Method and apparatus for setting selected recognition parameters to minimize an application cost function

A computer-implemented method employing at least one hardware implemented computer processor for setting selected automatic speech recognition parameters, the method comprising:

• by the at least one hardware implemented computer processor:
• defining a data set associated with operation of a speech recognition application, the data set including:
• i) a plurality of recognition states characterizing semantic progression of a user through the speech recognition application, and
• ii) a plurality of recognition outcomes associated with each recognition state;
• defining an application cost function characterizing for a selected user interaction with the speech recognition application an estimated cost of the user interaction for each recognition outcome;
• and
• setting one or more system performance parameters indirectly related to the user interaction to values that optimize the cost of the user interaction over the plurality of recognition states, wherein the one or more system performance parameters include at least one of:
• one or more speech recognition confidence score thresholds directly related to the recognition outcomes and one or more speech detection thresholds.