Introduction - Data Warehouse Systems: Design and Implementation - page 3

data warehouse can be broadly summarized as follows (this list by no means
attempts to be comprehensive):
￿ Reporting, such as dashboards and alerts.
￿ Performance management, such as metrics, key performance indicators
(KPIs), and scorecards.
￿ Analytics, such as OLAP, data mining, time series analysis, text mining,
web analytics, and advanced data visualization.
Although in this topic the main emphasis will be on OLAP as a tool to exploit
a data warehouse, many of these techniques will also be discussed.
In this chapter, we present an overview of the data warehousing field,
covering both established topics and new developments, and indicate the
chapters in the topic where these subjects are covered. We give in Sect. 1.1
a historical overview of data warehousing and OLAP, starting from the
early achievements. Then, we describe in Sect. 1.2 the field of spatial and
spatiotemporal data warehouses, which has been increasingly used in many
application domains. Finally, in Sect. 1.3 we describe new domains and
challenges that are being explored in order to answer the requirements of
today's analytical applications.
1.1 A Historical Overview of Data Warehousing
In the early 1990s, as a consequence of an increasingly competitive and
rapidly changing world, organizations realized that they needed to perform
sophisticated data analysis to support their decision-making processes.
Traditional operational or transactional databases did not satisfy the
requirements for data analysis, since they were designed and optimized to
support daily business operations, and their primary concern was ensuring
concurrent access by multiple users and, at the same time, providing recovery
techniques to guarantee data consistency. Typical operational databases
contain detailed data, do not include historical data, and perform poorly
when executing complex queries that involve many tables or aggregate large
volumes of data. Furthermore, when users need to analyze the behavior of
an organization as a whole, data from several different operational systems
must be integrated. This can be a dicult task to accomplish because of the
differences in data definition and content. Therefore, data warehouses were
proposed as a solution to the growing demands of decision-making users.
The classic data warehouse definition, given by Inmon, characterizes a
data warehouse as a collection of subject-oriented, integrated, nonvolatile,
and time-varying data to support management decisions. This definition
emphasizes some salient features of a data warehouse. Subject oriented
means that a data warehouse targets one or several subjects of analysis
according to the analytical requirements of managers at various levels of the