In this work, we present and evaluate the usage of an interactive web interface for browsing and correcting lecture transcripts. An experiment performed with potential users without transcription experience provides us with a set of example corrections. On German lecture data, user corrections greatly improve the comprehensibility of the transcripts, yet only reduce the WER to 22%. The precision of user edits is relatively low at 77% and errors in inflection, case and compounds were rarely corrected. Nevertheless, characteristic lecture data errors, such as highly specific terms, were typically corrected, providing valuable additional information.
Index Terms: speech recognition, user study, transcript correction, lectures
Cite as: Kolkhorst, H., Kilgour, K., Stüker, S., Waibel, A. (2012) Evaluation of interactive user corrections for lecture transcription. Proc. International Workshop on Spoken Language Translation (IWSLT 2012), 217-221
@inproceedings{kolkhorst12_iwslt,
author={Henrich Kolkhorst and Kevin Kilgour and Sebastian Stüker and Alex Waibel},
title={{Evaluation of interactive user corrections for lecture transcription}},
year=2012,
booktitle={Proc. International Workshop on Spoken Language Translation (IWSLT 2012)},
pages={217--221}
}