This is one of InLOC's direct properties.
| name | id |
|---|---|
| URI | http://purl.org/net/inloc/id |
| same as | dc:identifier |
| label | Identifier |
| domain | rdfs:Resource |
| range | rdfs:Literal |
| property of | LOCdefinition, LOCstructure, LOCassociation, subject, scheme, object |
| note | ids must be unique, and should be as persistent as possible |
In every case, InLOC recommends, but does not require, an HTTP URI for the form of the identifier. For conciseness and clarity for human checking, abbreviations may be introduced using well-known and recognised approaches. For instance: prefixes in Turtle; CURIEs for RDFa.
Within any particular InLOC document, file, package, etc. that is communicated, any one identifier must be given in exactly the same form throughout, with no resolution or expansion necessary to determine equality.
Any LOCstructure or LOCdefinition must only be referred to elsewhere in an InLOC document using its id. If it is desired to represent other identifiers for an element, these must be represented as extraID elements.
Fixed vocabularies for some scheme.ids
InLOC specifies fixed vocabularies for LOCassociation scheme.ids in two cases: for
- relationships (where the LOCassociation type is always "LOCrel")
- compound properties where the LOCassociation type is "by" (but not for other compound properties).
Extra constraints for optimised RDF
For the optimised RDF binding, the range of id must be a URI (xsd:anyURI) in all instances except as a property of object.