An OLAP semantic model is a structured multidimensional dataset; it has many dimensions and pre-calculates aggregations ahead of time. OLAP semantic models organize data in a hierarchical arrangement according to dimensions and measures. See MOLAP, ROLAP, and HOLAP.
The structure of a semantic model makes it easy to visualize or conceptualize data along various dimensions of a semantic model, making it easy to query and interact with the semantic model.
OLAP (Online Analytical Processing) is a set of operations that one can do on a dataset to analyze data, for example, pivoting, slicing, dicing, or drilling.
- Slice - is the act of picking a rectangular subset of a semantic model by choosing a single value for one of its dimensions. For example, the sales figures for all sales regions and all product categories of the company in a year can be "sliced" out of the data semantic model.
- Dice - is a slice on more than two dimensions of data. A dice, for example, shows the sales figures of a limited number of product categories, while the time and region dimensions cover the same range as the slice example.
- Drill up/down - allows you to navigate among levels of data ranging from the most summarized (up) to the most detailed (down).