Object tagging is a big topic in data governance conversations; it leads to discussions on taxonomies, enterprise and faceted search efforts and ultimately to AI with intent.
From the get go the data vault metadata tags staged in data vault are the all too familiar tags we see such as:
- Record source (mandatory): where the data came from
- Business key collision codes (mandatory): to ensure shared business keys from different platforms don’t clash
- Multi-tenant id (optional): a column used to designated multiple tenants/users of the data vault, a shared model but not shared content
- Load date timestamp (mandatory): to record when the record was loaded to the data vault
- Applied date (recommended): the metadata date time stamp of the packet of data loaded to the data vault
- Jobid / taskid (recommended): the job code of the processes that loaded that record
- Userid (optional): the system user id used to load that record
What about Jira-id to tie the project/scrum effort that was responsible for getting that record in?
What other metadata tags would make sense to add to your data vault?