classification is normally used for classifying the statistical units of a population in disjoint groups, so that the union of these groups is the population itself. For instance, the population of residents in Italy on January 1, 2018, can be classified by the classification “sex” in males and females. The population of enterprises in Lombardia on December 31, 2016, can be classified by the “classes of the number of employees” (enterprises with a number of employees up to 9, from 10 to 49, from 50 to 99, 100 and more).

Hence, in the field of statistics, a classification denotes a list of categories that describe how a phenomenon (measured by a variable) is associated with the statistical units of a population, composed by mutually exclusive (incompatible) and collectively exhaustive (universal) categories.

More complex examples of classifications are the hierarchical ones. In a hierarchical classification the categories are arranged in a tree-structure with two or more levels, where each level contains a set of mutually exclusive categories. The items of each level but the highest (most aggregated) are aggregated to the nearest higher level. An example is given by the classification “administrative-territorial units”: regions and provinces are two contiguous levels; every region is composed by the set of its provinces.

Classifications are also characterized by versions. Versions describe the modifications of the classification as time goes by. For instance, the classification of the economic activities (ATECO) was defined in its last version in 2009 (ATECO version 2009). This version substituted the former one, defined in 2002 (ATECO version 2002).