Tekstbaseret lemmaselektion

Abstract

Based on a reseach project at The Aarhus School ofBusiness, which includes a text corpus of about 1 800 000 words established in preparation for lexicographical processing, problems regarding the use of a text corpus for lemma selection are discussed. Among other things, the priority of a text corpus compared to that of other sources in this phase is discussed, and an efficient search method used as an alternative to traditional selection methods is suggested.