Confidence measure generation for speech related searching

摘要

A method of generating a confidence measure generator is provided for use in a voice search system, the voice search system including voice search components comprising a speech recognition system, a dialog manager and a search system. The method includes selecting voice search features, from a plurality of the voice search components, to be considered by the confidence measure generator in generating a voice search confidence measure. The method includes training a model, using a computer processor, to generate the voice search confidence measure based on selected voice search features.

主权项

1. A method of generating a confidence measure generator for use in a voice search system, the voice search system including voice search components comprising a speech recognition system, a dialog manager and a search system, the method comprising:
identifying initial voice search features from the voice search components, the initial voice search features comprising a first feature from the speech recognition system indicative of accuracy in recognizing an utterance to generate a query, a second feature from the search system indicative of accuracy in searching for results based on the query, and a third feature from the dialog manager indicative of characteristics of user dialog turns conducted in obtaining the utterance; determining a significance measure for each of the initial voice search features; selecting a set of voice search features from the initial voice search features, based on the determined significance measures, to be considered by the confidence measure generator in generating a voice search confidence measure indicative of a confidence that a search result answers a request input by a user; and training a model, using a computer processor, to generate the voice search confidence measure based on the selected set of voice search features.