EUROSPEECH 2001 Scandinavia
7th European Conference on Speech Communication and Technology

Aalborg, Denmark
September 3-7, 2001


Improved Word Confidence Estimation using Long Range Features

David D. Palmer (1), Mari Ostendorf (2)

(1) The MITRE Corporation, USA
(2) University of Washington, USA

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.

Full Paper

Bibliographic reference.  Palmer, David D. / Ostendorf, Mari (2001): "Improved word confidence estimation using long range features", In EUROSPEECH-2001, 2117-2120.