FW: Some MME questions
Mason, Ingrid
Thanks for letting me share this reply to your email with the group -
you've asked great questions. We'd be glad to take feedback on any or
all of these points.
>>Geo Spatial Data
// In regard to places, I'm assuming ANDS wants them geocoded so that
they can be found using the map interface to RDA. Will institutions be
expected or able to submit points/boxes/polygons to MME, or is MME going
to be doing some geocoding before output as RIF-CS?
Answer: We decided we'd start it the simple way and go with place names
in text as the entry level. I looked at various means of supporting
geo-spatial data. If a contributing organisation would like to submit
alternative values for geo-spatial data then we're open. We're in
discussions with Robina at ANDS about this right now. I'd anticipated:
text, iso19139dcmiBox, and dcmiPoint for the Hedda Morrison collection
here at the Powerhouse as the example.
Text: Beijing, China; Peking, China; Hong Kong
Box: name=Beijing; northlimit=40.3; southlimit=39.6; westlimit=115.8;
eastlimit=116.9
name=Hong Kong; northlimit=22.8 southlimit=21.99 westlimit=113.5
eastlimit=114.6
Point: name=Beijing north=39.9 east=116.4; name=Hong Kong north=22.3
east=114.16
We can certainly provide a space for that kind of data if people wish to
supply it. We're yet to discern what tools ANDS has that might do some
magic for us.
>>Parties Data
// I'm also wondering about party relationships. I'm assuming that the
RIF-CS generated by MME for each collection will include a related
object to a party record for the contributing institution - which
presumably will be a People Australia identifier. There might however be
some question as to whether the relationship should be modelled as
'isManagedBy', 'isOwnedBy' or 'hasCollector' (ownership might not be
completely straightforward). Will this be able to set on a
per-collection basis by contributing institutions?
Answer: Yes, we'll create a party object (of the collecting organisation
and data contributor) and link object that to the collection objects (of
the party). In a quick look I'm at advocating "isManagerOf" as the
attribute for the organisation, in that the Party manages the material
and this value can be set for the party object that relates to all the
collection objects (of the party). As to PeopleAustralia, we'll be
looking at integrating that where we can - see next answer.
>>PeopleAustralia IDs
// I'm also wondering about the People Australia ids themselves. I
presume many if not all of the participating institutions will have PA
entries that have been harvested from CAN. Will the MME project be
working with Basil to put this on a sustainable footing, so that the
entries can be modified/maintained irrespective of what might happen to
CAN itself. (I've just noticed that the NMA has two PA ids -
disambiguation fail!)
It can't be guaranteed that the data in CAN will enter PA (though in an
ideal world it would). So..I'm guessing that there is a question here
about MME provided via MME to PA to ensure ongoing viability. We're in
process on this one too... So thanks for highlighting it for others to
see and as a reminder.
>>People/Organisation Data
// And what about other party relationships? Many of the collections
will have a known collector who is not the collecting institution, so it
would be good to make this explicit with a 'hasCollector' relationship
to a party record. There doesn't seem any way in the current template to
document such relationships and make them available to MME.
Answer: We are interpreting those persons, i.e. the collector or the
generator of that collection material, as people/organisation data. We
aimed to use subject/type/local as a parking spot for this data. In
discussion just now with Lynne we're mooting using identifier/type/nla
to link (using the URI) through to the PeopleAustralia records for the
collector/generator recorded there. Our example is Hedda Morrison, she
was both the generator and collector of the material in the Hedda
Morrison Collection. http://nla.gov.au/nla.party-615689 We're not
looking at getting into the complexity of describing the relationship
that Hedda to that collection using the "hasCollector" or duplicating
party records. If the NLA (that maintains the party record for Hedda)
wishes to provide the link back to the collection, i.e.
"hasAssociationWith" via the party record that's another question and
possibly more useful means of reflecting that relationship? Do you have
other suggestions?
>>Harvesting Metadata Directly
// How are you looking to harvest collection records from institutions
to the MME?
We're in process with that and I'm waving the OAI-PMH flag, the API flag
and you tells us flag right now. I've no idea how many organisations
have OAI-PMH or API capacity, but this seems (to me at least) and piece
of useful infrastructure to have in place. For example the Powerhouse
has an API on the go that could be queried to pull out collection data
regularly. But... another organisation may have OAI-PMH. So perhaps if
you tell us how you'd prefer to work to that end, that could be really
helpful.
>>Python Client Library
// While I was at ANDS I started work on a little Python client library
for generating RIF-CS - it might be of some use to you down the track.
Answer: Sounds useful... More please... :-)
Thanks, more questions or responses anytime.
Ingrid
Ps. For those that want to post a reply and can't please just email me
so we can work out how your Google groups membership works
Ingrid Mason
Museum Metadata Exchange
w. http://www.powerhousemuseum.com/museumexchange/
e. muse...@phm.gov.au
Digital Project Manager
Powerhouse Museum
w. http://www.powerhousemuseum.com
e. ing...@phm.gov.au
p. 02 9217 0432
----------------------------------------------------------------------------------------------
This email and attachments are for the use of the intended recipient(s) only and may contain confidential or legally privileged information or material that is copyright of Powerhouse Museum or a third party. If you have received this email in error, please notify the sender immediately and then delete it. If you are not the intended recipient, you must not use, disclose or distribute this e-mail without the author's prior permission. Any views expressed in this message and attachments are those of the individual sender and the Powerhouse Museum accepts no liability for the content of this message.
-----Original Message-----
From: tim.sh...@gmail.com [mailto:tim.sh...@gmail.com] On Behalf
Of Tim Sherratt
Sent: Monday, 11 October 2010 11:35 AM
To: Mason, Ingrid
Subject: Some MME questions
Ingrid,
Hope I'm not jumping the gun, but I'm keen to get things going at the
NMA and have some questions about the proposed collection description
schema.
In regard to places, I'm assuming ANDS wants them geocoded so that they
can be found using the map interface to RDA. Will institutions be
expected or able to submit points/boxes/polygons to MME, or is MME going
to be doing some geocoding before output as RIF-CS?
I'm also wondering about party relationships. I'm assuming that the
RIF-CS generated by MME for each collection will include a related
object to a party record for the contributing institution - which
presumably will be a People Australia identifier. There might however be
some question as to whether the relationship should be modelled as
'isManagedBy', 'isOwnedBy' or 'hasCollector' (ownership might not be
completely straightforward). Will this be able to set on a
per-collection basis by contributing institutions?
I'm also wondering about the People Australia ids themselves. I presume
many if not all of the participating institutions will have PA entries
that have been harvested from CAN. Will the MME project be working with
Basil to put this on a sustainable footing, so that the entries can be
modified/maintained irrespective of what might happen to CAN itself.
(I've just noticed that the NMA has two PA ids - disambiguation fail!)
And what about other party relationships? Many of the collections will
have a known collector who is not the collecting institution, so it
would be good to make this explicit with a 'hasCollector' relationship
to a party record. There doesn't seem any way in the current template to
document such relationships and make them available to MME.
How are you looking to harvest collection records from institutions to
the MME?
Lots of assuming and wondering there - but any details gratefully
received. And of course I'm more than willing to contribute ideas and
suggestions from my experience at both ends of the process. Speaking of
which, while I was at ANDS I started work on a little Python client
library for generating RIF-CS - it might be of some use to you down the
track.
Cheers, Tim
--
Tim Sherratt (t...@discontents.com.au)
National Museum of Australia
Adjunct Associate-Professor, Digital Design + Media Arts Research
Cluster, Faculty of Arts and Design, University of Canberra
Words - http://www.discontents.com.au
Experiments - http://wraggelabs.com
@wragge on Twitter