In the data warehouse theory a common term to describe the logical representation of data is the OLAP cube. OLAP cubes can be thought of as extensions to the two-dimensional array of a spreadsheet as the data will be arranged as an element of a multi-dimensional cube¹⁾. Often OLAP is described as a category of software technology that enables analysts, managers and executives to gain insight into data through fast, consistent, interactive access to a wide variety of possible views of information that has been transformed from raw data to reflect the real dimensionality of the enterprise as understood by the user²⁾. Practical Example in a company: An analysis of financial data by product, by type of revenue, by cost, by city, and by comparing actual data with a budget.

A cube contains measures, dimensions, hierarchies and levels³⁾. The numeric facts you will find in a cube are called measures and are categorized by dimensions. The cube structure is often created from a star schema or snowflake schema of tables in a relational database. Each of the elements of a dimension could be summarized using a hierarchy. An example for hierarchies could be: May 2009 is summarized into Second Quarter 2009, which again is summarized in the Year 2009.

Each individual point in a cube is referred to as a cell. Due to that fact, cubes support the following basic operations⁴⁾:

• Slicing (cut subsets of the data cube for on-screen viewing)
• Dicing (is a slice on more than two dimensions of a data cube. A smaller sub cube is generated, which contains parts of the data from the original one.)
• Pivoting (rotation of the data in order to provide an alternative presentation of data)
• Drill-Down (to zoom into deeper levels of consolidation. The user navigates among levels of data ranging from the most summarized (up) to the most detailed (down).)
• Roll-Up (opposite operation to drill-down. The user navigates among levels of data ranging from the most detailed (down) to the most summarized (up).)
• Drill-Across (calculations and modeling applied across dimensions. The coexisting dimensions on the same hierarchy level are looked at, e.g. other regions, other products, other months.)
• Drill-Through (reach-through to underlying detail data)

Every dimension represents one 'variable', which can be analyzed separately. Due to this possibility a cube can be used to answer countless questions. For example:

How many books did we sell last week in Berlin?

How many books are in stock?

Which shop sold most books?

Which shops increased its sales within the past 3 years?

The data will be saved in a multi-dimensional way (MOLAP), a relational approach (ROLAP) or in hybrid-configuration (HOLAP)⁵⁾. Traditional OLAP products are also called “multidimensional OLAP” (MOLAP) because they summarize transactions into multidimensional views ahead of time. ⁶⁾ The data is organized into a cube structure. The cube can be rotated by the user, which is of help for example for financial summaries. Relational OLAP tools extract data from relational databases.⁷⁾ ROLAPs are usually using data that has a large number of attributes, which is the reason why the data cannot be easily placed into a cube structure. For example, customer data with numerous descriptive fields are typically ROLAP candidates, rather than financial data. HOLAP (Hybrid Online Analytical Processing) is a combination of ROLAP and MOLAP. HOLAPs store one part of the data in a MOLAP and another part in a ROLAP system. This allows a tradeoff of the advantages of each.