Glossary Updates6 new terms added to the glossaries · October 11, 2026, 13:38 CEST
AI TechDocKnowledge

Glossary · 1 · Documentation models

Semantic documentation

Also known as: Semantic technical documentation, Semantically enriched documentation

German: semantische Dokumentation

In technical communication, semantic documentation is documentation whose meaning is represented in machine-readable form — by semantic markup within the content and by metadata about the content that refers to controlled vocabularies, taxonomies or ontologies — so that software can select, filter, link, deliver and check the information by what it is about, not only by where it stands in a document.

  • Documentation models
  • Metadata
  • Semantics

In one sentence

Semantic documentation represents the meaning in and about content in machine-readable markup and metadata, so software can select and deliver it.

Example

Each maintenance topic of a packaging machine carries metadata for the product variant, the component, the lifecycle phase and the information type, taken from a controlled vocabulary; a service app shows a technician only the topics for the variant and component scanned on site.

How it applies

  • Two layers: Semantics in content is expressed by semantic markup that names what a piece of text is — a parameter, a hazard, a part number. Semantics about content is expressed by metadata whose values come from a taxonomy or an ontology, so that a topic is described by product, component, audience, lifecycle phase or information type.
  • Standards and vocabularies: In DITA, subject scheme maps define controlled values for attributes and their hierarchies. In iiRDS, every iiRDS Topic is described with RDF metadata from the iiRDS ontology, so that documentation can be delivered in a standard package to content delivery portals and other consumers. RDF and JSON-LD are the general formats for such statements on the web.
  • Machinery documentation: Semantic documentation lets a portal or app filter instructions by machine variant and situation, link fault messages on an HMI to the matching troubleshooting topic, and supply information to systems such as the asset administration shell. The metadata has to be governed like the content; wrong metadata delivers wrong information. Semantic enrichment makes documentation easier to find and process; it does not make it compliant or its content correct.
  • AI-based retrieval: Metadata and typed content help retrieval systems select the right passages and keep them in context; see AI-ready content and retrieval-augmented generation.

Semantic documentation vs. Every Page is Page One

Every page is page one is a writing model: it sets requirements for how each page is written so that a human reader can start on it. Semantic documentation is a representation model: it makes the meaning of content explicit for machines. The two are complementary — metadata can record the context an EPPO page establishes in its text, and an EPPO page gives a semantically selected topic a reader can use.

External references