2. What is a scholarly construct?
• An agreed pattern for describing a particular concept. For example,
– How is object production defined?
– What is a visual depiction?
– What are the different acquisition / provenance scenarios?
– How are object parts represented?
• Independent of any particular data implementations. YCBA <> BM
• Promotes data harmonisation and analysis across different
organisations.
Yale Centre
British Art
(CDWA)
British
Museum
(SPECTRUM)
3. Schema Standards & Customisation
Content: Jenn Riley
Design: Devin Becker
Work funded by the Indiana University Libraries White Professional Development Award
Copyright 2009-2010 Jenn Riley
5. Ontological Knowledge
Representation
• Ontologies describe concepts
and relationships, but you need
to get the right balance.
– Too high level - squeeze the
goodness out of museum data
into artificial boxes (e.g. Dublin
Core)
– Specialisation prevent
agreement on constructs.
Parmenides – the
nature of reality
7. General Aggregators too General
The Institution The Aggregator
o The further the data is from the originator the less assumptions can be made
about it.
o Default to a common set of fields.
o Challenge for retaining knowledge is not to prescribe a common set of fields
but to find a common set of generalisations within a domain to harmonise
different datasets.
o Using aggregator models as a primary publication models prevents the
formation of meaningful scholarly constructs.
The curator / researcher
8. • Real world ontology that matches the richness of museum
records.
• The only purpose built ontology that can adequately
represent a British Museum record.
• The only ontology that allows relevant and practical cross
organisational constructs.
• The only ontology that allows practical collaborative
enrichment beyond the BM record. Open Constructs!
• rosetta stone.pdf
9. E.g. Normalised Acquisition Constructs
Construct 1 - Acquired From
• Bequeathed by
• Donated by
• Exchanged with
• From
• Purchased from
• Transferred from:
• Unclaimed item:
Construct 2 - Received Custody From
• On loan from
Construct 3 - Acquired Through (intermediary)
• Purchased through
• Bequeathed through
• Donated through
• Exchanged through
Construct 4 - Acquisition Motivated By
• In Honour of
• In Memory of
Construct 5 - Found By
• Collected by
• Excavated by
16. • Visualise different data sources against the
CRM.
• Use scholarly constructs as a knowledge base
for ontology mapping.
• Plug-in local vocabularies.
• Manage the relationship and changes
between data producers and multiple
aggregators.
1. As per slide. Pattern, independent – harmonise
I would say construct because these are constructs that don’t just help scholarly analysis but also help engagement.
Example – The British Museum and Yale University. They have original Constables and we have prints that are variations and which have different information recorded in different ways. A construct allows us to bring all the information together and use computer modelling techniques.
We use to many data models and in any event customise them
There is little point in even trying to attempt to create middleware to bring different databases together – it simple isn’t sustainable or maintainable.
Not just the usual vocabularies like materials, techniques etc
But attribution vocabularies that can be specialist for individual institutions