The Modeling Room: (M4) Handle realistic exceptions - (A20) Handle customer types
The Modeling Room, module 4, article 20 of 26. Give each customer type explicit attribute rules, keep unknown, not applicable, and invalid apart, and build one specialist view whose joins preserve the common customer identity and counts.
Who should read it?
Students learning modelstorming, analysts, report builders, data modelers working with business experts
Prerequisite
Keep a recorded occurrence distinct from its covered population, as in article 19
Learning objective
Give each customer type explicit attribute rules without losing the common customer identity and counts
Artifact
A customer type card, a common customer table, and one specialist view with a tested join rule
Why this matters
A customer-data audit lists Northline Couriers as incomplete because its date of birth is blank. Anna Bennett could invent a date to satisfy the form. The invented date then enters age calculations and mailing selections, and months later nobody remembers which dates are real. The better repair changes the rule: a company does not have the attribute the audit is demanding.
What you'll learn
- How to give each customer type explicit attribute rules while keeping one customer identity and the all-customer counts.
- How to build one specialist view whose joins do not multiply rows.
Your output: a customer type card, a common customer table, and one specialist view with a tested join rule.
New terms: defining characteristic, exclusive attribute