Why conceptual data modeling?
Data modeling has a long tradition and is still very important today. Data modeling has become popular at the beginning of the 1970´s when databases entered the software market, and data were no longer designed according to an application´s needs and stored in separate files.
Since that time data were considered as a resource that has to be designed with care in a data modeling process. Today, data modeling has become a very essential activity within the life cycle of integrated information systems, especially in the context of shared applications.
Data modeling is typically done in three phases (see Figure 1):
- conceptual (or semantic)[1] design
- logical design, and
- physical (or technical) design.
Figure 1: Phases of database design
The deliverables of these phases are the conceptual (or semantic) data model[2], the logical data model, and the physical data model, respectively. The most important model, however, is the conceptual data model. It is the basis for the derivation of the two other data models.
The objective of conceptual data modeling is to describe data structures in terms of entity types, relationship types and attributes or classes, associations and attributes, if object-oriented data modeling techniques are applied. The focus is on semantics, technological or implementation details are out of scope.
In my following articles I want to describe the ER model types that are available in ARIS and to bring out the differences between these model types.
[1] The terms “conceptual“ and “semantic“ are used as synonyms.
[2] The term “schema” is often used as a synonym for “data model”.
