Database Reference
In-Depth Information
TABLE 4.1
School Performance Assessment
School
Road
Area
City
Postcode
Assessment
Bay
Elm Lane
Bove
Eagleton
EA1 7GG
5
Court Place
Brambles Road
Bove
Eagleton
EA1 8QT
3
Beech House
West Street
Central
Eagleton
EA2 9WW
8
St John's
Derby Road
Hasly
Eagleton
EA2 1AC
6
Deprivation Areas
Health Index
Area
Deprivation Score
Postcode
Index
Bove
3.9
EA1 7GG
97.2
Hasly
3
..
Central
1.2
EA1 8QT
88
Merriby
2.3
..
EA2 WW
93.4
..
EA2 1AC
93.7
deprivation are identified by the name of the area and the health issues by postcode.
We are able to relate this data using conventional means such as relational data-
bases as there are common factors that can draw the data together. The use of GIS
in this case is not necessary.
Table 4.1 shows the relational database tables for schools, health, and deprivation
data. The address element of the school will enable us to associate the school both
to the areas of social deprivation via the area element of the address and to health
issues data through the postcode element. It is now possible to look at the correla-
tion (without implying causation) that social deprivation and local health issues have
with the school performance results; however, this can only be achieved by stating
the relationships using an SQL Query; the associations that exist between the data
are not explicit.
However, if the data were published as Linked Data, then an initial step in under-
standing these data structures and converting them to a common model would be sig-
nificantly reduced, if not removed altogether, and the relationships between the various
elements represented as explicit links rather than revealed indirectly through an SQL
Query. By publishing the links on the Web, there is a third benefit as now these links
are available to others. Table 4.2 and Figure 4.2 show how the data is represented.
In conclusion, the Semantic Web can improve the processes of locating and pre-
senting data to a GIS for analysis, but the GIS is much better at performing arithmetic
analysis. The Semantic Web is better suited to cases where the task can be accom-
plished through inference or graph analysis. In both examples, the representation of
different datasets in the common data model of Linked Data reduces effort required
to massage data so that it is suitable for analysis, and the links themselves can be
made available to others, providing onward benefits.
Search WWH ::




Custom Search