| CPC G10L 15/22 (2013.01) [G06F 30/27 (2020.01); G10L 15/063 (2013.01); G10L 15/1815 (2013.01); G10L 15/183 (2013.01); G10L 2015/0635 (2013.01)] | 20 Claims |
|
15. A method for updating machine learning models based on sentiment of outcomes of communications within an interactive voice response system, the method comprising:
receiving, at an interactive voice response system, a plurality of communications transmitted by human callers;
responding, at the interactive voice response system, to the plurality of communications;
recording, at a hardware-processor-based voice to text transformation unit program component, the plurality of communications;
selecting, at a hardware-processor-based selector unit program component, a subset of the plurality of communications from the plurality of recorded communications, the subset of the plurality of communications comprising communications including an original outcome that includes negative sentiment expressed by human callers;
simulating, at a hardware-processor-based simulator unit program component, each communication included in the subset of the plurality of communications;
enabling, at the at a hardware-processor-based simulator unit program component, a user to swap out one or more original parameters and/or one or more original prediction models for one or more modified parameters and/or one or more modified prediction models used by the interactive voice response system to form responses to communications included in the subset of the plurality of communications;
modifying, at the at a hardware-processor-based simulator unit program component, based on the swap out of the one or more original parameters and/or one or more original prediction models, the original outcome of communications included in the subset of the plurality of communications; and
displaying, at the at a hardware-processor-based simulator unit program component, the modified outcome of communications included in the subset of the plurality of communications.
|