Business intelligence is a huge segment of the software world. Gartner Group estimates that sales in this area surpassed $10 billion in 2010, with Oracle as the second largest vendor in the category. At the heart of these analytical-oriented applications are dimensional data models with OLAP as a critical component for achieving high-performance.

If you want to do development work in this area or understand how to maximize its value, read this book. A professional in the world of analytics and business intelligence needs an understanding of OLAP’s specific data modeling principles, its analysis capabilities, and its relationship to other information technologies. All are presented here in a systematic fashion written in an easy-to-follow style.

The Multidimensional-Data Modeling Toolkit takes you on an instructional journey into the world of OLAP. You will get an in-depth look at its analytical possibilities as well as comparison with the approaches used in data mining and statistics. You will learn the design issues and get step-by-step programming instructions for solving real-world problems. You will learn techniques rarely taught in university or technical training programs that will help set you apart as an expert.

The Multidimensional-Data Modeling Toolkit will take you under the covers and show you what happens inside of Oracle’s Analytic Workspaces where the multidimensional magic occurs. Programming instruction is based on the Oracle 10g database, but most of the statements shown will work with other editions of the database, such as Oracle 9i and 11g, and even earlier editions of the technology found in stand-alone products such Oracle Financial Analyzer and Oracle Sales Analyzer

The data analysis principles presented are universal and can be applied to any OLAP system, for example Essbase, Cognos, Business Objects, and Microsoft Analysis Services. Whether you are new to business intelligence or a seasoned practitioner, you should find The Multi-dimensional Data Modeling Toolkit with plenty of valuable insights to offer.

Section Overviews

Preparing for the Journey

In this section, you will be introduced to business analytics in a multidimensional framework. You will get a glimpse into its essential benefits and defining characteristics. You will see a concrete illustration of its capabilities with sample reports based on four-dimensional data. You also will get your first peek into the analytical environment inside which the succeeding explorations will take place and get your introduction to the Oracle OLAP programming language.

Mastering Dimensions

Next, you’ll zero in on the element that underlies everything: the dimension itself. You’ll find out how to use dimensions as a powerful tool for structuring data for analysis. You’ll see the process of creating, describing, relating, and categorizing dimensions. You’ll see how to organize dimensions into hierarchies and explore practical issues that can lead to complications.

Designing Multidimensional Data

In this section, you’ll look at the core elements of the OLAP framework: how multidimensional data is stored, how data can be calculated at display time, how to perform filtering operations, and how to work with sets and combinations. You’ll also see what can be done with data: multidimensional data breakouts, dimension profilers, calculating shares and percentages, time series smoothing and digital filters, segmentations, manipulations with non-numerical data, and aggregations.

Graphical User Interfaces for OLAP

Until now, the focus has been on data modeling and creating analytical components within the dimensional framework using Oracle’s powerful OLAP command language. In this section you will see the graphical interfaces for populating those components with data from a star schema data warehouse that was designed using Kimball techniques. You’ll also see interfaces end users will employ to create reports and conduct sophisticated analyses. This will included report and chart layouts as well as filtering operations.

OLAP DML Programming

Now you are ready to write DML programs. You’ll already have a great background in the language by this point, so writing programs will be but a small step forward. We won’t just go over the syntax of how to create, save, and run a program; we’ll look at the typical kinds of things done with programs in the business intelligence arena. You’ll learn practical programming techniques through a series of examples based on those very applications.

Analytical Alchemy

Here’s where you’ll learn the secrets of analytical alchemy with multidimensional data. I’ll introduce two key ideas: the OLAP advantage and "dimensionalizing" an analysis. Through them, I’ll characterize the benefit that OLAP offers and articulate a key design principle that you can use to create unique and powerful measures. Application of the principle will be thoroughly illustrated via a powerful OLAP-based customer-tracking model.

Given a full-fledged OLAP analysis platform, you may be curious to know how it compares to the other analysis platforms—statistics and data mining. In the final chapter of this section, I will discuss the relationship between these differing approaches.

Disclaimer

This site and the products and services offered on this site are not associated, affiliated, endorsed, or sponsored by Oracle, nor have they been reviewed tested or certified by Oracle.