232 Part II • Information Technology
Vector-based GISs are widely used in public
administration and utilities and, arguably, are the most
common approach used in business. Vector systems
associate features in the landscape with either a point, a
line, or a polygon. Points are often used to represent small
features such as ATMs, customer addresses, power poles,
or items in motion, such as trucks. Lines are for linear
features such as roads and rivers and can be connected
together to form routes and networks. Polygons represent
areas and surfaces, including lakes, land parcels, and
regions—such as sales territories, counties, and zip codes.
The relationships between the vector elements are called
their topology; topology determines whether features
overlap or intersect. Vector systems can distinguish, for
example, an island in a lake, two roads crossing, and
customers within a 2-mile radius of a retail site. However,
vector data are not continuous; the resulting overlaps and
gaps between features affect presentation and analysis and
thus can require a specialist’s attention.
The most common data model for both vector and
raster data is the coverage modelin which different layers
or themes represent similar types of geographic features in
the same area and are stacked on top of one another (see
Figure 6.4). Like working with transparent map overlays,
the layers allow different geographic data to be seen
together, and they facilitate geographic manipulation and
analysis.
Most GIS technologies today effectively combine
both types of data, often using raster data sets for realism
and vector data for roads, administrative boundaries, and
locations. By employing both types of data, geographic
analysis can answer questions such as the following:
- What is adjacent to this feature?
- Which site is the nearest one, or how many are within
a certain distance? - What is contained within this area, or how many are
contained within this area? - Which features does this element cross, or how
many paths are available? - What could be seen from this location?
If you have looked for your house on Google Earth,
you have used the zoom feature as well as panning and
centering your display based on a map overlay—these are
basic capabilities of any GIS. Other basic capabilities
include finding the distance between two points, searching
for and labeling specific features, and turning symbols and
layers on and off on demand. These functions underlie
spatial manipulation such as intersection and union, the
assignment of geographic references to addresses through
geocoding, and standard query language support for inter-
acting with descriptive (attribute) data. Once limited to
high-end workstations, advanced GIS applications—now
on the desktop or palmtop—automate sophisticated deci-
sion support tasks such as finding the shortest/fastest/safest
route from A to B or grouping sales or service territories to
minimize internal travel distance, equalize potential, or
omit the fewest prospects.Issues for Information Systems Organizations
Business applications of GISs are often initially introduced
into a company to support a single function such as market
research or field service. Experience shows us that GISs
soon spread within and across groups. Thanks to the matu-
rity of GIS tools, organizations can acquire off-the-shelf
geographic technologies with scripting languages, applica-
tion program interfaces with popular desktop software pack-
ages, and Internet-based interactive mapping packages (such
as Microsoft’s MapPoint Web Service). As the cost of GIS
technologies has dropped and technical sophistication has
risen, even computationally demanding functions such as
route optimization and territory assignment are widely avail-
able (do your competitors already have these functions?).
Managing geographic technology options, now that
they are available on familiar platforms, may be less
challenging to a typical IS organization than managing
spatial data. Obvious geographic data (which you will
want to buy, not build) include base maps, zip code maps,
street networks, and advertising media market maps. Other
data with spatial elements are spread around in internal
company databases, including customer locations, loca-
tions of company warehouses and distribution centers, and
the location of fixed and movable assets. Recently, a new
wrinkle in spatial data management has arisen: Handheld
GPS devices now allow users to collect their own data and
download it to a PC.
Because the value for a business in “going spatial”
comes from bringing internal and external data together, IS
personnel can expect to get an education in cost and
quality issues for geographic data. For example, although
geographic files for zip codes are often included at no
additional cost in packaged desktop GIS software, the U.S.
Postal Service updates zip codes on an ongoing basis,
resulting in decay in the accuracy of existing data. For
geodata coverage outside the United States, one can expect
fewer choices in terms of available data, and the data that
are available are likely to be less accurate, more difficult to
obtain, and more expensive.
Many people have been exposed to mapping tech-
nologies through household-name Internet sites such as
Yahoo! Maps or Google Earth. The “GIS engines” behind
these sites come from a less well-known pool of vendors