Database Reference
In-Depth Information
truth of a statement is independent on whether it is known to be true or false by any
single person. That one person does not know a fact to be true does not mean that it
is false. This may seem to be intuitively obvious, but it is not the supposition made
by most databases. Here, the closed world assumption applies: It is presumed that a
statement is false unless it is known to be true in the database. This strategy is valid
in that most databases are built to be a closed solution: They assume their universe
comprises only the things of interest to the database; therefore, if a fact is not stated
in the database, it is not true, or at least not true within the scope (narrow worldview)
of the database. The scope of the database therefore defines the limits of knowledge.
This is a useful assumption to make for most databases as it means that data can
be checked for compliance to the data model. But, it is not valid for the Web as the
scope of the Web is without limits. The Web is also decentralized and has no global
data model; it grows organically, and therefore it is not possible
a priori
to know how
new information will it, and it is certainly inevitable for there to be conflict between
the data on the Web. Things could not be more different from the orderly, defined,
and controlled world of a database.
On the Web, the open world assumption holds sway. This is an extremely impor-
tant observation. It means that for the Semantic Web there can be no single world-
view, no God's eye view, no one model that fits all. It also means that
facts
may
conflict with each other, and different people may resolve these conflicts differently
depending on their own worldviews.
The degree of agreement on the Semantic Web is at the lowest level possible, with
the standard languages (Resource Description Framework [RDF] and Web Ontology
Language [OWL]) used to encode knowledge. Beyond this, and within the knowl-
edge encoded in OWL, the open world assumption applies. To think sensibly about
the Semantic Web, and indeed the Web as a whole, one has to assume that one's own
knowledge is incomplete, and that publishing on the Web contributes to a greater
pool of knowledge.
The Semantic Web also provides an explicit and formal description of the data
through the use of ontologies. An important aspect of this is that the ontology is
independent of both the data and any application code. This makes the ontology very
visible and provides greater flexibility because ontologies are easy to share and reuse
or adapt for specific situations.
4.2.2 G
eoGraphic
i
nformation
With a field as diverse as GI, there are no comparable founding principles as there
are with the Semantic Web. Whereas the Semantic Web is built on the open world
assumption, there is no such positive assertion for GI. However, and accepting the
caveat that because of this diversity there can be no universal statements, it is fair
to say that for the most part GI is founded on the closed world assumption. This is
a reasonable statement given that the technologies most commonly used to imple-
ment GI today, Geographic Information Systems (GIS) and relational databases, are
implemented based on the closed world assumption. Furthermore, this assumption
has also been carried forward into Open Geospatial Consortium (OGC) standards.
Consider the following short paragraph describing an OGC feature: “Any feature may
