Error classification in a spoken learner corpus of Czech native speakers

Abstract

The poster presents a system for tagging errors and spoken language specific features in the Czech subcorpus of LINDSEI (Louvain International Database of Spoken English Interlanguage). Since it is based on an error tagging system for a written learner corpus, it focuses mainly on the categories important for spoken language that should facilitate working with the tagged corpus in future.

Keywords

• learner corpus
• spoken language
• English
• error tagging