US 12,229,784 B2
Systems and methods for utilizing feedback data
Priyanka Bhatt, Haryana (IN); Shankara Bhargava, Santa Clara, CA (US); Akshay Kumar, Rajasthan (IN); and Neeraj Agrawal, Agra (IN)
Assigned to Walmart Apollo, LLC, Bentonville, AR (US)
Filed by Walmart Apollo, LLC, Bentonville, AR (US)
Filed on Jun. 30, 2021, as Appl. No. 17/364,798.
Prior Publication US 2023/0004988 A1, Jan. 5, 2023
Int. Cl. G06Q 10/10 (2023.01); G06N 20/00 (2019.01); G06Q 10/06 (2023.01); G06Q 30/016 (2023.01); G06Q 30/02 (2023.01); G06Q 30/06 (2023.01); G06F 40/30 (2020.01); G06Q 10/0837 (2023.01); G06Q 30/0601 (2023.01); G10L 15/22 (2006.01)
CPC G06Q 30/016 (2013.01) [G06N 20/00 (2019.01); G06F 40/30 (2020.01); G06Q 10/0837 (2013.01); G06Q 30/0613 (2013.01); G10L 15/22 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A system comprising:
one or more processors;
a set of memory resources to store a set of instructions that when executed by the one or more processors, cause the system to:
receive a first set of data, the first set of data including information indicating a first set of resolved user sessions of first customers, each of the first set of resolved user sessions having an associated summary and a corresponding agent indicated intent;
train a machine learning model based on the first set of data that includes information regarding resolved user sessions and corresponding agent indicated intents;
based on the first set of data, determine a set of utterances of the first customers and for the set of utterances a corresponding set of intents for the first customers;
receive a second set of data, the second set of data including information indicating a second set of user sessions of second customers, each of the second set of user sessions having an associated determined utterance and corresponding interaction information of a second customer;
based on the second set of data including information indicating determined utterances and corresponding interaction information of the second customers, determine whether a corresponding intent of each utterance of the set of utterances of the first customers is validated to be a correct intent or not, to generate validated intent data for the first customers based on:
determining, among the corresponding set of intents for the first customers, a first intent that was not validated based on the second set of data,
disassociating the first intent from the set of utterances; and
retrain the machine learning model based on the second set of data and the validated intent data.