ORIGINAL COMPANY-CONTEXT PRACTICE
Evolve a controlled research vocabulary without rewriting history
Original practice scenario: A hypothetical research platform renames sample categories and splits one category into two. Existing studies still reference the old category. Design a compatible vocabulary model.
Written for independent practice using company context; this is not a verified question from a company interview.
A REASONED APPROACH
Model answer.
Use stable concept identifiers and versioned meaning rather than treating display labels as keys. A category split changes classification semantics and cannot be represented faithfully by a simple rename.
Work through it
- Store concept identity, labels, validity and relationships such as replacement or narrower category, preserving the vocabulary version used by each source record.
- Define whether historical records can be reclassified from available evidence; retain unknown or ambiguous assignments when the old category lacks sufficient detail.
- Support current and historical views through explicit mappings, testing merges, splits, retired concepts and localized labels with clear provenance.
Tradeoffs to discuss
- A current simplified vocabulary aids discovery but may lose historical distinctions; parallel versioned views improve fidelity while increasing mapping maintenance.
Common pitfalls
- Do not assign every old record to one new category merely to satisfy a required field.
- Two labels that look similar may encode different experimental meanings and should not automatically share an identifier.