Incremental learning of MAP context-dependent edit operations for spoken phone number recognition in an embedded platform

Error-corrective post-processing (ECPP) has great potential to reduce speech recognition errors beyond that obtained by speech model improvement. ECPP approaches aim to learn error-corrective rules to directly reduce speech recognition errors. This paper presents our investigation into one such approach, incremental learning of maximum a posteriori (MAP) context-dependent edit operations. Limiting our dataset to spoken telephone number recognition output, we have evaluated this approach in an automotive environment using an embedded speech recognizer in a mobile device. We have found that a reduction of approximately 44~49% in speech recognition string errors can be achieved after learning.