The concept (see skos:Concept) is defined by a natural language label. SKOS provides three properties for attaching labels to a concept, the most important being skos:prefLabel (see; is a preferred label for a concept, only a single preferred label is allowed per language) and skos:altLabel, which we are discussing here.

These are alternative labels for a concept, e.g. synonyms, abbreviations and acronyms (there can be many).

In thesauri prior to SKOS it was very often referred to as the non-preferred term: use for entry term; entry term; non-descriptor. A term that is not assigned to documents but is provided as an entry point in a thesaurus or alphabetical index. A non-preferred term is followed by a reference to the appropriate preferred term or terms, e.g., hounds USE dogs.

Alternate lexical labels and relationships between labels


The first characterizations of concepts are the expressions used to refer to them in natural language: their labels. SKOS provides three properties for attaching labels to concept resources: skos:prefLabel, skos:altLabel and skos:hiddenLabel. Each property implies a specific status for the label it introduces, ranging from a strong, one-to-one denotation relation to a string to aid search. These properties are formally defined as pairwise disjoint. This means, for example, that it is an error if a concept has the same literal as both a preferred label and an alternative label.

skos:prefLabel, skos:altLabel and skos:hiddenLabel provide single labels. These are all subproperties of rdfs:label, and are used to bind a skos:Concept to a simple RDF literal, which is a string (e.g., “love”) combined with an optional language label (e.g., “en-US”).

Alternate lexical labels

The skos:altLabel property allows you to assign an alternative lexical label to a concept. This is especially useful when assigning labels beyond the one preferred for the concept, for example, when synonyms need to be represented:

ex:animals rdf:type skos:Concept;
skos:prefLabel “animals”@en;
skos:altLabel “creatures”@en;
skos:prefLabel “animaux”@fr;
skos:altLabel “creatures”@fr.

Note that the representation of synonyms for preferred labels is not the only use of skos:altLabel. Near-synonyms, abbreviations and acronyms can be represented in the same way:

ex:fao rdf:type skos:Concept;
skos:prefLabel “Food and Agriculture Organization of the United Nations”@en;
skos:altLabel “FAO”@en.

Note on bottom-up posting: It is also possible to use skos:altLabel to represent cases of bottom-up posting [ISO-2788]. That is, when a concept aggregates more specialized notions that are not explicitly introduced as concepts in the considered KOS:

ex:rocks rdf:type skos:Concept;
skos:prefLabel “rocks”@en;
skos:altLabel “basalt”@en;
skos:altLabel “granite”@en;
skos:altLabel “slate”@en.

However, while SKOS is not intended to replace existing guidelines for KOS design [ISO-2788, BS8723-2], the reader should be aware that their upward publication is not recommended. A more appropriate KOS for this domain would introduce a skos:Concept for each rock type considered (basalt, granite and slate) and assert it as a narrower ex:rock concept.

Linkages between tags

Some applications require the creation of explicit links between tags associated with concepts. For example, consider the relationship between a preferred label for a concept “Corporation” and its abbreviation “Corp.” coined as an alternative label, or a translation link between two labels in different languages: “Cow”@en and “Vache”@fr. The use of SKOS lexical labeling properties, e.g. skos:prefLabel, is restricted to RDF literals. Therefore, these labels cannot be the subject of an RDF statement, and no direct relationship between them can be asserted.

To solve this representation problem, the SKOS vocabulary has been extended with an optional extension for labels, SKOS-XL [SKOS-REFERENCE]. This extension introduces a skosxl:Label class that allows labels to be treated as first-order RDF resources. Each instance of this class will first be attached to a single RDF literal via the skosxl:literalForm property. Consider the example where the concept “Food and Agriculture Organization of the United Nations” is labeled by both the official name and the acronym of the institution. The two labels can be represented as follows:

ex:FAOlabel1 rdf:type skosxl:Label;
skosxl:literalForm “Food and Agriculture Organization of the United Nations”@en.
ex:FAOlabel2 rdf:type skosxl:Label;
skosxl:literalForm “FAO”@en.
Instances of skosxl:Label can be related to concepts via properties (skosxl:prefLabel, skosxl:altLabel, skosxl:hiddenLabel) that reflect standard literal-based labeling constructs. Finally, these instances can be linked together using skosxl:labelRelation declarations:

ex:FAO rdf:type skos:Concept;
skosxl:prefLabel ex:FAOlabel1;
skosxl:altLabel ex:FAOlabel2.
ex:FAOlabel2 skosxl:labelRelation ex:FAOlabel1.

However, this solution is not complete: an “acronym-sensitive” application would lose the actual information that the two labels are indeed in an acronym relationship. Such an application would also miss the direction of the link. Therefore, SKOS-XL users are encouraged to specialize skosxl:labelRelation to fit the specific requirements of their application, as in the following case:

ex:isAcronymOf rdfs:subPropertyOf skosxl:labelRelation.
ex:FAOlabel2 ex:isAcronymOf ex:FAOlabel1.

Note that the SKOS-XL data model ensures that the use of this pattern remains compatible with standard SKOS labeling practice. If an instance of skosxl:Label is attached to a concept by, for example, a skosxl:altLabel statement, it follows from the SKOS-XL data model that the literal form of the skosxl:Label instance is related to this concept by a standard skos:altLabel statement. In the above example, ex:FAO therefore has “FAO”@en” as an alternative label (literal).

Fact checker: Conrad

More Information

Note the following concepts and definitions:

  • Controlled vocabulary: prescribed list of terms or rubrics, each of which has an assigned meaning. Controlled vocabularies are designed for use in classifying or indexing documents and for searching. They typically contain a single preferred term for each concept or entity with links to that term from non-preferred terms. They may also show the relationships between terms.
  • The descriptor: use of the preferred term.
  • The equivalence relationship: relationship between two terms representing the same concept. When two or more such terms are found in the same monolingual thesaurus, one of them is designated as the preferred term and the other(s) as the non-preferred term(s); the relationship is known as intra-vocabulary equivalence. When both terms are preferred terms in different thesauri, the relationship is known as inter-vocabulary equivalence.
  • Indexing. Intellectual analysis of the subject matter of a document to identify the concepts represented in it, and assignment of corresponding preferred terms to enable information retrieval. The term “subject indexing” is often used for this concept, but in a context that does not deal with other elements such as authors or dates, “indexing” suffices.
  • Metadata. This is data describing the characteristics of a document. metadata is essentially a catalog record, providing (a) access points by which document records can be classified or retrieved and (b) descriptive information, by which the relevance of a document can be assessed without consulting it in its entirety. Preferred terms or annotations selected during the indexing process are often applied as metadata elements to describe the subject matter of a document.
  • The multilingual thesaurus. A thesaurus that uses more than one language, in which each concept is represented by a preferred term in each of the languages, and there is a single structure of hierarchical and associative relationships between concepts that is independent of language.
  • The non-descriptor: use the non-preferred term.
  • The preferred term: use for the descriptor. A term specified by a controlled vocabulary for use in representing a concept in indexing. For example: schools; school uniform; schooling costs; tuition. A preferred term should preferably be a noun or noun phrase.
  • The orphan term. It is a preferred term that has no hierarchical relationships.
  • Post-coordinate indexing. This is an indexing system in which the subject of a document is parsed into its constituent concepts by an indexer, but the preferred terms so assigned are not combined until selected by a user at the search stage. Compare this with precocious indexing.
  • Pre-combined indexing. This is an indexing system in which the preferred terms assigned to a particular document are syntactically combined into one or more sequences that represent the only combinations available for retrieval. Compare with post-coordinate indexing.
  • The string: a sequence of preferred terms representing a compound concept in a precoordinate indexing system.
  • The string: a sequence of preferred terms representing a compound concept in a precoordinate indexing system.
  • The thesaurus. It is a controlled vocabulary in which concepts are represented by preferred terms, formally organized so that paradigmatic relationships between concepts are made explicit, and preferred terms are accompanied by entries of synonyms or quasi-synonyms.



, ,




Leave a Reply

Your email address will not be published. Required fields are marked *