ISCA Archive Eurospeech 2001
ISCA Archive Eurospeech 2001

Improved word confidence estimation using long range features

David D. Palmer, Mari Ostendorf

This paper describes experiments in improving word confidence estimation using document- and task-level features of the hypothesized word sequence from a recognizer. The improved confidence estimates are shown to improve information extraction performance, specifically named entity (NE) recognition. The detected names can then be used to further improve confidence estimation in a multi-pass NE recognition framework.