US 11,922,444 B1
Systems, methods, and media for classifying electronic voice of the customer data using emotion profiles generated for a plurality of clusters
Sihan Zha, Jersey City, NJ (US); Chaitra Vishwanatha Hegde, Mountain View, CA (US); Preethi Raghavan, Belmont, MA (US); Zhengzheng Pan, Fishers, IN (US); and Nathaniel Young, Ann Arbor, MI (US)
Assigned to FMR LLC, Boston, MA (US)
Filed by FMR LLC, Boston, MA (US)
Filed on Jan. 31, 2023, as Appl. No. 18/103,775.
Int. Cl. G06Q 30/0204 (2023.01); G06F 16/65 (2019.01)
CPC G06Q 30/0204 (2013.01) [G06F 16/65 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A method for emotion profile prediction using Voice of Customer (VoC) data, the method comprising:
obtaining a plurality of user profiles for a plurality of different users, wherein each of the plurality of user profiles includes user metadata for a different user;
forming a plurality of clusters by a processor of a computing device, wherein each of the plurality of clusters includes a subset of user profiles for a subset of users, wherein formatting the plurality of clusters is based on a determination that the subset of user profiles includes related user metadata;
generating, by the processor, an emotion profile for each of the plurality of clusters, wherein each emotion profile is generated utilizing the VoC data f rom the subset of users with a pretrained language model;
receiving, by the processor, new VoC data and new user metadata for a new user; identifying a selected cluster of the plurality of clusters based on a determination that the new user metadata for the new user correlates to selected metadata of a selected subset of user profiles for the selected cluster;
determining, by the processor and in response to identifying the selected cluster, at least one deviation between emotion attributes of the new VoC data and a selected emotion profile of the selected cluster; and
determining an action within a set of predetermined actions, wherein the action is determined based on the at least one deviation.