James, as the previous responses to your posts imply, I think you're asking a question which has many layers of answers. I agree with a lot that's already been said, but thought I'd add on a bit.
Is OAI-PMH here to stay? I'm not sure that's the key question to worry about. OAI-PMH is simply the bridge your data crosses. The bridge isn't really all that important. It can be replaced with other data transfer technologies once they come along and mature.
However, the data you walk across that bridge is what'll determine whether it was all worth the effort. If you'd like to aggregate metadata from a variety of sources and offer a meaningful search across that aggregation, your data content will be your key concern. (Just to clarify terms, I use data content to denote whatever goes into a database field or in between the open/close XML tag). Unfortunately, that's also the least tractable issue in the entire equation if you're dealing with pre-existing metadata. Even if you agree on the same data structure (like DC, CDWA Lite or LIDO) and all the mappings are perfect, the data content will still determine what'll get retrieved.
In summary: I think you need to think about how to analyze the data your taking in, and how to strategically mitigate some of the inevitable data content issues so searching is satisfactory. Your data transfer protocol doesn't have much impact on this issue.
If you do decide to go down the OAI-PMH route, you might have some use for the tools we're offering as part of the Museum Data Exchange project:
COBOAT (http://bit.ly/95vdMO) is a free conversion tool which extracts metadata out of a collections management system, and transforms it into CDWA Lite. (The tool can also be tought to transform to other outputs - created by CogApp.)
OAICatMuseum (http://bit.ly/bDpqSP) is an open source data content provider which speaks DC and CDWA Lite.
We've used these tools in the Museum Data Exchange to harvest data from 9 museums - in essence, the process you'll go through. All of this is written up in a short version here http://bit.ly/9b13yL, and in all its gory details here http://bit.ly/aU3nkP. These reports also include some interesting findings about interoperability of the data content from these 9 museums you might want to look at regardless of whether you use OAI-PMH or not.
Cheers,
Günter Waibel
OCLC Research
voice: +1-650-287-2144
Günter blogs at ... http://www.hangingtogether.org
Follow me on Twitter ... http://twitter.com/GuWa
-----Original Message-----
From: Museums Computer Group [mailto:[log in to unmask]] On Behalf Of James Donohue
Sent: Wednesday, August 04, 2010 8:54 AM
To: [log in to unmask]
Subject: OAI-PMH and the future
Hi all
The BFI is working with UK regional film archives as part of the Screen Heritage UK project to create a 'Union Search' application that will allow the public to see details of all the archives' holdings via a single service. The system is designed to aggregate data from a variety of different sources and map it onto an agreed minimum metadata standard for film archives, which is then presented to the public via a web-based search interface.
My question is not about the front-end though: we are currently considering the best ways to extract source archive metadata, with eyes both on what methods can be supported by the present partners, and what methods are most likely to enable other film and video archives to participate in the near future. It is our intention to offer a simple CSV-based submission method for those archives with no accessible database online. However we would also like to offer a more sophisticated 'pull'-based alternative, for which OAI-PMH has been suggested.
Based on your insights, will OAI-PMH remain the standard for harvesting museum and archive metadata for the forseeable future? Clearly the Protocol has had a moderate level of uptake since 2002, but do you see the number of services supporting it increasing or waning? Do you think the technical burden of implementing the standard (at both ends) inhibits adoption? And are you aware of any other standards or harvesting methodologies that you think might become gradually supplant OAI-PMH in the coming years, or do you think it is here to stay?
Apologies for descending on your list and throwing in such a complex question - from reading the archives I could see the right expertise is here but I couldn't find an answer.
Thanks for any advice you may be able to offer!
Regards,
James Donohue
Archive Online Technical Manager
British Film Institute
****************************************************************
For mcg information visit the mcg website at
http://museumscomputergroup.org.uk/
To manage your subscription to this email list visit
http://museumscomputergroup.org.uk/email-list/
****************************************************************
****************************************************************
For mcg information visit the mcg website at
http://museumscomputergroup.org.uk/
To manage your subscription to this email list visit
http://museumscomputergroup.org.uk/email-list/
****************************************************************
|