Data-quality is the degree to which data satisfies defined fitness-for-use requirements — accuracy, completeness, consistency, timeliness, and validity — as measured against a specification and sustained through governance processes. It is a human-made evaluative construct: the concept of data quality does not exist in raw data itself; it is imposed by human systems (databases, compliance frameworks, ML pipelines) that define, measure, and enforce thresholds for acceptable data. The persistence mechanism is institutional practice: data governance policies, quality assurance workflows, and the standardized metrics (completeness rate, error rate, freshness SLA) that organizations maintain and audit. [formal: data qualitas | substrate: behavior | horizon: hours | explicit: yes | epoch: 0.01]
Accepted ontology entry
data-quality
Data-quality is the degree to which data satisfies defined fitness-for-use requirements — accuracy, completeness, consistency, timeliness, and validity — as measured against a specification and sustained through governance processes. It is…
Definition
Why it is in scope
A human-made practice and set of metrics for evaluating how well data serves its intended use — accuracy, completeness, consistency, timeliness, and validity. Built to persist through audit frameworks, quality-assurance protocols, and standardized measurement systems.
Names and aliases
- data-qualityen · CANONICAL
Relations from this entry
- cmr9uz3vv00elhcxfruyltnd4DEPENDS_ON →
Data-quality is the practice of evaluating data through metrics — remove measurement and data-quality cannot operate at all. Measurement is the operational mechanism by which data-quality functions. This passes the removal test (Law 8): without measurement, the concept of evaluating data quality ceases to function.
- cmrnpwjwq02y6d1nl4am3t71xSERVES →
Data-quality is built and maintained for the sake of decision-making. Its designed purpose is to ensure decisions rest on accurate, reliable data. Without data-quality, decision-making operates on garbage. Law 8d: the servant (data-quality) points at the master (decision-making).
Relations to this entry
- cmrz2cszc02s2ekkxlce5xkls← SERVES
Documentation is maintained for the sake of data-quality — it records data lineage, provenance, and quality metrics. Its designed purpose includes supporting data governance practices. Law 8d: servant→master direction correct.
- cmrpq44f107t2d1nlsmovn40n← INSTANCE_OF
Dark-data is a specific kind of data-quality issue: data that exists but cannot be used due to missing metadata, wrong format, or other quality deficiencies. A competent speaker would call dark-data 'a data-quality problem.' Nearest kind.
Record identity
- Created
- Jul 24, 2026, 3:06 PM UTC
- Content hash
- f3da804c03505f8348af44b70a62e7bbdc3285a4387fcb9cf12ad7c4a5f8cd86