From Thesaurus to Framenet

Publikation: Bidrag til bog/antologi/rapportKonferencebidrag i proceedingsForskningpeer review

245 Downloads (Pure)

Abstract

High-quality semantic data from a Danish thesaurus linked with valency information from a
Danish dictionary allows us to compile a frame lexicon (Berkeley FrameNet style) for Danish
in a very efficient way. In the paper we present the thesaurus as well as the dictionary and
argue that they both represent valuable background information for assigning semantic frames
to the Danish vocabulary. The resulting partial frame lexicon is tested in an annotation task
where the semantic role inventory from English is directly transferred and made available for
annotations of Danish. While simply aiming at reaching the highest possible frame coverage of
the Danish vocabulary by reusing existing English frame and role inventories, we discuss the
advantages and the drawbacks of the proposed method. The gained experiences from the work
will be considered when scaling up the framenet resource to cover all verbs.
OriginalsprogEngelsk
TitelElectronic Lexicography in the 21st Century : Proceedings of eLex 2017 conference
RedaktørerIztok Kosem, Carole Tiberius, Miloš Jakubíček, Jeæena Kallas, Simon Krek, Vit Baisa
Antal sider23
ForlagLexical Computing CZ
Publikationsdato2017
Sider1-22
StatusUdgivet - 2017
BegivenhedeLex 2017 - electronic Lexicography in the 21st Century : Lexicography from Scratch - Leiden, Holland
Varighed: 18 sep. 201721 sep. 2017

Konference

KonferenceeLex 2017 - electronic Lexicography in the 21st Century
Land/OmrådeHolland
ByLeiden
Periode18/09/201721/09/2017

Citationsformater