
Recently, Library and Archives Canada formed a working group
to begin implementing PREMIS and METS around digital assets. Quite a bit
of time has been spent surveying our digital holdings and listing the metadata
requirements for the material. As we work through the PREMIS Data
Dictionary, a number of implementation issues have come up. I was hoping
that the group could share some experiences on these issues.
1) The rights
entity in PREMIS seems to be designed to track curatorial rights: 1) granting
permission to archive an object; 2) authorization to take preservation actions.
It organizes permission statements, agents associated with permissions,
actions, etc, but can it be used to record and manage access rights? What
approaches have other institutions been taking when implementing PREMIS rights
and recording access rights? Has anyone implemented a rights registry?
2) Since we are
looking to house our PREMIS metadata inside the METS structure, I was
wondering, has anyone done any mapping of PREMIS data elements to the METS
schema?
3) Finally,
following the METS/PREMIS implementation thread, has anyone applied PREMIS and
METS to complex, born-digital objects such as websites? Are these schemas
proving to be scalable to objects like websites, which consist of complex
structures, thousands and thousands of files, and multiple formats? Has
this been applied to a domain web harvest?
I realize that all of these PREMIS questions are 'nested'
inside of the larger METS implementation issues, but I was hoping to pick the
groups collective brain on this.
Sam Generoux
Library and Archives
Research and Innovation