Skip to main content
Records Authority
Category: Metadata Standards

Descriptor

Simply put

A descriptor is a word, phrase, or feature used to describe or identify something so it can be recognized and found again. In recordkeeping and information management, descriptors help label and organize records so that they can be located and understood. The exact meaning of the term varies considerably depending on the field in which it is used.

Formal definition

In general usage, a descriptor is something, typically a word, phrase, or characteristic feature, that serves to describe or identify an item. The term carries distinct technical meanings across different domains: in software and data systems it may refer to a structure that defines a data element (for example, a data descriptor that specifies the definition of a table, including elements such as the table name), while in computer vision a feature descriptor denotes a representation designed to enable tasks such as efficient object search within images. Because the evidence available does not include a recordkeeping-specific definition, practitioners should treat these as illustrative senses rather than a settled records management definition, and should scope the term to the particular system or standard in which it appears. Note that a descriptor is a means of description or identification and should not be confused with the item it describes.

Why it matters

The term "descriptor" carries no single settled meaning, and its practical significance depends heavily on the domain and system in which it appears. In general usage it refers to a word, phrase, or characteristic feature that serves to describe or identify something. For information professionals, the risk lies in assuming a common definition across contexts: a descriptor in a software data structure, in a computer vision system, or in an indexing vocabulary may each denote something quite different. Treating these senses as interchangeable can lead to misconfigured systems, ambiguous specifications, and metadata that fails to serve its intended purpose of retrieval and identification.

Because the evidence available does not provide a recordkeeping-specific definition, practitioners should be cautious about importing the term into records management contexts without first anchoring it to a particular standard, system, or controlled vocabulary. Where the term is encountered, its scope should be established explicitly by reference to the environment in which it is used. This discipline matters most where descriptors govern how records are labeled, indexed, and subsequently located, since imprecise or inconsistent descriptor usage typically undermines the findability and comprehensibility of records over time.

A further, recurring point of confusion is the distinction between a descriptor and the item it describes. A descriptor is a means of description or identification; it is not the record, data element, or object itself. Preserving this distinction helps avoid conflating metadata with content, which is a foundational concern in information management practice.

Who it's relevant to

Records managers and information governance officers
Those responsible for how records are labeled and located should be alert to the term's ambiguity. Since the available evidence offers no settled recordkeeping definition, they should scope any use of "descriptor" to the specific system or controlled vocabulary in place and avoid treating it as a standard records management concept without qualification.
Metadata and classification specialists
Practitioners designing or maintaining metadata schemes and indexing approaches encounter descriptors as a means of describing and identifying items so they can be found again. They should keep the descriptor distinct from the item it describes, and document precisely which sense is intended in a given scheme.
Systems and data professionals
Those working with data systems may encounter the technical sense of a data descriptor, which defines a data element such as a table, including elements like the table name. They should interpret the term according to the software or standard in use rather than a generic definition.
Professionals working across domains
Anyone integrating tools or specifications from adjacent fields, such as computer vision, where a feature descriptor supports object search in images, should confirm the domain-specific meaning before applying the term. The senses illustrated here are distinct and not interchangeable.

Inside Descriptor

Controlled term
A descriptor is typically a term drawn from a controlled vocabulary, thesaurus, or authority file rather than free text, so that the same concept is represented consistently across a collection.
Subject or content representation
A descriptor commonly stands for the subject matter, function, activity, or attribute of a record or aggregation, supporting retrieval by what the record is about rather than only by its title.
Relationship to a classification scheme
In many recordkeeping systems descriptors are linked to a classification scheme or functional taxonomy, helping to place records within a structured arrangement that reflects business activity.
Metadata element
A descriptor often functions as a metadata element attached to a record, aggregation, or entity, contributing to the description that supports identification, management, and retrieval over time.
Scope and usage guidance
Well-formed descriptors are typically accompanied by scope notes or usage rules that clarify the intended meaning, distinguish near-synonyms, and reduce inconsistent application by different users.

Common questions

Answers to the questions practitioners most commonly ask about Descriptor.

Is a descriptor the same thing as a keyword?
Not exactly, though the terms are often used loosely as synonyms. A descriptor is typically a preferred term drawn from a controlled vocabulary or thesaurus, selected to represent a concept consistently across a collection. A keyword, by contrast, is often any free-text term a user might apply or search on, without control over synonyms or variant forms. The distinction matters because descriptors are intended to reduce ambiguity and bring related material together under a single agreed term, whereas uncontrolled keywords can scatter related records across many variant expressions. Depending on organizational policy, some systems blur this line by allowing both, but professionally the value of a descriptor lies in its controlled, standardized nature.
Does applying descriptors amount to classifying a record?
Not in the strict recordkeeping sense. Classification, as generally understood, concerns placing a record within a structured scheme that reflects business functions and activities and that often carries consequences for aggregation, retention, and disposition. Descriptors are typically a form of indexing or subject description that support retrieval and access by describing what a record is about. The two can coexist and complement one another, but they serve different purposes: classification situates a record within a controlled organizational structure, while descriptors add access points. Treating descriptor assignment as a substitute for classification can leave records without the contextual and lifecycle controls that classification is meant to provide.
How should descriptors be selected and assigned in practice?
In many implementations, descriptors are drawn from an established controlled vocabulary, thesaurus, or authority file rather than composed ad hoc. Assignment may occur at capture, either manually by staff, semi-automatically through system suggestions, or through automated means, depending on the tools and policies in place. Consistency is generally improved by documenting rules for how many descriptors to apply, how to choose among related terms, and how to handle new concepts. Organizations often designate responsibility for maintaining the vocabulary so that terms remain coherent over time.
Who is typically responsible for maintaining the set of descriptors an organization uses?
Responsibility depends on organizational structure, but maintenance of a controlled vocabulary or thesaurus often sits with records or information management staff, sometimes in collaboration with subject-matter experts and system administrators. Maintenance work can include adding new preferred terms, retiring obsolete ones, managing relationships between terms, and reconciling variant forms. Because uncontrolled growth or inconsistent editing can undermine retrieval, many organizations establish governance for changes rather than allowing individual users to alter the vocabulary at will.
How can descriptors be kept consistent across a large volume of records?
Consistency is commonly supported by constraining the choices available to those applying descriptors, for example by drawing terms from a managed pick-list or authority-controlled field rather than free text. Documented indexing rules, training, and periodic review of applied terms can help reduce drift over time. Some systems offer term validation or suggestion features to steer users toward preferred terms. The degree of consistency achievable will depend on the tools available, the clarity of the vocabulary, and the resources devoted to quality control.
What should be considered when introducing or changing descriptors over time?
When vocabularies evolve, organizations generally need to consider how existing records described under earlier terms will remain findable. Retiring or renaming a descriptor without accounting for previously assigned instances can fragment retrieval. Practices such as retaining relationships between old and new terms, or documenting when and why changes were made, can help preserve continuity of access. The appropriate approach depends on the capabilities of the system and the organization's policies for managing change to controlled vocabularies.

Common misconceptions

A descriptor is just a keyword the user types in freely.
While descriptors and free-text keywords both aid retrieval, a descriptor is usually a controlled term governed by a vocabulary or authority source, whereas an uncontrolled keyword may be any word chosen ad hoc. The controlled nature is what gives a descriptor its consistency, though practice varies by organizational policy.
Descriptors and the classification scheme are the same thing.
A classification scheme typically provides an overarching structure for arranging records, often reflecting functions and activities, while descriptors are terms used to characterize subject or content. They frequently work together, but they are distinct components and should not be conflated.
Assigning descriptors is a purely descriptive act with no bearing on records management.
In many systems descriptors support not only retrieval but also processes such as classification and, indirectly, the application of retention and disposition rules where those are tied to functional or subject categories. Their role therefore often extends beyond simple labeling, depending on how the system is configured.

Best practices

Draw descriptors from a controlled vocabulary, thesaurus, or authority file rather than free text, so the same concept is represented consistently across the collection.
Provide scope notes or usage guidance for each descriptor to clarify intended meaning and to distinguish it from near-synonyms, reducing inconsistent application by different users.
Maintain the alignment between descriptors and any associated classification scheme, keeping the two distinct in purpose while ensuring they work together coherently.
Govern the vocabulary through a defined maintenance process, reviewing, adding, and retiring descriptors as business functions and terminology evolve, in line with organizational policy.
Train those who apply descriptors and, where feasible, apply consistency checks so that terms are assigned reliably and retrieval remains dependable over time.
Record descriptors as structured metadata associated with the relevant record or aggregation, so they persist and remain usable across the record lifecycle.