Chapter 1
The Evolution Of The Data
Warehouse
While the use of databases as a
vehicle for complex data analysis is new, the need to perform
complex data analysis has been with us for centuries. Answering
“what-if” questions, simulating the introduction of a new product,
or determining the most profitable products are all legitimate
business needs, and the advent of the data warehouse did not herald
the first time that computers had been used to solve these types of
tasks.
In fact, computers have been used to
solve complex types of data analysis problems since their earliest
commercial inception in the 1950s. Essentially, the nature of the
questions that data warehouses solve has not changed in four
decades. The summarization of numbers, the aggregation of associated
data, and data variance analysis is nothing new.
The Evolution Of Data Management
Platforms
Regardless of the sophistication of
a database manager, it remains true that all databases are
constructed from simple data structures such as linked lists,
b-trees, and hashed files. In reviewing the building blocks of
database systems, it is possible to gain a historical perspective on
the evolution of databases, and to remember the past, so that we are
not condemned to repeat it.
The following pages are a historical
review of database evolution, showing each of the enhancements that
were introduced with each new architecture. It is also important to
review the problems inherent to each database architecture. As you
will see, there are striking similarities between object-oriented
databases and earlier database architectures. By understanding the
historical reason that object-oriented databases have evolved into
their present form, we can gain insight into the future trend and
directions of databases.
and include the URL for the page.
