Solving Problems in Industry - Java Data Mining: Strategy, Standard, and Practice - page 24

Chapter
2

Solving Problems in Industry

No problem can stand the assault of sustained thinking.
—Voltaire

Data mining has long been used to solve important problems across
industries [Apte 2002] [Berry/Linoff 2004] [Gloven 2002]. One of

the biggest challenges to using data mining is knowing to which
types of problems it can readily be applied. Too often, a discussion of
data mining delves into a specific algorithm and its myriad features.
However, success with data mining goes beyond choosing the right
algorithm. Understanding the requirements in a particular domain,
and the types of problems and approaches to solving those problems,
is the best starting place to reap the benefits of data mining technol-
ogy. This chapter explores various business problems that are com-
mon to many industries and that can be addressed with a strategy
based on data mining.
As Internet and digital storage technology enables volumes of
data to grow astronomically, so does the need for analyzing that
data to extract useful information. However, data mining is not only
for large institutions with terabytes of data. Data mining's benefits
can be leveraged by companies both big and small, from large finan-
cial institutions to local car dealerships, those with millions of
customers, and those with hundreds, and those with scientific as
well as manufacturing process data analysis needs.

25