Global Positioning System Reference
In-Depth Information
Each of the classes could have subclasses and those subclasses could have
subclasses, and so on in a treelike hierarchy. Thus <radio base station> could have
a subclasses <long range> and <hotspot>. The domain experts would decide on
the definitions and these would be recorded in the standard.
The second building block is property. Within a class or subclass, an entity
may have properties; for example, a <hotspot> may have a <hotspot standard>
since it could be a WiFi hotspot or a Bluetooth hotspot (or in the future some other
as yet unknown standard). The advantage of this approach is that it lends itself to
modification and extension and is thus future-proof.
When an actual situation is being defined, the building block of “individuals”
or instances is needed. For example, a specific hotspot will need its <type> to
actually have a value (i.e., <WiFi>). A specific base station would have a mapped
location that would take the value of a specific latitude and longitude.
Following the OWL ontological primitives there needs to be a way to deal
with the resources. For example, the definition of a base station will actually be
located as a Web resource, all of which have Uniform Resource Identifiers
(URIs). The familiar Web page Uniform Resource Locator (URL) is a type of
URI. To make sure everything described is unique and not ambiguous, the concept
of “namespaces” is used. When a URL of any sort is examined in detail, it can be
seen that the format /something/something-else/ … is used so that although rather
long and tedious, the actual URL is absolutely unique since the namespace is
always extensible (and thus can be made unique).
A method is needed to create real low-level relationships so that values can be
assigned. The syntax is described in XML, which does not describe anything per
se but rather is a framework used to define domain-specific markup languages.
For example, if GML and SML were viewed as text on a page, their formats
would be similar but the specific terms would be unique. Stahl and Haupert [8]
describe how XML-encoded indoor building models can be useful for positioning
systems. If the actual data is viewed, it will appear to be in the markup scripting
form that is standardized as the Extensible Markup Language (XML). When
viewed, it can be seen to be similar in format to the earlier HTML of the early
Web but is much more versatile. (A good way to see these markup languages is to
view a Web page on a standard browser and on its toolbar select <view> then
<source> .)
To be able to generate lists of data to be included within any instances of a
Whereness ontology (which is really a distributed database) there needs to be a
way to associate objects (i.e., instances of the classes or subclasses) with
appropriate properties and their values. The Resource Description Framework
(RDF) is used. OWL describes terms of the ontology and their hierarchies, but it is
RDF that actually deals with the information (and uses the domain vocabulary that
OWL was used to define). Each line of RDF is rather like one record in a
database: lists of statements serializing the real-world information.
RDF is a powerful concept and it predated OWL and underpins much of the
Semantic Web. Powers [9] describes the practical use of RDF. Three pieces of
Search WWH ::




Custom Search