Lexikografiska aspekter på Internet som källa till informellt språkbruk

Abstract

This paper treats Internet-based corpora in a lexicographical perspective, which first of all requires a quick look at the terms normative and descriptive lexicography. However the main part of the paper presents research on the compilation of Internet-based corpora, and compares that type of corpora with the traditional kind. This includes a discussion of the notion of the representativeness of the corpus, with reference to comparison between Internet-based corpora and traditional corpora such as the BNC. It is noted that Internet-based corpora offers the possibility to capture language from registers that hitherto has been unrepresented or at least underrepresented in traditional corpora. Reference is being made to some experiences of compilation of Swedish web corpora, notably how to evaluate the differences in word frequency, when different corpora are compared.