Research Program: Paradigms

The Atomic State Paradigm

The Atomic State Paradigm attempts to synchronize the data model with the business' high level event management process model. The centerpiece is the domain of the highly derivable status attribute of each entity.

Rationale

As entity instances live their lives, they pass through a series of logical states.
We must understand the transition rules affecting these states to correctly model the business' data and provide the process rules that will support process modeling.
Remember: The instance isn't in a particular state because the Status Code attribute says so.... The Status Code attribute has its value because the entity instance is in that state.
Do not be trapped by the illusion that the field is more important than the business rule.
Since states are decomposable, any set of field values will represent ranges in the entity instance's life, obscusing more discrete stages.
Distinctions within these ranges result in much of the exception processing that must be built into our computer systems today.
Event + State = Process!
More discrete stages will prevent massive processing changes as the business rules change.
If logical data attributes are identified that are unique to specific states, then consider creating subtype entities to more accurately describe their characteristics, especially relationships with other entity types.

Trigger

Multiple business events being conducted on data whle in a common business state depending upon the value of other attributes in the data model.

This paradigm is taught in the Quality-Based Data Modeling seminar.

Return to... DOQS Home Page, DOQS Research Page