US 11,954,705 B2
Low entropy browsing history for ads quasi-personalization
Michael Kleber, Newton, MA (US); Gang Wang, Jersey City, NJ (US); Daniel Ramage, Seattle, WA (US); Charlie Harrison, Mountain View, CA (US); Josh Karlin, Mountain View, CA (US); and Moti Yung, New York, NY (US)
Assigned to Google LLC, Mountain View, CA (US)
Filed by Google LLC, Mountain View, CA (US)
Filed on Aug. 22, 2022, as Appl. No. 17/892,699.
Application 17/892,699 is a continuation of application No. 16/698,548, filed on Nov. 27, 2019, granted, now 11,423,441, issued on Aug. 23, 2022.
Application 16/698,548 is a continuation in part of application No. 16/535,912, filed on Aug. 8, 2019, granted, now 11,194,866, issued on Dec. 7, 2021.
Claims priority of provisional application 62/887,902, filed on Aug. 16, 2019.
Prior Publication US 2022/0391947 A1, Dec. 8, 2022
Prior Publication US 2023/0222542 A9, Jul. 13, 2023
Int. Cl. G06F 7/02 (2006.01); G06F 16/00 (2019.01); G06F 16/951 (2019.01); G06Q 30/0251 (2023.01); H04L 9/08 (2006.01); H04L 9/32 (2006.01)
CPC G06Q 30/0255 (2013.01) [G06F 16/951 (2019.01); H04L 9/085 (2013.01); H04L 9/32 (2013.01); H04L 2209/42 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method for anonymization to provide pseudo-personalized clustering, comprising:
aggregating, by one or more computing devices, a vector received from a client device into a matrix with vectors obtained from other client devices;
calculating, by the one or more computing devices, a dimension reduction of the matrix to obtain a dimension reduced matrix representing the aggregated vectors obtained from the client device and the other client devices;
determining, by the one or more computing devices, clusters of the dimension reduced matrix;
adjusting, by the one or more computing devices, a classifier model based on the identified clusters and singular vectors of the dimension reduced matrix;
transmitting, by the one or more computing devices and to the client device, at least some of the singular vectors and weights of the classifier model;
receiving, by the one or more computing devices and from the client device, a request for content including a cluster identifier generated by the client device using the at least some of the singular vectors, the weights of the classifier model, and features of resources accessed through a first application executing at the client device; and
transmitting, to the client device and responsive to the request, content selected using parameters of one of the clusters corresponding to the cluster identifier received in the request.