Coherence Modeling for the Automated Assessment of Spontaneous Spoken Responses

This study focuses on modeling discourse coherence in the context of automated assessment of spontaneous speech from non-native speakers. Discourse coherence has always been used as a key metric in human scoring rubrics for various assessments of spoken language. However, very little research has been done to assess a speaker's coherence in automated speech scoring systems. To address this, we present a corpus of spoken responses that has been annotated for discourse coherence quality. Then, we investigate the use of several features originally developed for essays to model coherence in spoken responses. An analysis on the annotated corpus shows that the prediction accuracy for human holistic scores of an automated speech scoring system can be improved by around 10% relative after the addition of the coherence features. Further experiments indicate that a weighted FMeasure of 73% can be achieved for the automated prediction of the coherence scores.