A symbol, string, or token that unambiguously denotes an entity — a resource, person, organisation, device, dataset, or concept — within a defined scope, enabling that entity to be referenced, located, compared, and linked independently of its descriptive attributes. Identifiers are the foundational primitive of naming systems: their guarantees of uniqueness, persistence, and resolvability within a namespace determine what URIs, UUIDs, DIDs, and registry codes can be trusted to do.
Semantic Classification
Content
Definition
An identifier is a name whose job is denotation: it picks out exactly one entity within a stated scope so that systems and people can refer to that entity reliably. Unlike a description, an identifier carries no obligation to say anything true about what it names — “urn:isbn:0451450523” identifies a book whether or not you know its title. This separation of reference from description is what makes identifiers composable: records, links, indexes, and protocols can all be built on the stable string while the entity’s attributes change.
Identifier schemes vary along a small set of design axes. Scope defines where uniqueness holds — locally (a database primary key), organisationally (an employee number), or globally (a URI, a UUID). Assignment may be centralised through a registry, hierarchical through delegated Namespace authority (DNS, ISBN), or decentralised through cryptographic self-generation, as in content hashes and Decentralised Identifier methods. Persistence governs whether an identifier may ever be reassigned; resolvability determines whether it can be dereferenced to reach the entity or its description; and opacity distinguishes meaningless surrogate keys from semantically loaded codes that embed classification and risk breaking when the world changes.
In this graph, the identifier class anchors a family of specialisations and mechanisms: centralised identifiers issued by registries, self-sovereign decentralised identifiers, URI schemes that syntactically partition the global identifier space, and encoding conventions such as Multicodec that make an identifier’s format self-describing. Reliable identifiers are in turn the substrate on which Digital Identity, linked data, and interoperable registries are built.
Technical Details
-
Global schemes: URI/IRI (RFC 3986/3987) for web-scale naming; URN namespaces for location-independent names; UUID (RFC 9562) for coordination-free 128-bit uniqueness.
-
Sector registries: ISBN and DOI for publications, ORCID for researchers, LEI for legal entities, GTIN for trade items — centralised or federated authorities guaranteeing uniqueness and metadata.
-
Cryptographic identifiers: content-addressed hashes (Git commits, IPFS CIDs) and DID methods derive the identifier from key material or content, making integrity verifiable without a registry.
-
Persistence practice: “cool URIs don’t change” — persistent-identifier infrastructure (DOI, ARK, PURL) separates the stable name from mutable locations via resolution services.
-
Design pitfalls: reusing identifiers, embedding mutable semantics (smart keys), and scope collisions when local identifiers leak into global contexts are the classic failure modes.
Current Landscape
-
The UUID standard was fully modernised by RFC 9562 (Proposed Standard, May 2024), which obsoletes RFC 4122 and adds sortable, timestamp-ordered UUIDv7 plus the reordered UUIDv6 and free-form UUIDv8 — a direct response to database and distributed-systems needs where random UUIDv4 keys fragment index locality.
-
W3C Decentralized Identifiers (DID) Core v1.0 became a W3C Recommendation in July 2022, defining a cryptographically verifiable, registry-independent identifier that (unlike a UUID) is both resolvable and provable; DID methods and the accompanying Verifiable Credentials work remain the fastest-moving area of new identifier design.
-
The URI/IRI family (RFC 3986/3987) still underpins web-scale naming, while ISO/IEC 9834 / ITU-T X.667 keep the OID and UUID registries technically aligned across ISO and the IETF.
-
Persistent-identifier infrastructure (DOI, ORCID, ROR, ARK) continues to expand in research and cultural-heritage settings, reinforcing the “cool URIs don’t change” separation of stable name from mutable location.
Sources: