US 11,727,266 B2
Annotating customer data
Linsong Chu, White Plains, NY (US); Pranita Sharad Dewan, White Plains, NY (US); Raghu Kiran Ganti, Elmsford, NY (US); Joshua M. Rosenkranz, Westchester, NY (US); and Mudhakar Srivatsa, White Plains, NY (US)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by International Business Machines Corporation, Armonk, NY (US)
Filed on Aug. 2, 2019, as Appl. No. 16/530,046.
Prior Publication US 2021/0034964 A1, Feb. 4, 2021
Int. Cl. G06F 16/00 (2019.01); G06N 3/08 (2023.01); G06F 16/28 (2019.01)
CPC G06N 3/08 (2013.01) [G06F 16/282 (2019.01); G06F 16/285 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
summarizing touchpoints into k-hot encoding feature vectors, wherein the touchpoints are customer interactions;
mapping the feature vectors onto an embedding layer;
predicting a hierarchical data sequence using the embedding layer and the feature vectors;
extracting the feature vectors that are most influential in predicting the embedding layer based on gradient analysis of terms that influenced embedding; and
outputting the touchpoints associated with the most influential feature vectors.