Språkteknologiens behov for leksikalsk informasjon

Abstract

The information to be included in a lexicon depends on its intended use. Automatic syntactic analysis of a large and varied text corpus requires detailed information on inter alia countability, compounds, spelling and inflectional variants, neologisms, and multiword expressions, information that cannot always be harvested from traditional dictionaries. In this article it is shown how deep syntactic analysis of a corpus can contribute to the enrichment of lexical resources.