US 11,868,375 B2
Method, medium, and system for personalized content delivery
Amit R. Kapur, Venice, CA (US); Steven F. Pearman, Redondo Beach, CA (US); and James R. Benedetto, Hermosa Beach, CA (US)
Assigned to Yahoo Assets LLC, New York, NY (US)
Filed by Yahoo Assets LLC, Dulles, VA (US)
Filed on May 23, 2019, as Appl. No. 16/420,811.
Application 16/420,811 is a continuation of application No. 14/085,585, filed on Nov. 20, 2013, granted, now 10,346,436.
Application 14/085,585 is a continuation of application No. 12/968,251, filed on Dec. 14, 2010, granted, now 8,615,442, issued on Dec. 24, 2013.
Claims priority of provisional application 61/286,517, filed on Dec. 15, 2009.
Prior Publication US 2019/0278787 A1, Sep. 12, 2019
This patent is subject to a terminal disclaimer.
Int. Cl. G06Q 30/00 (2023.01); G06F 16/28 (2019.01); G06Q 30/0251 (2023.01)
CPC G06F 16/285 (2019.01) [G06Q 30/0269 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A computer implemented method for generating personalized content for display on a user device, the method comprising:
receiving, over a network, electronic content, the electronic content comprising clickstream data of a user from servers of one or more websites;
normalizing the electric content by extracting text for analyzing and discarding text for analyzing, the text for discarding including HTML tags, advertisements, and disclaimer links;
analyzing, by a processor, the extracted text to determine lexical features;
determining, by the processor, at least one topic of the electronic content based on the lexical features;
generating, by the processor, at least one content graph representing the at least one topic of the electronic content and a popularity of the electronic content;
generating, by the processor, an interest graph for the user, the interest graph including one or more attachment levels of the user for one or more topics, the one or more attachment levels based on the clickstream data of the user;
determining, by a processor of an interest module, future interests of the user based on comparing a velocity and an acceleration for each of the one or more topics on the interest graph for the user, the velocity calculated by finding an attachment difference between the attachment level at a first time and the attachment level at a second time and dividing the attachment difference by a time difference, the time difference being the difference between the first time of the attachment level and the second time of the attachment level, the acceleration calculated by determining a rate of change of the velocity;
computing, by the processor, an interest score for each future interest based on comparing the content graph and the future interests of the user;
generating, by a processor of a personalization module, personalized content including order ranked hyperlinks for the user, the order ranking based on the interest scores for each future interest; and
transmitting, by the processor, over the network, the generated personalized content for display on an electronic device of the user.