What is a domain model, why does every enterprise need one, and how do you build one?
A domain model is a conceptual model of the key business objects and their relationships within an area of interest — not a database schema or a physical data model, but a high-level logical map of what an organisation knows and cares about, spanning all business units and activities. Its primary value is shared understanding: when business and IT personnel use the same terms consistently, requirements are clearer, integration is simpler, and the endless miscommunication that plagues large programmes is significantly reduced.
This white paper introduces domain modelling from first principles using a notation derived from UML and extended with concepts from James Odell's work — covering concept boxes, relationship lines, subtype arrows, composition diamonds, role markers, and decomposition conventions. It explains when and how to use each construct, and illustrates them with worked examples from banking. The paper then covers how to enrich the model with metadata: for each business object, recording ownership, user community, storage location, criticality, volumes, retention, privacy requirements, and more — turning the domain model into a foundation for data governance, information security policy, migration planning, and master data management.
Written as an accessible introduction for architects, analysts, and business stakeholders, the paper is deliberately concise — five pages — making it a practical quick-start reference for any team beginning an information architecture or enterprise architecture initiative.
Pages: 5
Originally published as a white paper by Graham McLeod, Inspired/Promis Ltd, January 2009.