COMPEL Glossary / metadata
Metadata
Metadata is data that describes other data -- information about a dataset's source, format, creation date, quality metrics, ownership, access permissions, update frequency, and usage history.
What this means in practice
Rich metadata enables AI teams to discover relevant datasets, evaluate their fitness for a particular use case, understand their limitations, and comply with governance requirements. Without metadata, data assets are opaque: teams cannot determine what a dataset contains, how reliable it is, who owns it, or whether they are allowed to use it for AI training. In the COMPEL maturity model, metadata management maturity is assessed as part of Domain 6 (Data Management and Quality), with organizations progressing from no metadata (Level 1) through standardized metadata with business glossary entries and lineage documentation (Level 3) to rich contextual metadata enabling self-service data discovery (Level 4+).
Why it matters
Without metadata, data assets are opaque: teams cannot determine what a dataset contains, how reliable it is, who owns it, or whether they are allowed to use it for AI training. Rich metadata transforms data from a mystery into a managed asset, enabling self-service discovery that accelerates AI project initiation. Organizations with poor metadata management force AI teams into time-consuming manual data investigation for every new project.
How COMPEL uses it
Metadata management maturity is assessed as part of Domain 6 (Data Management and Quality) during Calibrate. COMPEL tracks progression from no metadata (Level 1) through standardized entries with lineage documentation (Level 3) to rich contextual metadata enabling self-service discovery (Level 4+). The Model stage designs metadata standards, the Produce stage implements management tooling, and the Evaluate stage measures metadata completeness and usage.
Related Terms
Other glossary terms mentioned in this entry's definition and context.