Abstract | ||
---|---|---|
The reliability of the confidence score is very important in Spoken Dialog System performance. This paper describes a set of experiments with previously collected off-line data, regarding the set of features that should be used in the computation of the confidence score. Three different regression methods to weight the features were used and the results show that the incorporation of the confidence score given by the speech recognizer improves the confidence measure. |
Year | DOI | Venue |
---|---|---|
2012 | 10.1007/978-3-642-28885-2_45 | PROPOR |
Keywords | Field | DocType |
dialog system confidence score,speech recognizer,confidence score,spoken dialog system performance,incorporating asr information,different regression method,confidence measure,off-line data | Confidence score,Spoken dialog systems,Confidence measures,Regression,Spoken dialog,Computer science,Speech recognition,Artificial intelligence,Natural language processing | Conference |
Citations | PageRank | References |
2 | 0.41 | 6 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
José Lopes | 1 | 29 | 4.81 |
Maxine Eskenazi | 2 | 979 | 127.53 |
Isabel Trancoso | 3 | 906 | 113.87 |