This article presents the written Slovenian learner corpus KOST, focusing on its position among other learner corpora for other target languages. In terms of the sociolinguistic position of the target language, KOST can be compared with approximately one-tenth of more than 190 learner corpora. With its design, current size of almost 835,000 words, partially tagged language errors, and free access to data, KOST is fully comparable to these corpora and thus a useful resource for various forms of language research.
|