THE DATA WAREHOUSE - Fundamentals of Database Management Systems - page 337

In the past, files were developed to support individual applications that we would
now classify as DSS applications. For example, the five-year sales trend analysis
for retail stores described above has been a fairly standard application for a long
time and was always supported by files developed for it alone. But, as DSS activity
has mushroomed, along with the rest of information systems, having separate
files for each DSS application is wasteful, expensive and inefficient, for several
reasons:

Different DSS applications often need the same data, causing duplicate files to be
created for each application. As with any set of redundant files, they are wasteful
of storage space and update time, and they create the potential for data integrity
problems (although, as we will see a little later, data redundancy in dealing with
largely historical data is not as great a concern as it is with transactional data).

■

While particular files support particular DSS applications, they tend to be
inflexible and do not support closely related applications that require slightly
different data.

■

Individual files tied to specific DSS applications do nothing to encourage other
people and groups in the company to use the company's accumulated data to gain
a competitive advantage over the competition.

■

Even if someone in the company is aware of existing DSS application data that
they could use to their own advantage (really, to the company's advantage),
getting access to it can be difficult because it is ''owned'' by the application for
which it was created.

■

When we talked about the advantages of data sharing earlier in this topic,
the emphasis was on data in transactional systems. But the factors listed above
regarding data for decision support systems, which in their own way largely parallel
the arguments for shared transactional databases, inevitably led to the concept of
broad-based, shared databases for decision support. These DSS databases have
come to be known as '' data warehouses .'' In this chapter, we will discuss the
nature, design, and implementation of data warehouses. Later in the chapter we will
briefly touch upon some of their key uses.

13-A S MITH &N EPHEW

CONCEPTS

IN ACTION

S mith & Nephew is a leader in
the manufacture and marketing of medical devices.
Headquartered in London, UK, the company has over
7,000 employees and operations in 34 countries.
Smith & Nephew focuses on three areas of medical
device technology, each run by a separate business
unit. In orthopedics, Smith & Nephew is a leading
manufacturer of knee, hip, and shoulder replacement
joints, as well as products that aid in the repair of broken
bones. In endoscopy, the company is the world leader

in arthroscopic surgery devices for minimally invasive
surgery of the knee and other joints. Last, the company
is the world leader in providing products and techniques
for advanced wound management. All of this from a
beginning in 1856 when Thomas J. Smith opened a
pharmaceutical chemist shop in Hull, England. And, yes,
he later brought his nephew into the company.
Smith and Nephew supports its orthopedics prod-
ucts business with a state-of-the-art data warehouse. This
data warehouse incorporates daily sales and inventory