Print

Print


 Hi Elizabeth,
 
   
 
  Thanks Kai for your contribution in the discussion!


We
 discussed your question today at the PREMIS EC and we want to ask you 
some extra questions to be able to answer. The collection  is connected 
to one accession record and the accession record is the IE. Have
 we understood it correctly with that assumption? Could you give us some
 more information on the type of collection and if its possible to split
  it up into more IE's? 
There have been a discussion regarding large image sequences and its 
events at the PREMIS Implementators maillist PIG. Perhaps that could be 
to some aid? You can sign up to that list using the instruction  here
http://www.loc.gov/standards/premis/, and look into the archive for that
mail thread.
 
   
 
  We look forward to hear from you and give help!
 
   
 
  Best regards,
 
  Karin Bredenberg, member of PREMIS EC
 





Hälsningar/Regards
Karin Bredenberg
 
****************************************************************
Karin Bredenberg
Teknisk rådgivare, metadata /Technical advisor, metadata
Riksarkivet/Swedish National Archives
Avdelningen för offentlig informationshantering/Public Information
Management
P.O. Box 12541, SE-102 29 Stockholm, Sweden
Tel + 46(0)10-476 71 23
SMS + 46(0)70-349 74 55
riksarkivet.se


>>> Kai Naumann <[log in to unmask]> 17-03-02 16:45 >>>
Hi All,
 
we are considering the same problem, but don’t work on PREMIS
implementation and from the definition of whether an object is an IE or
not. But anyway, if this advances science …
 
We have the following ideas:
 
Events are recorded as attributes of an object. An object can be a whole
collection, a smaller batch, a distinct cataloguing unit, or a single
file. 
They can also be recorded as a summary attribute of an object that
regards several sub-objects (e.g. checksum verification result on an
accession of 3000 files attached to a collection object).
Events can be recorded as simple success message or documented success
message.
Events can be recorded in a human-readable or both human- and
machine-readable way.
 
Best regards,
Kai
___
Dr. Kai Naumann | Landesarchiv Baden-Württemberg | Staatsarchiv
Ludwigsburg | Ref. 53
Arsenalplatz 3 | 71638 Ludwigsburg | Fax 07141 64854 6311 | E-Mail
[log in to unmask] | Tel. 07141 64854 6331 
 
 
Von: PREMIS Implementors Group Forum [mailto:[log in to unmask]] Im
Auftrag von Peter McKinney
Gesendet: Mittwoch, 1. März 2017 22:50
An: [log in to unmask]
Betreff: [PIG] FW: [digital-curation] recording many PREMIS events


 
Hi All,
 
I’m forwarding this email below to the PIG list to see if anyone has
comments for Elizabeth,
 
Best from wellington,
Pete
 
From: [log in to unmask]
[mailto:[log in to unmask]] On Behalf Of Elizabeth
England
Sent: Thursday, 2 March 2017 5:44 a.m.
To: Digital Curation
Subject: [digital-curation] recording many PREMIS events


I'm working on a many terabyte collection that has to be processed in
smaller batches because of its size. It all links back to one accession
record (in ArchivesSpace); in PREMIS speak there are many objects but
one intellectual entity. The best solution I've come up with for
recording PREMIS-like events is creating multiple message digest
calculations, etc. as I go, and indicating what portion of the
collection each event refers to using free text fields in the
ArchivesSpace record.




I'm wondering how others have handled recording PREMIS-like events for
large collections that are processed iteratively, or more generally,
what are you considering the "intellectual entity" you create
PREMIS-like events for?




Thank you!




Elizabeth




Elizabeth England
National Digital Stewardship Resident
The Sheridan Libraries
Johns Hopkins University
3400 N Charles St
Baltimore, MD 21218
410-516-8787
[log in to unmask]


-- 
You received this message because you are subscribed to the Google
Groups "Digital Curation" group.
To post to this group, send email to [log in to unmask]
Visit this group at https://groups.google.com/group/digital-curation.
For more options, visit https://groups.google.com/d/optout.