Identification and existence dependency
The method of extending cardinalities via the specification of lower and upper limits as discussed in chapter
ERM - eERM extensions enables certain dependencies between data objects to be defined.
By definition, relationship types and reinterpreted relationship types do not exist autonomously, but come into being due to the existence of the entity types they link. This means that they depend on other entity types in terms of both identification and existence.
In addition, there are entity types that depend on the existence of other entities even though they have their own key attribute. These dependencies may be the result of a grouping operation, for example. Thus, in the example of the following figure, a department requires at least one assigned workstation, while defining a workstation, in turn, implies that it be assigned to a department. As shown in the following figure, these existence-related dependencies are expressed by specifying the complexity. A (min, max) notation uses (1,1) and (1,*). The definition of existence-related dependencies within the data model results in conditions for the referential data integrity when implemented. In other words, complying with these conditions ensures that the consistency of the database contents is preserved even after certain transactions have been performed. In the example below, this means that a department can be deleted only if all workstations assigned to this department are also deleted.