Improved word confidence estimation using long range features

Published 3 September 2001

David D. Palmer, Mari Ostendorf

DOI10.21437/eurospeech.2001-499

Citations5

TL;DR

The improved confidence estimates are shown to improve information extraction performance, specifically named entity (NE) recognition, and can then be used to further improve confidence estimation in a multi-pass NE recognition framework.

Abstract

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.

Keywords

Computer Science

BiometricsClassification and Regression Trees.

23,841 Citations1984Alexander Gordon, Leo Breiman +3 more

Statistics for biology and healthGeneralized Linear Models

20,419 Citations2005

Posterior probability decoding, confidence estimation and system combination

234 Citations2000G. Evermann, PC Woodland

The word lattices produced by the Viterbi decoder were used to generate confusion networks, which provide a compact representation of the most likely word hypotheses and their associated word posterior probabilities.

Estimating confidence using word lattices

169 Citations1997Thomas Kemp, Thomas Schaaf

In experiments on spontaneous human-to-human speech data the use of word lattice related information signi cantly improves the tagging accuracy and this work exploits theUse of such word lattices as information sources for the measure-of-con dence tagger JANKA.

Neural-network based measures of confidence for word recognition

120 Citations2002M. Weintraub, Françoise Beaufays +3 more

A novel method to combine different knowledge sources and estimate the confidence in a word hypothesis, via a neural network, is described, and a measure of the joint performance of the recognition and confidence systems is proposed.

A probabilistic approach to confidence estimation and evaluation

81 Citations2002Larry Gillick, Shosuke Ito +1 more

A novel way of estimating confidences for words that are recognized by a speech recognition system is proposed, which makes use of generalized linear models as a means for combining various predictor scores so as to arrive at confidence estimates.

Computer Speech & LanguageEvaluation of word confidence for speech recognition systems

59 Citations1999Man-Hung Siu, H. Gish

Way in which to quantify the performance of confidence measures in terms of their discrimination power and bias is discussed and two different performance metrics are analyzed: the classification equal error rate and the normalized mutual information metric.

Speech CommunicationRobust information extraction from automatically generated speech transcriptions

24 Citations2000David D. Palmer, Mari Ostendorf +1 more

A robust system for information extraction from spoken language data is described, using a state topology designed for explicit modeling of variable-length phrases and class-based statistical language model smoothing to produce state-of-the-art performance for a wide range of speech error rates.