logo for DanNet

På dansk

Download DanNet

January 3, 2013: V. 2.2 of DanNet available for download from the download page.


Pedersen, B.S, S. Nimb, J. Asmussen, N. Sørensen, L. Trap-Jensen, H. Lorentzen. DanNet – the challenge of compiling a WordNet for Danish by reusing a monolingual dictionary. Not yet published in Language Resources and Evaluation, Computational Linguistics Series. 28 s.

DanNet Linguistic Specifications

The linguistic specifications for DanNet are now available in English from the download page.

What is DanNet?

DanNet is a research- and development project concerned with the development of a Danish lexical semantic wordnet; i.e. a language resource where the semantic relations between words are expressed in a formal language and thereby made usable for IT systems dealing with intelligent information handling.

What is a wordnet?

A wordnet is a collection of linguistic data where the individual senses are connected by means of the semantic relations that hold between them, such as hyponymy (e.g. cup — tea cup) and meronymy relations (e.g. handle — tea cup), as well as purpose relations (such as container — contain). Through these relations, a complex semantic network is established as is already seen for American English in Princeton WordNet and for several European languages in EuroWordNet.

Illustration: Diagram of a wordnet

How can DanNet be used?

The most obvious application of DanNet regards different kinds of intelligent navigation in Danish textual material, such as content-based querying, intelligent information retrieval and data mining. With the immense amounts of text material available in digital form today, flexible and intelligent navigation facilities become indispensable. To this end, navigation tools are getting more and more refined, and conceptually organised vocabularies such as wordnets are constituting a valuable background resource for this development. It is important that such new tools are also supported for Danish.

What does DanNet build on?

The project builds partly on the expertise of the two research and development groups Det Danske Sprog- og Litteraturselskab (DSL) and Center for Sprogteknologi (CST), University of Copenhagen, partly on two newly completed data collections for Danish: Den Danske Ordbog and Den danske sprogteknologiske ordbase STO/SIMPLE.

Who finances DanNet?

The project is supported by Statens Humanistiske Forskningsråd for the period 2005-2008.

Development of the resource continues in the years 2009-2011 in the DK-CLARIN project which is finansed by "Forskningsrådet for Kultur og Kommunikation".

Logo for Københavns Universitet DSL