Data Warehouses and the Semantic Web - Data Warehouse Systems: Design and Implementation - page 548

level member. For instance, it is stated that the member ns1:J belongs
to level ex:GovernmentO ceRegion and that such member aggregates over
the member ns2:921 in level ex:Country . Note that the dimension instances
are treated analogously to QB, using the SKOS vocabulary.
ThedatacubeofFig. 14.5 can be represented using QB4OLAP as follows:
ex:Geography a qb:DimensionProperty .
ex:Time a qb:DimensionProperty .
ex:Year a qb4o:LevelProperty ; skos:closeMatch db:Year ; qb4o:inDimension ex:Time .
ex:Household a qb:MeasureProperty .
ex:HouseholdCS a qb:DataStructureDefinition ;
qb:component [qb4o:level ex:UnitaryAuthority] ;
qb:component [qb4o:level ex:Year] ;
qb:component [qb:measure ex:Household ; qb4o:hasAggregateFunction qb4o:sum] .
There are some relevant differences with the QB data cube shown in
the previous section. The data cube schema is declared as an instance
of the class qb:DataStructureDefinition . The level property qb4o:level is
used to specify the levels in the cube schema. For example, the triple
ex:HouseholdCS qb:component [qb4o:level ex:Year] tells that ex:Year is a
level of the cube. Analogously, the triple ex:Household a qb:MeasureProperty
tells that ex:Household is a measure in the cube, like in QB. In addition,
the aggregate function corresponding to this measure is defined using the
property qb4o:hasAggregateFunction .
An instance of the cube is represented in QB4OLAP as follows:
ex:dataset-hh a qb:DataSet ; qb:structure ex:HouseholdCS ;
rdfs:label '' HouseholdinUK '' @en .
ex:o1 a qb:Observation ; qb:dataSet ex:dataset-hh ;
ex:UnitaryAuthority ns0:00mc ; ex:Year db:2007 ; ex:Household 58 .
The data set ex:dataset-hh is defined as an instance of the class qb:DataSet ,
linked to the data cube using the property qb:structure .Further,anobserva-
tion ex:o1 is defined as an instance of the class qb:Observation .Theinstance
represents the cell in the cube with values ns0:00mc for UA , 2007 for Year ,
and 58 for the measure Household . Note that observations are defined using
QB classes and properties. Therefore, the observations in an existing cube
expressed in QB do not need to be changed in order to express the cube in
QB4OLAP.
OLAP Operations over QB4OLAP
We now show a possible implementation of some of the OLAP operations
over a data cube defined using the QB4OLAP vocabulary. The operations
arewritteninSPARQL.