capturing hypotheses

[sniemetz]

Hello

I would love to get my hands on the matching hypotheses - I see them in the log for each match, but would like to exposes these when I post the response to my app.

This is mainly for QA purposes where I want to see if (and which) other grammar matches were considered (thus dropping the confidence for each).

Does the system provide for this? Can't see anything in documentation.

Thanks!

sven

[support]

Hi Sven,

Sorry, but our system does not provide any way to capture the hypotheses for a given utterance. Upon a grammar match, you have access to the following in the form of application variables:

application.lastresult$.confidence - Confidence level for this utterance. A value of 0.0 indicates minimum confidence, and a value of 1.0 indicates maximum confidence.

application.lastresult$.utterance - The raw string of words that were recognized for this interpretation.In the case of a DTMF grammar, this variable will contain the matched digit string.

application.lastresult$.inputmode - The mode in which user input was provided: dtmf or voice.

application.lastresult$.interpretation - Contains the interpretation, or the exact match to the active grammar.

application.lastresult$.recording - Contains the last utterance if the global property "recordutterance" is set and audio was collected.

application.lastresult$.recordingsize - Contains the last utterance recording size in bytes.

application.lastresult$.recordingduration - Contains the last utterance duration size in milliseconds.

Regards,
Plum Support