Data quality, traceability and protection are important aspects for the use of data in analytical processes.
With the methods and technical tools of data governance, the process of data management is monitored and documented, data assets and sources are documented technically and functionally, and access is controlled as needed. This creates the basis for any decision-making, whether manual or automated by AI / ML.
Completeness
A new key value, a changed source column or simply an error in the system: the reasons for a hard termination of a job, or - even worse - unnoticed incomplete data management are manifold.
Correctness
Technically correct does not necessarily mean functionally correct! Even if data is complete, Data Governance establishes the context by cataloging the data so that users can check and evaluate the technical correctness for the use case.
Origins analysis
The origin of the data ("lineage") is an important criterion for deciding for or against a data asset in a use case. Likewise, the reverse case ("impact analysis") reveals important aspects when it comes to the assessment for properties to be changed.
Access protection
Based on business or technical metadata, manually or automatically, access to data must be managed and controlled, or data must be anonymized or pseudonymized.