Multidimensional Databases

Multidimensional databases is a key technology in the enabling of interactive analyses of large amounts of data for decision-making purposes. Multidimensional databases differ from previous technologies by viewing data specifically as multidimensional cubes, which have proven to be particularly well suited for data analyses. Multidimensional models more specifically categorize data as either facts—which are the entities of interest w.r.t. the analyses at hand, e.g., a purchase in a retail business—or dimensions, which are mostly textual and characterize the facts, e.g., the location of the purchase, the product being purchased, and the time of the purchase. Facts have associated numerical measures that capture the properties of the facts that are to be studied, e.g., the total sales price. This article describes multidimensional concepts such as cubes, facts, dimensions, measures, and querying, including advanced issues, and it introduces implementation techniques such as MOLAP, ROLAP, and pre-aggregation.