Automatic Detection of Syntactic Errors in an ASR-based CALL System
We present a new method, called SynPOS: Syntactic analysis using POS-tags. SynPOS is applied to a corpus of spoken human-machine interactions. The results show that language learners of Dutch often make syntactical errors, that there are many different types of syntactical errors, and that their frequencies vary a lot. This information can be used next to select errors and develop exercises for CALL systems.