From STLab
In order to represent the statistic data on properties used in the analyzed datasets we use four different tables. Notice that the data needed to fill the tables have been already retrieved in most cases. We only need to present the data in separate table for analysis purposes.
Furthermore, the same data has to be available in RDF form so that we can query them. In order to represent the such data in RDF you can use the ontology available at http://ontologydesignpatterns.org/ont/lod-analysis-properties.owl (prefix prop:)
Name of the column
| Explanation
| Associated semantic entity
|
Property
| The property under analysis
| rdf:Property
|
Domain
| The property domain asserted in the vocabulary.
| prop:DefinedUniverse, prop:hasDomain
|
Range
| The property range asserted in the vocabulary
| prop:DefinedUniverse, prop:hasRange
|
Table 1: Properties as they are defined in their vocabulary (with domain and range)
Name of the column
| Explanation
| Associated semantic entity
|
Property
| The property under analysis
| rdf:Property
|
# triples
| The number of triples that use the property
| prop:NumberOfTriplesInDataset, prop:isDatasetNumberOfTriplesOf, path:inDataset, prop:numberOfTriples
|
Observed Universe
| The actual universes (couples of subject type and object type) used in the triples (they can be more than one) of a certain dataset.
| prop:hasObservedUniverseInDataset, prop:ObservedUniverseInDataset, path:inDataset, prop:aboutObservedUniverse, prop:ObservedUniverse, prop:hasSubjectType (domain prop:ObservedUniverse), prop:hasObjectType (domain prop:ObservedUniverse)
|
% of subject type
| The percentage of usage of the subject type of a certain universe
| prop:hasSubjectPercentage
|
% of object type
| The percentage of usage of the object type of a certain universe
| prop:hasObjectPercentage
|
Table 2: Properties, frequency of use, universes and their percentages.
Name of the column
| Explanation
| Associated semantic entity
|
Property
| The property under analysis
| rdf:Property
|
Maps to
| The top level property to which the property can be mapped.
| prop:mapsToTopLevelProperty
|
Type of mapping
| The type of relation for describing the mapping
| owl:equivalentProperty, rdfs:subPropertyOf
|
Table 3: Properties aligned to top-level properties defined in general Content ODPs.
Name of the column
| Explanation
| Associated semantic entity
|
Property
| The property under analysis
| rdf:Property
|
Universe
| The universe (subject type object type) as they are used in the triples for the property (only one for each row).
| prop:ObservedUniverse
|
Maps to
| The top level classes to which the properties is aligned to.
| prop:TopLevelUniverse, prop:TopLevelClass, prop:mapsToTopLevelUniverse
|
Type of mapping
| The type of relation for describing the mapping
| owl:equivalentClass, rdfs:subClassOf
|
Table 4: Property universes aligned to top-level classes defined in general Content ODPs. Notice that the same property used with n different couples (subject type, object type) has to be repeated n times in the table, because we have to produce n different alignments for its universes.