OLAP is the abbreviation for Online Analytical Processing. This refers to the provision ofdata for analysis purposes. The most common form of OLAP data structures are so-called OLAP cubes. This is a non-relational data structure designed specifically for analysis.
Widely used tools for creating OLAP cubes are, for example, Microsoft SQLServer Analysis Services or IBM Cognos Business Intelligence Power Play. An OLAP cube is composed of properties (called dimensions) and key figures (called facts).
The dimensions form a multidimensional structure. This is where the name cube comes from, in reference to the three-dimensional basic shape of the same name. In OLAP cubes, the number of dimensions is theoretically unlimited. Each dimension consists of hierarchically structured, distinct values, the so-called attributes. Such proficiencies could be the month, another axis could include sales, net sales, and price, and the third axis could include products 1 through n. All possible combinations of characteristics of all dimensions form the set of account points in the cube. For each such account point, key figures can be defined, e.g. sales per month and region.
These key figures can now be aggregated across the different hierarchy levels of the dimensions or other groupings. This makes OLAP cubes very suitable for analyzing detailed data and exploring unknown relationships. OLAP cubes are therefore often used in controlling departments. However, OLAP cubes are only suitable for number-oriented evaluations. They are less suitable for text-oriented evaluations or lists, classic evaluations or scorecards.