Rule Library Hygiene Check
Description
This recipe evaluates the overall health and governance quality of the Data Quality (DQ) Rule Library by systematically analyzing existing rule metadata. It identifies duplicate rules applied multiple times to the same data elements, orphan rules that are not mapped to any object, and stale rules that have not been executed recently. The recipe consolidates these findings into standardized KPIs and a Rule Library Hygiene Score, providing stewards with clear visibility into rule library issues and enabling informed cleanup, optimization, and governance decisions.
Details
As data quality programs mature, rule libraries often accumulate duplicate, orphan, and stale rules that create governance blind spots, increase execution overhead, and confuse stewardship teams; this recipe provides a structured and repeatable way to assess overall rule library hygiene and identify concrete cleanup opportunities.
Business value
- Reduces governance debt by systematically identifying redundant and unused rules
- Lowers operational cost by highlighting rules that add no active value
- Improves audit readiness with measurable and explainable hygiene metrics
Who benefits
- Data Quality Stewards responsible for rule ownership and maintenance
- Data Governance teams conducting periodic health reviews
- Platform administrators managing execution efficiency
What you receive
How the recipe works (guided flow)
Step 1 — Analyze rule definitions and mappings
The recipe scans rule metadata and object mappings to establish the full inventory of rules in the environment.
Step 2 — Detect duplicate and overlapping rules
Rules applied to the same columns with similar logic are grouped to identify redundancy at the column level.
Step 3 — Identify orphan and stale rules
Rules without object mappings or without recent executions are flagged as governance risks.
Step 4 — Compute hygiene metrics and publish results
Aggregated metrics and a standardized hygiene score are calculated and published as a reusable dataset.
Sample insights (illustrative)
| Insight Category | What the recipe discovered | Business Impact |
|---|---|---|
| Redundancy | Multiple rules validating the same column with similar logic | Unnecessary execution cost and steward confusion |
| Unused assets | Rules not mapped to any object | Governance noise with no quality value |
| Rule freshness | Rules not executed in the last 30 days | Potential misalignment with current data validation needs |
Make sure the following ingredients are available in your workspace:
- Connected datasets are crawled and profiled
- Columns have distinct counts and top values generated