Metadata is the lexicon of the digital organisation, making data easily findable & identifiable; consistently interpretable, manageable & usable throughout your organisational ecosystem. Metadata actuates your systems. Metadata is, of itself, data: and is richer and more important than the data it describes. Often, metadata forms part of systems of record and is immensely valuable in making AI or ML meaningful.
Types of metadata
There are different types of metadata, each meeting a different set of organisational needs:
Administrative
process metadata: derived, gives the status of the data within a particular process or workflow;
governance metadata: should detail how the item is to be retained, managed, stored, or erased, for control, compliance & audit purposes;
rights metadata: precisely details distribution, licensing, publication or other digital property rights.
New item list
Descriptive
descriptive metadata: describes what the digital asset itself is;
classification metadata: give the hierarchical context of data within an overarching taxonomy;
content metadata: gives context within a content topic to related items;
technical metadata: provide special (often proprietary, scientific or industry-specific) technical information about the data, its digital origins or how it was created;
metric metadata: describing any measurement or metric parameters that apply to the data, especially necessary if downstream conversion is anticipated.
Structural
structural metadata: provide logical context to the data within virtual logic, physical data model or structure;
object metadata: how data should be handled by the system, platform or cloud;
programmatic metadata: programmable cloud metadata, for writing or customising cloud functions.
Meta
when the state of the metadata itself needs to be described, e.g. time-stamped, for audit or data protection purposes;
when the metadata requirement is itself subject to change e.g. as regulatory requirements evolve.
Metadata audit use case
A metadata audit is recommended if you have acquired data sources through corporate acquisition, merger, or departmental consolidation! Otherwise, you might only know of a potential privacy, governance or security exposure when it’s too late!
A metadata audit will:
assess the quality of existing metadata;
assess the quality of metadata salience against current taxonomy;
assess quality of vocabulary & metrics, control & standardisation;
assess the quality of current metadata against business need or use cases;
assess the level of metadata completeness & note any remediation potential;
advise any accuracy issues which might affect analytics;
advise any potential governance, audit or compliance issues;
advise any security, credentials, rights management implications or exposure;
advise quick win metadata opportunities which might benefit supply chain, sales or customer relations processes;
advise any feasible opportunities for realising value through the use of historic metadata for comparative analysis;
advise on metadata suitability for any forthcoming AI/ML or knowledge initiatives;
outline recommendations about improving how the organisation acknowledges & treats metadata.
