Data Quality & Taxonomy Governance
Data normalization · taxonomy mapping · QA governance · analytics readiness
A public-safe reconstruction of the systems required to turn inconsistent operational exports into governed, analysis-ready data. The workbook begins with deliberately messy synthetic source records and applies alias mapping, title standardization, learning taxonomy, activity normalization, date validation, severity logic, and an exception-review framework.
All organizations, people, products, source records, dates, and issue counts are synthetic.
Results at a glance
| Metric | Synthetic result |
|---|---|
| Raw source records | 180 |
| Clean / pass | 157 |
| Review required | 23 |
| High-severity rows | 10 |
| Medium-severity rows | 13 |
| QA pass rate | 87.2% |
Exception mix
Job title — 15
Learning taxonomy — 10
Company mapping — 6
Activity type — 4
Event date — 4
Counts are synthetic and intentionally seeded to demonstrate exception handling.
What the project demonstrates
- Company alias normalization to approved organization IDs
- Two-stage title standardization from source variants to canonical title, job family, and seniority
- Learning-title mapping to capability and sub-capability
- Activity-stage normalization across inconsistent source values
- Date validation and quarter derivation
- High/medium severity logic and an exception registry for human review
- Governance rules defining ownership, resolution paths, preventive controls, and a publish gate
- Source-system quality monitoring and analytics-readiness KPIs
Workbook architecture
Executive Dashboard → quality KPIs, exception mix, source-system comparison, and governance controls.
Raw Source Data → deliberately inconsistent synthetic inputs.
Cleaned Output → formula-driven normalization, mappings, QA status, and severity.
Mapping Tables → controlled company aliases, canonical titles, learning taxonomy, and activity values.
QA Exceptions → unresolved issues with severity and recommended resolution.
Governance Rules → field-level standards, owners, resolution paths, and preventive controls.
Data Dictionary + Portfolio Guide → definitions, review guidance, and privacy framing.