Database Reference
In-Depth Information
quality and integrity of the database ensured. The system also does not have to worry
about anything declared as out of scope. But, the closed world assumption does not
work well in eclectic environments simply because different models coexist; there-
fore, assumptions that may be valid for one model may be invalid or irrelevant for
another. The use of GI itself is becoming more eclectic as its uses broaden, par-
ticularly through the informal application of GI that marks much of the growth of
Web-based services.
Closed applications require standards that enforce a single model in place of
the many that may have previously existed to make it easier for organizations to
share data. Spatial data infrastructures (SDIs) can be seen as an embodiment of
the International Organization for Standardization (ISO) and OGC standards that
exist for the GI community; they provide a physical framework for the exchange of
GI through shared and standardized models and transport mechanisms. But, such
solutions are not scalable to the Web as a whole, or rather to the Web
community
as
a whole. It is simply impossible to enforce a single world model. The problem is of
course that the GI community's view is not the only view, and even within the GI
community there are many different views. Other communities have their own stan-
dards and frameworks to exchange data. And, the issue is further complicated by the
fact that an organization is rarely a member of a single community. As a result, there
are plenty of occasions when an organization has to take data structured according
to one set of standards and translate it (usually in conjunction with other data) to con-
form to another set of standards. An insurance company concerned with exchanging
property-based data may have to transport data in both GI and insurance industry
standards and will therefore require internal mechanisms to translate between these
standards. The translation process itself is normally done within the company, and
the translated data is unavailable to any other organization facing the same problem.
While this hiding of data may protect a commercial interest, it is inefficient and is
likely to be less attractive in noncompetitive situations such as government or other
data in the public sphere.
The need to enforce standards to efficiently exchange information within a
community has had the unfortunate side effect of reinforcing the insularity of that
community. Another unintended victim of this approach is the end user; the com-
promises made to develop the necessary standards are often made at the expense
of usability of the data for the end user. Computational and databasing efficiencies,
which are often important drivers of standards, do not necessarily provide the best
end-user solutions in terms of data content or comprehension. The availability of
data that without standards might not have been otherwise accessible does mitigate
inconveniences in usability. Nonetheless, standards do not always provide end users
with an optimal solution.
Like many other communities, the GI community has a tendency to think of itself
as the center of the universe, however unconsciously this happens. From this per-
spective, GI is king; everything else is secondary. It is undeniable that in many, many
instances geography is an important element, but it is not always so, and an over-
inflation of this importance can be distorting. Consider the following: An oft-stated
fact
within the GI community is that 80% of all organizational data is geographic.
This statement frequently occurs in publications and talks given by those in the
