Om udvælgelse af ord til Den Danske Ordbog

Abstract

The Danish Dictionary is to contain approximately 100 000 entries selected from a text corpus of 40 million running words and four comprehensive existing dictionaries. A method of automatic selection was developed on the basis of two entry lists, which were "manually" selected from small intervals of the alphabet. Two similar entry lists were then selected computationally. The linguistic importance of a word was calculated on the basis of the number of corpus instances and the number of existing dictionaries in which the word occurs. This led to the formulation of some 30 criteria for the automatic selection of entries.