Please enter at least 2 characters!
EuroVoc is a multilingual thesaurus which was originally built up specifically for processing the documentary information of the EU institutions.
It is a multi-disciplinary thesaurus covering fields which are sufficiently wide-ranging to encompass both Community and national points of view, with a certain emphasis on parliamentary activities. EuroVoc is a controlled set of vocabulary which can be used outside the EU institutions, particularly by parliaments.
The aim of the thesaurus is to provide the information management and dissemination services with a coherent indexing tool for the effective management of their documentary resources and to enable users to carry out documentary searches using controlled vocabulary.
However, EuroVoc has limitations.
The EuroVoc thesaurus covers all the activity fields of the European Union institutions:
Some fields are more highly developed than others because they are more closely related to the EU’s centres of interest. Thus, for example, the names of the regions of each EU Member State are in EuroVoc but not those of third countries.
The grouping of descriptors into fields is to a certain extent arbitrary. One of EuroVoc’s distinctive features is the limitation of polyhierarchy. Descriptors which could fit into two or more subject fields are thus generally assigned only to the field which seems the most natural for users, in order to facilitate the management of the thesaurus and limit its volume.
The maintenance team at the Publications Office collects and examines the proposals, put forward by the national parliaments, the European institutions or agencies and the private users. A maintenance system has been established to deal with the administration and maintenance of the thesaurus. The maintenance team coordinates the work of the Maintenance Committee and is responsible for IT and technical developments and for monitoring translations.
The Maintenance Committee made up of representatives of the various EU institutions, votes on the various proposals and decides on the amendments to be made to the thesaurus.
The Steering Committee supervises the project and officially adopts each new version.
All EuroVoc users with proposals and suggestions for amendments are invited to submit a maintenance form available in the ”CONTRIBUTE” section of this website.
On 31 January 2020 EuroVoc 4.10.1 was published. It contains the modifications related to the withdrawal of the United Kingdom from the European Union. It is available for download in XML, SKOS-Core, SKOS-AP-EU (RDF), Marc-XML and Excel format. The Release notes are available for download in PDF format.
On 17 December 2019 EuroVoc 4.10 was published. It is available for download in XML, SKOS-Core, SKOS-AP-EU (RDF), Marc-XML and Excel format. The Release notes are available for download in PDF format.
member of the Committee of the Regions
European Data Protection Board
adequacy decision (EU)
EU candidate countries
EU acceding countries
The table below provides some figures about the vocabularies to which EuroVoc is aligned:
Eclas (European Commission Library)
EIGE Thesaurus and Glossary (Insitute for Gender Equality)
ESCO (European Commission)
GEMET General Multilingual Environmental Thesaurus ( EEA)
Thesoz - Thesaurus Sozialwissenschaften
Inspire feature concept dictionary (EC Joint Research Centre)
LCSH (Library of Congress Subject Headings)
MESH Medical Subject Headings
Rameau (Bibliothèque Nationale de France)
STW Thesaurus for Economics
UNBIS Thesaurus (United Nations)
In 1982, following a comparative study on the documentary languages used internally in the institutions, the European Parliament and the European Commission decided to construct a multilingual thesaurus covering the fields which were of interest to the European institutions and would comply with the relevant international standards.
The work was based on:
After an initial test carried out by the European Parliament and the Publications Office, the first edition of the EuroVoc thesaurus was published in 1984 in seven languages (Danish, Dutch, German, Greek, English, French and Italian) in two volumes: an alphabetical thesaurus and a subject-oriented thesaurus. This thesaurus immediately came into use at the European Parliament and the Publications Office.
The second edition of the thesaurus, published in 1987, was the result of a thorough revision of the first edition, based on the indexing experience gained at the European Parliament and the Publications Office. Two new language versions (Portuguese and Spanish) were added to the seven original languages. Three additional versions — the multilingual thesaurus (1987), the permuted thesaurus (1989) and the terminogrammes edition (1990) — accompanied the two original versions (alphabetic thesaurus and subject-oriented thesaurus).
The second edition of EuroVoc met with increasing interest outside the EU institutions among national parliaments, in national and regional administrative bodies and even in the private sector.
In 1989, the European Parliament and the Publications Office, respective owner and manager of the thesaurus, took the initiative to organise a first seminar for EuroVoc users. This seminar provided guidance for decisions on the management and maintenance of the thesaurus.
One of these decisions was to set up a maintenance structure which would take account of the proposals of all thesaurus users, irrespective of whether they worked in the EU institutions. An update of the second edition was published in 1990 in the form of a fascicule.
The third edition of EuroVoc was published in 1995.
Following observations from users concerning the accuracy of certain language equivalents, the different language versions of EuroVoc were revised thoroughly by a multilingual team of translators and terminologists from the European Commission.
The third edition comprises three volumes per language, instead of the five volumes of the previous edition.
After consulting the users, the permuted and alphabetical presentations of the thesaurus were merged. The new alphabetical presentation provides descriptors in their semantic context and includes references for non-descriptors.
The new subject-oriented version comprises the associative relationships present in the terminogrammes edition. The publication of terminogrammes was abandoned following the difficulties experienced when publishing the second edition and owing to the fact that users were less interested in this type of presentation.
Compared with the previous edition of the thesaurus, 566 descriptors were added and 26 deleted.
Based on this edition, EuroVoc has been used as an indexing tool for the documentary databases of the Publications Office for the production of catalogues and tables of the Official Journal.
In October 1999, the Management Committee of the Publications Office entrusted the task of preparing a new edition of the EuroVoc thesaurus to two new interinstitutional committees — the Steering Committee and the Maintenance Committee. These two committees are made up of representatives of the institutions involved in the project: the European Parliament, the Council of the European Union, the European Commission, the Court of Justice and the Court of Auditors.
The Steering Committee is responsible for directing the EuroVoc project. In particular, it determines the objectives and priorities, lays down the overall timetable, decides on the use of resources and draws up the provisional budget. The Steering Committee is the authority which officially adopts each new version. This committee meets twice a year under the chair of the European Parliament. The Publications Office takes charge of the technical organisation of these meetings and provides the committee’s secretariat.
The Maintenance Committee, made up of permanent representatives of the institutions (mainly specialist documentalists), is a key component in the system. It has the task of examining the proposals for updating the EuroVoc thesaurus, and meets twice a year. The Publications Office takes charge of the technical organisation of these meetings and coordinates the work. A standing group in the maintenance team is managed by the Publications Office.
Edition 3.1 was developed with a view to meeting the urgent indexing requirements. In addition to the terminological updates and 142 new descriptors, this edition exists in two new languages, namely Finnish and Swedish. On 22 December 2000, two Internet sites were opened on the Europa server:
The fourth edition of EuroVoc was disseminated on the Internet on 26 November 2002, following a thorough revision of the thesaurus.
In view of changes in spelling rules, corrections were made to the Dutch, English and German versions. In accordance with users’ wishes, no changes were made to the overall structure but 364 new descriptors were added, compared with Edition 3.1, and nine were deleted.
Since February 2003, the website has offered the option of downloading the subject-oriented and permuted alphabetical versions of EuroVoc in the form of PDF files, and of consulting a multilingual list. This list replaces the old multilingual printed version.
Edition 4.1 of EuroVoc was disseminated on the Internet on 13 February 2004.
EuroVoc now contains 6 501 descriptors: 63 new descriptors; 1 descriptor was deleted.
Edition 4.2 of EuroVoc was disseminated on the Internet on 15 July 2005. It incorporates the proposals submitted by the Spanish, Czech, Polish and Swedish parliaments, and by the geographical entities and regions of the enlarged Europe.
It includes 6 645 descriptors, 155 of which are new. EuroVoc is currently available in 21 official languages, plus Croatian.
Edition 4.3 of EuroVoc was disseminated on the Internet in January 2008 following a thorough revision of domains 72 (Geography) and 76 (International organizations). It contains 6797 concepts in which new regions of Bulgaria and Romania.
Edition 4.4 of EuroVoc was the result of a thorough revision among other things according to the concepts introduced by the 'Lisbon Treaty'. It includes 6 883 thesaurus concepts of which 85 concepts are new, 142 have been updated and 28 have been classified as obsolete concepts.
All the terms in Portuguese have been revised according to the Portuguese language spelling reform. The prior lexical value remains available as Non-Preferred Terms.
For historical reasons, some concepts have been defined as 'obsolete'. They remain in the thesaurus and keep their URIs. A reference to the new concepts to be used instead is indicated by the abbreviation 'USE INSTEAD'.
In the list below the concepts which have become obsolete with EuroVoc 4.4 are sorted by the Microthesauri to which they belong.
The release 4.4 introduces the compound equivalence, in which one 'Non-Preferred Terms' leads to an association of at least two 'Preferred Terms'. The association is introduced by the abbreviation 'USE+'.
In the details of the 'Preferred Term', the 'Non Preferred Term' is presented by the abbreviation 'UF+'.
Since Edition 4.4, the EuroVoc thesaurus has been published and made available over the Internet in 23 EU languages (Bulgarian, Croatian, Czech, Danish, Dutch, English, Estonian, Finnish, French, German, Greek, Hungarian, Italian, Latvian, Lithuanian, Maltese, Polish, Portuguese, Romanian, Slovak, Slovenian, Spanish and Swedish), plus Albanian, language of FYROM and Serbian.
All language versions have the same status: each preferred term (descriptor) in one language necessarily matches a preferred term in each of the other languages. A thesaurus concept includes by default all the language equivalent of the preferred term.
However, equivalence between the non-preferred terms (non-descriptors) in the different languages does not necessarily exist. Each language is characterised by its wealth and its semantic and cultural differences, and a concept in one language is not always represented in another language.
Release notes 4.4
Release notes 4.5
Release notes 4.6
Release notes 4.7
Release notes 4.8
Release notes 4.9
Release notes 4.9.1
Release notes 4.10
Release notes 4.10.1
The EuroVoc thesaurus "eu:EuroVoc" is defined as an instance of “eu:Thesaurus”, itself a subclass of the SKOS "Concept Scheme" class.
To this "eu:EuroVoc" instance is attached the list of languages for which the EuroVoc thesaurus concepts have a preferred label (via skos:prefLabel or xl:literalForm). The thesaurus supported languages are defined by values of the property "eu:supportedLanguage". Values of this property are instances of the class "eu:Language".
The attribute rdfs:label is used to represent the language independent name of the Thesaurus (EuroVoc).
The "eu:Language" class is a convenience class where each instance represents a language. The language names of the instances are represented using "rdfs:label" in each of the supported languages. The property value "eu:language" contains the ISO 639-2 Language Code (conforming the xsd:language value space).
By convention, the URI of an "eu:Language" instance is the registered public subject indicator
All the EuroVoc publication features are exported from the Thesaurus Maintenance System (back-office) export instance (eu:Export). Properties of eu:Export are:
The publishing details of the EuroVoc thesaurus are expressed using two data properties:
EuroVoc is split into 21 domains and 127 microthesauri. Each domain is divided into a number of microthesauri. A microthesaurus is considered as a concept scheme with a subset of the concepts that are part of the complete EuroVoc thesaurus.
As SKOS has no rules to define a hierarchy of concept schemes, eu:Domain and eu:Microthesaurus has been modelled as a EuroVoc specific classes.
Domains are represented as instances of the class "eu:Domain" and microthesauri as instances of the class "eu:MicroThesaurus", a rdfs:subClass of skos:ConceptScheme.
The property value "eu:domain" defines the supergroup relation between a microthesaurus and its domain. Any micro thesaurus belongs to exactly 1 domain.
The name of an eu:Domain uses skos:prefLabel, with one label per language.
The domain identifier is represented using the Dublin Core property "dc:identifier" and is part of the domain name. The format of the domain identifier is 2 digits.
The name of an eu:MicroThesaurus uses skos:prefLabel, with one label per language.
The microthesaurus identifier is represented by the property dc:identifier and is part of the microthesaurus name. The format of the microthesaurus identifier is 4 digits. The first 2-digits of the microthesaurus identifier correspond to the domain it belongs to.
One of the EuroVoc distinctive features is the limitation of polyhierarchy. However, polyhierarchy is accepted for domains 72 (Geography). Boolean property value "eu: hasPolyHierarchy "indicates whether the domain supports poly-hierarchy (true) or not (false).
The property "skos:hasTopconcept" defines the top thesaurus concepts in a microthesaurus, with no broader relationships.
This new model follows a concept-based approach where concepts are language-independent of terms which represent them.
"eu:ThesaurusConcept" represents the class of all EuroVoc concepts and is defined as subclass of skos:Concept.
Every EuroVoc concept is attached to at least two skos:ConceptScheme instances: the global scheme "eu:EuroVoc" and one or more instances of "eu:MicroThesaurus". This attachment is using the skos:inscheme property.
A new instance "eu:Country" has been defined as a sub-set of "eu:ThesaurusConcept”, to represent a country, identified by the language-independent property value "eu:isoCountryCode" (2-char ISO country code). This property is used in domain 72 (geography).
The property "eu:useInstead" is applied to EuroVoc obsolete concepts and provides the value (cross-reference) to the new concept in use.
The "skos:broader" property is attached to "eu:ThesaurusConcept" to assign the value (URI) of the broader thesaurus concepts. By convention, "skos:broader" is used to assert a hierarchical link between a thesaurus concept and its direct parent(s), like the standard thesaurus construction BT.
The "skos:related" property of "eu:ThesaurusConcept" sets up an associative relationship to a concept associated semantically.
Instances of the class "eu:ThesaurusTerm" represent the lexical representations (terms) of concepts in a given language. This class is a subclass of the xl:Label class.
All EuroVoc label relations are expressed between thesaurus terms and are language-dependent.
Instances of "eu:ThesaurusConcept" are linked to the corresponding terms, using the properties "xl:prefLabel" for preferred terms and "xl:altLabel" for non-preferred terms.
Terms of the thesaurus may be preferred terms or non preferred terms.
"eu:permutedLiteralForm" property provides the value of the permuted literal form of a term (eu:ThesaurusTerm) .
Two classes have been settled to make the difference between the equivalence (USE/UF) and compound equivalence (USE+/UF+) relationship:
"eu:EquivalenceRelationship" is the class of equivalence relationships between a preferred term and its equivalent simple non preferred terms and may have two property values:
"eu:CompoundEquivalence" is the class of relationships between two or more preferred terms to express a compound concept.
The class properties are:
Note that no equivalent concept is created in this case at the concept level.
Specific associations have been modelled to represent the relationships between a term (full name) and its acronym or short name.
These properties are specialisations of the generic "xl:labelRelation".
EuroVoc notes are modelled using subproperties of "skos:note" defined in the standard, namely "skos:scopeNote", "skos:historyNote", and "skos:definition"
The value of property "eu:termReleasedWithVersion" provides the number of the EuroVoc release that introduced the concept.
The fixed string 'n/a' indicates the historic release number is not available.
The "eu:toBeTranslated" property is used to mark a domain, microthesaurus or thesaurus term (preferred term) with no existing translation for a individual language.
The range value is the target language for a thesaurus term or a list of target languages separated by a comma for a microthesaurus or a domain.
However, equivalence between the non preferred terms (non descriptors) in the different languages does not necessarily exist. Each language is characterised by its wealth and its semantic and cultural differences, and a concept in one language is not always represented in another language.
In all the language versions, the preferred terms and non preferred terms are generally in the singular, and the plural is used when the singular does not correspond to normal usage.
Abbreviations have been avoided as far as possible in order to make the thesaurus easier to understand and use. Only the acronyms of well-known international organisations have been taken as preferred terms. Specific associations have been modelled to represent the relationships between a term (full name) and its acronym or short name.
The alphabetical sorting of terms is not organised according to Formex 4 Characater encoding.
Initially, the EuroVoc thesaurus was compiled in accordance with the standards of the International Standards Organisation:
Since 2008, both standards have been revised, extended and combined into a new "ISO 25964 - Thesauri and Interoperability with other Vocabularies" standard in two parts:
See official ISO 25964 website.
The EurVoc ontology is an extension of SKOS (Simple Knowledge Organization System) - W3C recommendation, including appendix B, SKOS eXtension for Labels (SKOS-XL).
EuroVoc uses semantic web technologies that reflect W3C recommendations and the latest trends in thesaurus standards.
EuroVoc 4.5: EuVoc ontology and SKOS-AP-EU application profile
The EuroVoc ontology is an extension of SKOS (Simple Knowledge Organization System) - W3C recommendation, including appendix B, SKOS eXtension for Labels (SKOS-XL).
The vocabularies imported in the ontology are:
EuroVoc resources have been modelled as direct extensions of the SKOS and SKOS-XL classes and properties. Some Dublin Core properties have been reused and re-declared without importing the complete DCMI RDF schemes. (dc:identifier)
You can use the following SPARQL queries in order to obtain the results as web services. The Cellar SPARQL endpoint is available on the following address: http://publications.europa.eu/webapi/rdf/sparql
If you are interested in using other web services please do not hesitate to contact us.
Search a term (Preferred or Non Preferred Term) in EuroVoc and return the URIs of the corresponding concept(s) or concept (s) matching the found term(s).
Description of the SPARQL query
Returns the concepts that contain the search term in their pref or alt labels. Provide the search terms in between quotation marks below and the search language as a 2 letters ISO code.
Definition of the Web service Parameters
Example query: "agricultural policy"
Search a concept URI in EuroVoc and return the corresponding labels (Preferred Terms or Non Preferred Terms) matching the found URI.
Provide the target concept in between angular brakets and optionally the search language as a 2 letters ISO (of no language is needed then just leave undef keyword meaning undefined). To restrict the search to alt or pref labels only then choose to keep one, the other or both relations.
Example query: http://eurovoc.europa.eu/48
The query returns the prefLabels and altLabels of the EuroVoc concept http://eurovoc.europa.eu/48
Search a concept URI in EuroVoc and return the corresponding Related Terms (RT).
Provide the target concept in between angular brakets. If you would like to search for broader terms then use the skos:broader relation instead.
Example query: http://eurovoc.europa.eu/4450
The query returns the related terms (RT) of the EuroVoc concept http://eurovoc.europa.eu/4450