Skip to content

Hidden Label

A string attached to a controlled-vocabulary concept that is indexed for retrieval but withheld from display — so a user typing a typo, deprecated synonym, or embarrassing prior name finds the concept while the result still shows only the sanctioned preferred term.

Core Idea

A hidden label is a string attached to a concept in a controlled vocabulary that is indexed for retrieval but deliberately withheld from public display. A user who types the typo, deprecated synonym, dialect variant, or embarrassing prior name finds the concept, but the surfaced label is the sanctioned preferred term. The device separates two functions a name conflates: recognition during search and authority during presentation. In SKOS it is skos:hiddenLabel, the third tier below prefLabel and altLabel; PubMed indexes "heart attack" as a hidden term for "Myocardial Infarction."

Scope of Application

The device applies wherever one substrate recurs — a retrieval index run over a catalog with separable index and display layers.

  • Thesauri and subject vocabularies — the canonical use: MeSH entry terms and SKOS hiddenLabel strings.
  • Product search — misspellings and competitor names mapped onto catalog items but off the page.
  • Wiki redirects — old or confusable titles routing to the current article, kept off its visible name.
  • DNS and email aliasing — addresses routing a message without appearing in the rendered signature.

Clarity

Naming the hidden label makes legible a behavior that looks like a defect: the search bar resolves a term the result page refuses to call itself by. With the concept, that mismatch is recognized as a deliberate split between recognition during search and authority during presentation. It poses a three-way decision the flat model cannot — accept a variant as input, display it as sanctioned, or exclude — and clarifies that an altLabel carries a display obligation while a hidden label carries none.

Manages Complexity

A vocabulary's headache is the unbounded variety of strings users send at a concept, forcing an impossible trade on a flat name list: admit each variant and bloat the public face with typos and slurs, or refuse it and lose findability. The hidden label splits "the names of this concept" into two separately sized surfaces governed by one per-variant routing call. The maintainer then reads off that the index is legitimately wider than the display, the gap being the absorbed long tail, and that a name's commitment is a one-bit fact.

Abstract Reasoning

The foundational move is a per-variant routing inference read off a string's character — accept-as-input-only, accept-and-display, or exclude — licensed by separating recognition from authority. A diagnostic move reclassifies an apparent defect (a search/display mismatch is the device working, not a stale index), paired with a commitment inference reducing a label's obligations to one bit. A predictive/boundary move follows from the width asymmetry: pouring variants onto the hidden tier widens findability while the display holds constant, presupposing an index-versus-display split.

Knowledge Transfer

Within information science the hidden label transfers as mechanism across every system running a retrieval index over a controlled vocabulary — the three-tier structure, the routing call, the width asymmetry, and the silent-update property carry intact from thesauri to product search to wiki redirects to DNS aliasing, because these are one substrate in different wrappers. Beyond that substrate the case is shared-abstract-mechanism: what carries to a truly different system is not "hidden label" but the parent primes markedness (a privileged canonical form and variants) and aliasing (multiple names for one referent). Those carry the portable lesson — recognition-canonical and display-canonical can come apart; the SKOS apparatus stays home.

Relationships to Other Abstractions

Local relationship map for Hidden LabelParents appear above the current abstraction, mutual partners to the right, and children below. Node labels state whether each abstraction is prime or domain-specific; colors identify relation types.Hidden LabelDOMAINDomain-specific abstraction: Preferred label — presupposesPreferred labelDOMAINPrime abstraction: Alias-to-Authority Mapping — is a decomposition ofAlias-to-Author…PRIME

Current abstraction Hidden Label Domain-specific

Parents (2) — more general patterns this builds on

  • Hidden Label presupposes Preferred label Domain-specific

    Hidden status requires a separate sanctioned Preferred Label that the result displays after the hidden string has performed only its recognition role.

  • Hidden Label is a decomposition of Alias-to-Authority Mapping Prime

    Hidden Label is Alias-to-Authority Mapping specialized so a variant is accepted by the resolution layer but withheld from every public name surface.

Hierarchy paths (9) — routes to 4 parentless roots

Neighborhood in Abstraction Space

Hidden Label sits in a crowded region of the domain-specific corpus (13th percentile for distinctiveness): several abstractions share nearly its structure, so a description that fits it tends to fit its neighbors too.

Family — Surface Form & Underlying Structure (23 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-07-12