N04-1028 this work with state-of-the-art confidence annotation methods . 6 Acknowledgments The
W07-0305 extracted from the utterance . Confidence annotation . From Phoenix , the set of parsed
W03-2706 structures allow us to express confidence annotation , or whether a value of a feature
W07-0305 hypotheses is passed to Helios , the confidence annotation component . Helios uses features
W15-4647 an approach to utterance level confidence annotation which aims at making an estimate
W11-2029 approach differs from supervised confidence annotation methods that learn a fixed confidence
P08-2055 effective dialogs , we generalize the confidence annotation to all the subtrees , the first
N10-1126 speech characteristics . The Helios confidence annotation uses a learned model based on
W15-4647 of non-understanding , such as confidence annotation ( Bohus & Rud - nicky , 2002
W07-0305 m. -RSB- ) , and then through a confidence annotation module that assigns a confidence
W07-1805 ways an unfortunate byproduct of confidence annotation , which might be circumvented
P08-2055 this research , we generalize confidence annotation algorithms to all sub parse trees
H05-1029 awareness by integrating existing confidence annotation schemes with correction detection
D12-1116 addressing the lexical gap . Using high confidence annotations for query expansion in ETLM ,
W07-0305 have developed tools for learning confidence annotation models by integrating information
W15-4647 features from turn 1-3 . Unlike confidence annotation ( Bohus & Rudnicky , 2002
W01-1619 approach . The best features for confidence annotation are concerned with the verification
W01-1618 approach . The best features for confidence annotation are concerned with the verification
W08-0102 advice . <title> Response-Based Confidence Annotation for Spoken Dialogue Systems </title>
P98-2237 lection , when fairly reliable confidence annotations from the speech recognizer are
hide detail