April 1st Fun with PROV, Dublin Core and some Others

2014, three months passed, and it looks like I have not started the blog yet. I guess a simple and short writing with a figure and a table could be a good start.

The figure is a mapping summary based on two W3C documents: Dublin Core to PROV Mapping (W3C Working Group Note 30 April 2013) and the direct mappings of Dublin Core to PROV mapping file.(URL: http://www.w3.org/ns/prov-dc-directmappings.ttl).

The table below is a brief view on W3C's task on Provenance Vocabulary Mappings (using SKOS mapping). This time, I only choose OPM, PREMIS, and Dubin Core, and use the direct mappings of Dublin Core to PROV mapping (owl:equivalentClass and rdfs:subPropertyOf) to relate PROV to DC.

Terms from Reference Model (OPM)	Provenance Models: Terms and Mappings
	PREMIS		Dublin Core		PROV
	Mapping	Model Term	Mapping	Model Term	Mapping(DC2PROV)	Model Term
opm:Process⃰	skos:relatedMatch	premis:Event	skos:relatedMatch	dcmitype:Event
opm:Artifact⃰			skos:narrowMatch	dct:BibliographicResource, dct:PhysicalResource	rdfs:subPropertyOf	prov:Entity
	skos:narrowMatch	premis:Object		dct:Collection, dct:MethodOfAccrual
opm:Agent⃰	skos:narrowMatch	premis:Agent	skos:exactMatch	dct:Agent	owl:equivalentClass	prov:Agent
opm:Account			skos:narrowMatch	dct:ProvenanceStatement	rdfs:subPropertyOf	prov:Collection
opm:used	skos:relatedMatch	premis:linkingObjectIdentifier⃰	skos:relatedMatch, skos:relatedMatch	dcterms:requires, dcterms:isRequiredBy
opm:wasDerivedFrom	skos:broadMatch	premis:relatedObjectIdentification	skos:narrowMatch	dct:source	rdfs:subPropertyOf	prov:wasDerivedFrom
				dct:replaces, dct:hasPart, dct:references
opm:wasGeneratedBy	skos:broadMatch	premis:relatedEventIdentification⃰	skos:broadMatch	dct:source	rdfs:subPropertyOf	prov:wasDerivedFrom
opm:wasControlledBy	skos:relatedMatch	premis:linkingAgentIdentifier⃰	skos:relatedMatch	dct:contributor	rdfs:subPropertyOf
opm:wasTriggeredBy			skos:broadMatch	dct:source	rdfs:subPropertyOf	prov:wasDerivedFrom

It's interesting to see the difference of the event concept in these vocabularies. For me, the concept of event is supposed to be the first class of the provenance. The changing character of temporal and spatial components distinguishes provenance information from pure metadata description. However, in PROV we can only have the prov:InstantaneousEvent that is a non first-class notion, and this is probably because PROV has its original proto reference, OPM, that uses opm:Process as one notion of three opmo:Node class.
--------------------------------------------------------------------------------------------------