What is a Conceptual Data Model Anyway?

Everyone has heard the term Conceptual Data Model, but do you know what is meant by it or where it came from?

The terms conceptual, logical and physical in the context of data architecture originated in the Zachman Framework. The Zachman Framework provides an ontology of components for the description of the architecture of any object or system. It does this by posing a set of 6 questions (what, where, when, who, how and why) as columns in a table with 6 rows, each representing a different perspective (Planner, Owner, Designer, Builder, Implementor and User).  Distinct artefacts are proposed at the intersection of the columns and rows.


Source – Zachman International (Zachman.com)

Data Modelling is concerned with the What, i.e. what does an object or system consist of; what is its structure.  The 3 perspectives of interest to data modelers are the Business Management (Owners), Architect (Designers) and Engineer (Builders) perspectives, representing what we have come to know as the Conceptual, Logical and Physical models respectively.

Knowing who the audience is for any particular model certainly helps us to understand what type of language, the level of detail and degree of rigour that should be used, but of itself, does not give a complete picture of the purpose of a model.
When someone tells me “this is our Logical Model”, I can infer that it is some sort of design model, but this label alone tells me nothing of the scope (enterprise, department, project, etc.) or scenario (as-is, to-be, etc.) let alone the methodology or notation system employed.

You may know what you mean when you label a model conceptual, logical or physical, but be aware that these labels are outdated, imprecise and incomplete, and do not aid communication between author and audience

One of the main functions of a model is communication.  This would be better achieved if modellers clearly stated from the outset what is the purpose, perspective, granularity, scenario and methodology of their model.  Furthermore, having these aspects clear in your own mind from the start and throughout your modelling work will make you a better modeler.

Sandhill will be looking at the recent and forthcoming changes to the erwin Data Modeler in our upcoming webinar on June 27th at 10pm EST.

Register Now