Skip to content

5 Avoid embedding meaning

Stian Soiland-Reyes edited this page Mar 12, 2016 · 7 revisions

Rule 5: Avoid embedding meaning

Permalink URI: https://w3id.org/id-rules/5

The structure and scope of collections evolve, as does scientific understanding; minimizing the meaning embedded in identifiers makes them less vulnerable to obsoletion. In human genetics many genes were initially identified based on disease association. Later the identification, nomenclature, and function of genes were separated into different activities.

It is still possible to embed precision within an identifier: for instance, an InChI string both identifies a chemical entity and defines its structure. Hence, meaning can be embedded where it is indisputable, unchangeable and/or useful to the data consumer (e.g. computer-processable). These rules of thumb apply especially to Local Resource Identifiers but also to the /path of the URIs (see Rule 4).

When assigning identifiers, define what kind of entity is being identified. This information must be provided as an available description and encoded where possible, using metadata landing pages. (Force 11 Data Citation Principles) (ARK identifier scheme).