US 12,307,499 B2
Item recommendation method and device for protecting user privacy and learning system
Jingren Zhou, Bellevue, WA (US); Bolin Ding, Redmond, WA (US); and Zitao Li, Hangzhou (CN)
Assigned to Alibaba Damo (Hangzhou) Technology Co., Ltd., Hangzhou (CN)
Filed by Alibaba Damo (Hangzhou) Technology Co., Ltd., Hang Zhou (CN)
Filed on Nov. 18, 2022, as Appl. No. 17/989,683.
Claims priority of application No. 202111405912.3 (CN), filed on Nov. 24, 2021.
Prior Publication US 2023/0162262 A1, May 25, 2023
Int. Cl. G06Q 30/00 (2023.01); G06N 20/00 (2019.01); G06Q 30/0601 (2023.01)
CPC G06Q 30/0631 (2013.01) [G06N 20/00 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A method implemented by a data device, the method comprising:
obtaining a locally collected scoring matrix, the locally collected scoring matrix being used to describe a scoring situation of multiple users on multiple items;
locally training an item expression matrix and a user expression matrix corresponding to the locally collected scoring matrix, and sending the locally trained user expression matrix to a server to cause the server to aggregate user expression matrices uploaded by multiple data devices, and process and send an aggregated second user expression matrix to the data device, wherein the multiple data devices comprise the data device; and
predicting degrees of preference of the multiple users for the multiple items according to the locally trained item expression matrix and the aggregated second user expression matrix if a training cut-off condition is met.