Language corpora: The Czech case

Abstract

Against background of the growing need of information, which for language used to be supplied in a rather limited way, the new solution found in language corpora and the way how this has been implemented is outlined and discussed. For the Czech language, this solution has materialized in the 100 million representative Czech National Corpus (CNC, 2000). In the following, a brief tour is offered through various stages of its build-up, characterizing both various corpora within CNC and giving some figures about proportions of various types of language represented. The last part of the contribution sets a minimal programme for further research and desiderata to be followed in general in this branch of important and international stream of modern science.