Via my "context lenses" (i.e., my subjective view
of the world) a unit of Data (or Datum) is like a
cube of sugar, each side representing a value factor along the
following dimensions:
- Identity -- via Resolvable URIs based Names for everything
- Data Representation Format Dexterity -- e.g., HTTP based
Content Negotiation which loosens the coupling between Data Model
Semantics and actual Data Representation (Syntax/Markup)
- Platform Agnostic Data Access -- e.g. via ubiquitous HTTP
- Change Sensitivity -- data warehouses are like real-world
warehouses, goods rot and perish overtime
- Provenance -- data about the data (metadata) that helps
establish "Who", "What", "When", "Where", and at least approximate
or guesstimate "Why"
- Data Mesh Navigability -- delivered via inference rules.
The quality of service factors above nullify many of the typical
concerns associated data driven business models, such as:
- Wholesale Imports (crawls) - where your data is crawled and/or
imported wholesale into a new data space with zero attribution to the
source
- Lossy Attribution -- attribution is delivered in literal form
which doesn't deliver branding fidelity across many value chain
layers or entire life cycle of a given data item
- Service Provisioning -- effectively build any business model if
you can align services with unambiguously identifiable consumers
with actual data items or across entire data spaces.