Database Reference
In-Depth Information
A set of measure values and attributes is associated with an observation
using the properties
qb:MeasureProperty
and
qb:AttributeProperty
in the DSD,
respectively. An instance of the above cube is defined as follows:
ex:dataset-hh a qb:DataSet ; rdfs:label
''
HouseholdinUK
''
@en ;
qb:structure ex:HouseholdCS .
ex:o1 a qb:Observation ; qb:dataSet ex:dataset-hh ; ex:Geography ns0:00mc ;
ex:Time
<
http://dbpedia.org/resource/2007
>
;ex:Household58;
sdmx-attribute:unitMeasure
<
http://dbpedia.org/resource/Thousand
>
.
Here,wedefineadataset
ex:dataset-hh
, whose structure is
ex:HouseholdCS
.
The observation
ex:o1
belongs to this data set. Values for the dimensions are
also given. Note that the year 2007 is represented using an IRI.
QB provides the
qb:concept
property, which links components to the
concept they represent. The latter are modeled using the
skos:Concept
class
defined in the SKOS
13
(Simple Knowledge Organization System) vocabulary.
Multidimensional Data Representation in QB
QB does not provide a mechanism to represent a multidimensional schema,
but it allows us to represent hierarchical relationships between members
of dimension levels using the SKOS vocabulary. A
SKOS concept scheme
defines the semantic relationships between concepts. Concepts are linked
to the concept schemes they belong to via the
skos:inScheme
property.
For example,
skos:broader
and
skos:narrower
enable the representation of
hierarchical relationships. For instance, the triple
country skos:narrower
region
represents a hierarchical relationship where
region
is at a lower level
than
country
. To provide an entry point to the broader/narrower concept
hierarchies, SKOS defines a
skos:hasTopConcept
property. Thus, hierarchical
relationships between dimension members are represented in QB as an
instance of
qb:DimensionProperty
with an associated
skos:ConceptScheme
.
Particular concepts in the concept scheme can be associated with the property
skos:hasTopConcept
, indicating that these values correspond to members at
the highest level of granularity in the dimension hierarchy. Members at lower
levels of granularity can be reached using the
skos:narrower
property. All of
the above imply that dimension members in QB may only be navigated from
higher granularity concepts down to finer granularity concepts. Note that in
general common OLAP operations (except drill down) traverse dimension
hierarchies in the opposite direction.
We show next an excerpt of a dimension
Geography
constructed from the
Ordnance Survey ontology introduced above, represented using QB:
ex:Geography a qb:DimensionProperty, qb:CodedProperty ; qb:codeList ex:geo .
ex:geo a skos:ConceptScheme ; skos:hasTopConcept ns2:921 .
Search WWH ::
Custom Search