CzeSL-SGT – a corpus of non-native speakers’ Czech with automatic annotation

The CzeSL-SGT corpus (Czech as a Second Language with Spelling, Grammar and Tags) includes transcriptions of essays written by non-native speakers of Czech, extending the “foreign” (ciz) part of the CzeSL-plain corpus by texts collected in 2013.

Word forms are tagged by word class, morphological categories and base forms (lemmas). Some forms are corrected and the resulting texts are tagged again. Original and corrected forms are compared and error labels are assigned. The annotation is assigned automatically, which necessarily results in some inaccuracy and error rate.

Most texts are equipped with metadata about the author and the text.

The corpus is available either for on-line searching using the search interface of the Czech National Corpus, or for download as a whole from the LINDAT data repository.

For more about the CzeSl-SGT corpus see http://utkl.ff.cuni.cz/%7Erosen/public/2014-czesl-sgt-en.pdf.

Citing CzeSL-SGT

Šebesta, K. – Bedřichová, Z. – Šormová, K. – Štindlová, B. – Hrdlička, M. – Hrdličková, T. – Hana, J. – Petkevič, V. – Jelínek, T. – Škodová, S. – Poláčková, M. – Janeš, P. – Lundáková, K. – Skoumalová, H. – Sládek, Š. – Pierscieniak, P. – Toufarová, D. – Richter, M. – Straka, M. – Rosen, A.: CzeSL-SGT: CzeSL-SGT – a corpus of non-native speakers’ Czech with automatic annotation, version 2 from 28 Sep 2014. Ústav Českého národního korpusu FF UK, Praha 2014. Available on-line: http://www.korpus.cz