TRENDS IN DATABASE TECHNOLOGY - Database Design and Development: An Essential Guide for IT Professionals - page 658

DETERMINATION
OF OBJECTIVES

PREPARATION
OF DATA

APPLICATION OF DATA
MINING TECHNIQUES

EVALUATION AND
APPLICATION OF
RESULTS

Define
Objectives

Select
Data

Extract
Data

Preprocess
Data

Mine
Data

Review
Results

Select
Promising
Patterns

Present
Results
(text/charts)

Apply
Results

All
Results

Results
Presen-
tation

Objec-
tives

Selected /
Extracted
Data

Selected
Results

Preprocessed
Data

Enterprise
Data
Warehouse

Enterprise
Operational
Systems

Figure 20-19 Knowledge discovery process.

the uncovered hidden knowledge manifests itself as relationships or patterns. Figure
20-19 illustrates the knowledge recovery process.

OLAP Versus Data Mining As you know, with OLAP queries and analysis, users
are able to obtain results from complex queries and derive interesting patterns. Data
mining also enables the users to uncover interesting patterns, but there is an essen-
tial difference in the way the results are obtained. Figure 20-20 clarifies the basic
difference by means of a simple diagram.
When an analyst works with OLAP in an analysis session, the analyst has some
prior knowledge of what he or she is looking for. The analyst starts with assump-
tions deliberately considered and thought out. In the case of data mining, however,
the analyst has no prior knowledge of what the results are likely to be.
Users drive OLAP queries. Each query may lead to a more complex query and
so on. The user needs prior knowledge of the expected results. It is completely dif-
ferent in data mining.

Data Mining in Data Warehouse Environment Data mining fits well in the data
warehouse environment. It plays a significant role in the environment. A clean and
complete data warehouse forms the bedrock for data mining. The data warehouse
enables data mining operations to take place. The two technologies support each
other. These are some of the major factors:

• Data mining algorithms need large amounts of data, more so at the
detailed level. Most data warehouses contain data at the lowest level of
granularity.
• Data mining flourishes on integrated and cleansed data. If your data extracts
from source systems, data transformation functions, and data cleansing proce-