228 Part II • Information Technology
models that had not shown up earlier, and these issues
were quickly resolved. SAS Text Miner enables analysts to
zero in on a single performance issue that can be identified
based on text (SAS Web site, 2010).
For over a decade, hospitals have been using data
mining in an effort to identify relationships in their patient
data. For example, one hospital system used data mining to
predict which patients suffering from congestive heart
failure were most likely, after being treated and released, to
be readmitted or, even worse, to die. Data mining identi-
fied unsuspected clusters of data involving patient care that
the hospital used as a starting point for making changes in
clinical procedures. More recently, hospitals are using data
mining to reduce hospital-acquired infections, which saves
lives and shortens hospital stays as well as saves money.
Blue Cross and Blue Shield insurers have launched
infection control programs in selected hospitals in
Alabama, New Jersey, New York, Pennsylvania, and Texas
using MedMined data mining software from Cardinal
Health, based in Dublin, Ohio. The insurance companies
are paying much of the cost of the software, and it is
paying off for them in terms of reduced reimbursements
for hospital-acquired infections. The data mining software
uses data from a hospital’s laboratory and admission-
discharge-transfer systems to develop infection scorecards
for each inpatient unit and identify patterns of statistical
changes that have occurred. Infection control personnel at
the hospital use these data to assess trends at the unit level
and examine processes with the unit staff. Because of data
mining, infection control personnel no longer spend their
time detecting patterns; instead, they automatically get the
data they need to support decisions on changing processes
at the unit level to avoid infections. So far, 14 hospitals in
New Jersey have saved an estimated $6.4 million by
preventing hospital-acquired infections, according to
Horizon Blue Cross and Blue Shield of New Jersey
(Anderson, 2008). Data mining offers exciting possibilities
for learning about customers, particularly for companies
that have well-established data warehouses.
Group Support Systems
In our discussion of groupware in Chapter 5, we noted
that the term collaborationor the phrase collaborative
environmentis often used as a synonym for groupware.
Collaboration, of course, can be aided by many tools,
including e-mail, IM, VoIP telephony, videoconferencing,
shared workspaces, and shared document storage. But
none of the major groupware systems includes a collabora-
tion tool designed to provide comprehensive support for
meetings, whether these meetings are in-person or virtual
meetings. Such a collaboration tool—one that is specifically
aimed at supporting meetings—is a specialized type of
groupware called a group support system (GSS).
GSSs are an important variant of DSSs in which the
system is designed to support a group rather than an
individual. GSSs, sometimes called group DSSs or
electronic meeting systems, strive to take advantage of the
power of a group to make better decisions than individuals
acting alone. Managers spend a significant portion of their
time in group activity (i.e., meetings, committees, confer-
ences); in fact, some researchers have estimated that
middle managers spend 35 percent of their work week in
meetings and that top managers spend 50 to 80 percent of
their time in meetings. GSSs represent an attempt to make
these group sessions more productive.
GroupSystems, the market leader in GSS software,
provides an excellent example for us to consider. The
GroupSystems GSS software was originally developed at
the University of Arizona; in 1989, Ventana Corporation
was formed to offer the GroupSystems GSS as a commer-
cial product. Ventana Corporation later changed its
name to GroupSystems, Inc., and the newest version of
the GroupSystems GSS is now called ThinkTank.
GroupSystems estimates that its “customer base represents
roughly 75 percent market share of the group intelligence
segment of collaborative computing” (GroupSystems Web
site, 2010a). GroupSystems customers include major
corporations such as Agilent Technologies, IBM, Procter &
Gamble, Southern California Edison, Verizon, and Wellpoint
and government organizations such as the Environmental
Protection Agency (EPA), the National Aeronautics and
Space Administration (NASA), the U.S. Air Force, and the
North Atlantic Treaty Organization (NATO).
In a typical in-person implementation of the original
GroupSystems GSS (see Figure 6.2), a computer-supported
meeting room is set up containing a PC for each participant,
all linked by a local area network (LAN). A large public
screen facilitates common viewing of information when
this is desired. GroupSystems, which is installed on each
machine in the network, provides computerized support for
idea generation, organizing ideas, prioritizing (such as
voting), and policy development (such as stakeholder
identification).
Each participant in a group session (e.g., a brain-
storming session) has the opportunity to provide input
anonymously and simultaneously via the PC keyboard.
This can encourage creative thinking because no one can
be ridiculed for a “stupid idea.” Each idea or comment is
evaluated on its merits rather than by who offered it.
Similarly, in a voting session the participants will not be
swayed by how someone else votes. Thus, a GSS such as
GroupSystems should generate more high-quality ideas as
well as decisions that truly represent the group.