Meta-data is often confused with meta-models, mirroring the confusion between data and information. That misconception may be progressively reconsidered as the result of two major developments, both calling for a uniform and principled description of data independently of business-specific contents: compliance with privacy regulations, and the training of Large language models (LLMs). Hence the benefit of defining meta-data as labels meant to characterise provenance and modus-operandi without being specific about actual contents.
Exemples
- A. D. Trump said ” “
- B. Forbes estimated Trump’s net worth at $3.1 billion
- C. 100.000 people demonstrated for …
- D. Year on year inflation has been
- E. Iceland earthquakes during the last 48 hours
- F. Books most frequently added to Goodreads members’ shelves, updated weekly
Who
Ownership (past and present) of data
- A. D. Trump, public
- B. Forbes, public
- C. Reuters, public
- D. Office for National Statistics, public
- E. Icelandic Meteorological Office, public
- F. Google, public
Where
Location (physical or symbolic) of the source of data
- A. NY Court
- B. Forbes
- C. Reuters
- D. idem
- E. idem
- F. idem
When
Life cycle (initial and revision timestamps) of data
- A. Oct 2, 2023
- B. Oct 3, 2023
- C. Nov 13, 2023
- D. Nov 15, 2023
- E. Nov 15, 2023
- F. Nov 15, 2023
What
Nature (symbolic, digital, physical) of data
- A. Text/Speech
- B. Numerical
- C. Numerical
- D. Numerical
- E. Document
- F. Document
How
Modality (managed, observed, assessed, predicted) of data
- A. Observed
- B. Assessed
- C. Assessed
- D. Measured
- E. Observed
- F. Managed
FURTHER READING
KALEIDOSCOPE SERIES
- Signs & Symbols
- Generative & General Artificial Intelligence
- Thesauruses, Taxonomies, Ontologies
- EA Engineering interfaces
- Ontologies Use cases
- Complexity
- Cognitive Capabilities
- LLMs & the matter of transparency
- LLMs & the matter of regulations
- Learning
- Uncertainty & The Folds of Time