Definition: A taxonomy is a controlled system for classifying and relating content, products, documents, or concepts through consistent terms and rules. It defines which categories exist, how they are named, and how they relate to one another so that information can be organized, retrieved, and reused. On a website, it can support navigation, filters, search, archive pages, and metadata, but it is not the same as the entire information architecture.
Table of contents
Elements of a Taxonomy
A taxonomy is more than a list of tags. For different people and systems to classify information consistently, it must describe the accepted terms and the rules governing their use. Common elements include:
- Concepts and terms: each concept represents an idea or class, while a term is the word used to name it.
- Preferred and alternative labels: a primary name avoids unnecessary variants, while synonyms help recognize other search expressions.
- Scope notes: they explain what a category includes or excludes and reduce ambiguous decisions.
- Relationships: they connect broader, narrower, or associated concepts. For example, the footwear category may contain trainers, while material describes a different dimension.
- Application rules: they specify who can create terms, when terms are assigned, and how duplicates, renaming, or obsolete categories are handled.
Terms can be represented by stable codes or identifiers so that changing a label does not break its connection to content. The W3C SKOS model can express concept schemes, labels, and semantic relationships in interoperable systems.
Ways to Organize Concepts
The structure depends on the domain and the tasks it needs to support. An organization can combine several models without treating them as unrelated classifications:
- Controlled list: it contains authorized terms without necessarily arranging them into levels. It is useful for statuses, formats, or content types.
- Hierarchy: it arranges concepts from general to specific. Each level should follow the same criterion to avoid mixing, for example, product types with brands or audiences.
- Polyhierarchy: it allows a concept to have more than one broader category when both relationships are valid.
- Faceted classification: it separates independent dimensions, such as type, material, use, price, or location, and allows them to be combined as filters.
A hierarchical taxonomy can support browsing paths, but not every relationship indicates membership. Connecting two related concepts does not mean that one is a subcategory of the other. This distinction prevents artificial trees and helps determine whether a hierarchy, facet, or associative relationship is appropriate.
Taxonomies on Websites and in Content Management Systems
On a website, a taxonomy classifies content and can provide the basis for menus, filters, search results, breadcrumbs, and pages that group items together. A content management system stores these assignments and makes it possible to reuse them in different interfaces. WordPress, for example, distinguishes among categories, tags, and custom taxonomies.
These components are related, but they are not equivalent. Metadata describes properties of a resource and can use values from a taxonomy. A folksonomy emerges from free-form user tagging, while a controlled taxonomy limits terms and their relationships. An ontology adds a more formal representation of entities, properties, and constraints.
Web architecture also covers navigation, page hierarchy, URLs, search, and other organization systems. Taxonomy is one of its components. It can support internal linking and topical grouping, but it does not guarantee a better experience or greater organic visibility on its own. Archive pages provide value only when they have a clear purpose and do not create duplicates or combinations with insufficient content.
Designing and Maintaining a Taxonomy
Design starts with real content and tasks, not with an ideal tree created in isolation. An inventory helps identify existing concepts; search queries and support questions reveal the vocabulary people use; and techniques such as card sorting help show how different people group information.
The process usually involves several connected decisions:
- Define the scope: establish which resources will be classified and for which users, channels, and tasks.
- Gather the vocabulary: extract terms from content, domain specialists, and the language used by users.
- Normalize and relate: choose preferred labels, merge duplicates, document scope, and build consistent hierarchies or facets.
- Test retrieval: determine whether people can find, filter, and distinguish the expected content without relying on internal knowledge.
- Establish governance: assign owners, addition and retirement criteria, version control, and rules for changes affecting URLs, filters, and existing content.
- Evaluate usage: review unassigned terms, overly broad categories, searches without results, overlaps, and paths that do not help complete a task.
There is no universal review frequency. A taxonomy should be updated when the catalog, language, navigation needs, or systems that use it change. Maintaining identifiers, equivalences, and a change log allows it to evolve without losing the traceability of previous classifications.