Korpusbaserede relationsoplysninger og lemmatisering af flerordsforbindelser

Abstract

Of the approximately 100000 entries selected for The Danish Dictionary, not all will be one-word entries. Some of them will be subentries consisting of word groups. These subentries are described in detail as if they were ordinary entries, but we also have other and shorter ways of presenting information about phraseology and valency. Bearing these different possibilities in mind, we examine the concordances and word distribution reports generated from a text corpus of 40 million running words - and make the selection of subentries for the dictionary.