Data Analytics: Exploiting the Data Warehouse - Data Warehouse Systems: Design and Implementation - page 323

Chapter 9
Data Analytics: Exploiting the Data
Warehouse
Analytics can be defined as the discovery and communication of meaningful
patterns in data. Organizations apply analytics to their data in order to
describe, predict, and improve organizational performance. Analytics uses
descriptive and predictive models to gain valuable knowledge from data
and uses this insight to guide decision making. Analytics relies on data
visualization to communicate insight. We can distinguish several variations of
analytics depending on the kind of data to be analyzed. While data analytics
copes with traditional structured data, text analytics refers to the analysis
of unstructured textual sources such as those found in blogs, social networks,
and the like. Web analytics refers to the collection, analysis, and reporting of
web data. Finally, visual analytics combines automated analysis techniques
with interactive visualizations, providing effective means to interactively
explore large and complex data sets for decision making.
In this chapter, we focus on data analytics in the context of data
warehousing, that is, on the exploitation of the data collected in the
warehouse to support the decision-making process. We describe several tools
that can be used for this purpose, namely, data mining, key performance
indicators (KPIs), and dashboards. We start in Sect. 9.1 by presenting the
most widely used data mining tasks and the techniques that implement
them. We focus on decision trees, clustering, and association analysis and also
comment on other techniques like regression and pattern analysis. Then, in
Sect. 9.2 , we discuss the notion of KPIs. Finally, in Sect. 9.3 , we explain how
dashboards are used to display KPIs and other organizational information
in a way that the managers can take timely and informed decisions. In all
cases, we implement these techniques over the Northwind case study using
Microsoft Analysis Services.
Note that this chapter is not intended to be a comprehensive presentation
of these topics as there are many topics entirely devoted to each one of them.
We give an introduction to these topics in the context of data warehouses
and point at the end of the chapter to popular references in these domains.