Databases Reference
In-Depth Information
The Purpose of Dimensional Models
A specific modeling technique has evolved in order to support the types
of queries and analyses that businesses require. This technique is called
dimensional modeling . This approach has been applied to data warehousing for
nearly thirty years and is supported by a wide variety of database platforms
and data access or business intelligence tools. Dimensional models support
the business perspective of the data, and today's technology ensures that they
can be effectively implemented.
Dimensional modeling is a formal data modeling technique that is used
to organize and present data for analytical and reporting use. The focus
is on the business perspective and representation of data. The goal is to
free the data that has been captured and stored by the operational systems
and make it available to the business community. Regardless of how data
is structured, business people will ask questions based upon their frame of
reference. This perspective is driven by the basic characteristics of the industry
and how the company is organized, so why not organize the data to reflect
this business perspective? The two primary goals for dimensional modeling
are ease of use and query performance . These are the principles that guide the
entire dimensional modeling process.
There are other data modeling techniques that play an important part in
overall systems development. They help ensure that the data itself and the rela-
tionships between different data elements are clearly defined. For operational
systems, it is important that the data be organized to facilitate transaction
processing. This includes ensuring transaction integrity and speed. The type
of modeling used for operational system design is called entity-relationship
(E-R) modeling . This may also be referred to as normalized modeling .Onespecific
form of E-R modeling represents the data in third normal form (3NF) .There
is a complete discipline surrounding this approach to data modeling. This is
mentioned to acknowledge the value and purpose of E-R modeling for opera-
tional system design. The following two sections examine the main objectives
of dimensional modeling.
Ease of Use
In order to ensure that people will use a data warehouse, the data must be
presented in a manner that makes sense to them. If it is too confusing or
does not mirror the way the business runs, then people are not likely to
use it. Therefore, the dimensional model must cleanly represent the basic
components of the business. In addition, the model must be presented in terms
that are used by the business.
A well-designed dimensional model should be obvious to the business
community and be met with a confirmation that it is indeed correct. If it causes
Search WWH ::




Custom Search