IP Library Granted Patent US 12,608,415
Granted Patent B2
US 12,608,415 · App. 18/220,493 · Granted Apr 21, 2026

Methods and systems for personalized screen content optimization

Inventor: Kyle Miller (Durham, NC)
Assignee: Adeia Guides Inc.
G06F16/435G06F16/24578G06F16/41G06F16/438
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,608,415
App. No.
18/220,493
Granted
Apr 21, 2026
Kind
B2
Abstract

Systems and associated methods are described for providing content recommendations. The system selects, using a multi-armed bandit solution model, a first plurality of content categories based on a reward score of each content category. The categories are displayed. When a user selects an item from the displayed categories, the system finds all categories that include the selected item, but rewards only the category with the highest score. The system selects, using the multi-armed bandit solution model, the second plurality of content categories based on the updated reward score of each content category. The categories are then displayed. The system may also repeat the steps to refine the multi-armed bandit solution model.

Claims (45)

1 . A method comprising:

selecting, by a server executing a content recommendation application, based on demographical data of a user, a first plurality of applications, from a plurality of applications, that are associated with a first plurality of categories selected by the recommendation application using a multi-armed bandit selection algorithm, from a plurality of categories, for recommendation on a device, wherein each category of the first plurality of content categories is associated with a reward score;

receiving, at the server, from the device a request to access an application from the first plurality of applications;

identifying, by the server, multiple categories, from the first plurality of categories, that include the requested application;

only increasing, by the server, reward score of one content category, from the multiple identified categories, that have the highest reward score;

tracking, by the server, the reward score of only those content categories whose reward score is increased to determine whether the reward score continues to increase over a predetermined period; and

in response to determining that the reward score does not continue to increase:

using the increased reward score, which includes any continued increases, to select a second plurality of content categories based on the increased reward score of each multiple identified categories using the multi-armed bandit selection algorithm;

selecting, based on demographical data of the user, a second plurality of applications from the second plurality of content categories; and

causing the device to generate for display, identifiers of the second plurality of applications, wherein the display of the identifiers of the second plurality of applications is generated based on instructions transmitted by the server to the device.

2 . The method of claim 1 , wherein, determining that the reward score does not continue to increase comprises, determining that the increase in reward score over the predetermined period is negative, zero, or below a predetermined positive number.

3 . The method of claim 1 , wherein determining that the reward score does not continue to increase over the predetermined prior of time comprises:

maintaining a counter to track a number of iterations of selections of the content categories whose reward score is increased; and

continuing to track the number of iterations until a predetermined number of iterations are reached.

4 . The method of claim 3 , further comprising, continuing to track the reward score after the predetermined number of iterations are reached until a determination is made that the reward score does not continue to increase.

5 . The method of claim 3 , further comprising, tracking a history of rewards and using data related to history of rewards and number of iterations in determining the second plurality of application from the second plurality of content categories.

6 . The method of claim 3 , further comprising, switching to an exploitation stage upon determining that the reward score does not continue to increase.

7 . The method of claim 6 , wherein the exploitation stage relates to selecting the second plurality of applications from the second plurality of content categories based on which content categories have the highest reward score after a predetermined number of iterations are reached.

8 . The method of claim 1 , further comprising, in response to determining that the reward score continues to increase, selecting the second plurality of applications from the second plurality of content categories based on a random technique.

9 . The method of claim 1 , further comprising:

switching from an exploration stage to an exploitation stage upon determining that the reward score does not continue to increase over the predetermined period of time.

10 . The method of claim 1 , further comprising, grouping the first plurality of applications in a group based on their association with the first plurality of categories.

11 . A system comprising:

communications circuitry configured to access a database containing content items; and

control circuitry configured to:

select, based on demographical data of a user, a first plurality of applications, from a plurality of applications, that are associated with a first plurality of categories selected by a recommendation application using a multi-armed bandit selection algorithm, from a plurality of categories, for recommendation on a device, wherein each category of the first plurality of content categories is associated with a reward score;

receive from the device a request to access an application from the first plurality of applications;

identify multiple categories, from the first plurality of categories, that include the requested application;

only increase reward score of one content category, from the multiple identified categories, that have the highest reward score;

track the reward score of only those content categories whose reward score is increased to determine whether the reward score continues to increase over a predetermined period; and

in response to determining that the reward score does not continue to increase:

use the increased reward score, which includes any continued increases, to select a second plurality of content categories based on the increased reward score of each multiple identified categories using the multi-armed bandit selection algorithm;

select, based on demographical data of the user, a second plurality of applications from the second plurality of content categories; and

cause the device to generate for display, identifiers of the second plurality of applications, wherein the display of the identifiers of the second plurality of applications is generated based on instructions transmitted by the server to the device.

12 . The system of claim 11 , further comprising, the control circuitry configured to switch from an exploration stage to an exploitation stage upon determining that the reward score does not continue to increase over the predetermined period of time.

13 . The system of claim 11 , wherein determining that the reward score does not continue to increase over the predetermined prior of time comprises, the control circuitry configured to:

maintain a counter to track a number of iterations of selections of the content categories whose reward score is increased; and

continue to track the number of iterations until a predetermined number of iterations are reached.

14 . The system of claim 13 , further comprising, the control circuitry configured to track a history of rewards and using data related to history of rewards and number of iterations in determining the second plurality of application from the second plurality of content categories.

15 . The system of claim 13 , further comprising, the control circuitry configured to continue tracking the reward score after the predetermined number of iterations are reached until a determination is made that the reward score does not continue to increase.

16 . The system of claim 11 , further comprising, the control circuitry configured to switch to an exploitation stage upon determining that the reward score does not continue to increase.

17 . The system of claim 16 , wherein the exploitation stage relates to the control circuitry configured to select the second plurality of applications from the second plurality of content categories based on which content categories have the highest reward score after the predetermined number of iterations are reached.

18 . The system of claim 11 , further comprising, in response to determining that the reward score continues to increase, the control circuitry configured to select the second plurality of applications from the second plurality of content categories based on a random technique.

19 . The system of claim 11 , wherein, determining that the reward score does not continue to increase comprises, the control circuitry configured to determine that the increase in reward score over the predetermined period is negative, zero, or below a predetermined positive number.

20 . The system of claim 19 , further comprising, the control circuitry configured to group the first plurality of applications in a group based on their association with the first plurality of categories.

Assignments (2)
CHANGE OF NAME Recorded Oct 3, 2024
From: ROVI GUIDES, INC.
To: ADEIA GUIDES INC.
Reel/Frame 069106/0231 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Aug 23, 2023
From: MILLER, KYLE
To: ROVI GUIDES, INC.
Reel/Frame 064685/0995 →
Continuity (3)
Continuation 17458994 · Aug 27, 2021
Continuation 16454834 · Jun 27, 2019
Related Publication 20230350937A1 · Nov 2, 2023
References Cited (27)
US 9224105B2 · Cornelius et al. · 2015 [cited by applicant]
US 10242381B1 · Zappella et al. · 2019 [cited by applicant]
US 10715869B1 · Miller · 2020 [cited by applicant]
US 11586965B1 · Di Benedetto et al. · 2023 [cited by applicant]
US 11709886B2 · McInerney · 2023 [cited by examiner]
US 20120016642A1 · Li et al. · 2012 [cited by applicant]
US 20140019225A1 · Guminy · 2014 [cited by examiner]
US 20140047562A1 · Stepanov et al. · 2014 [cited by applicant]
US 20140272847A1 · Grimes · 2014 [cited by examiner]
US 20140282676A1 · Joergens · 2014 [cited by examiner]
US 20150051973A1 · Li et al. · 2015 [cited by applicant]
US 20150199715A1 · Caron et al. · 2015 [cited by applicant]
US 20150319479A1 · Mishra et al. · 2015 [cited by applicant]
US 20160353145A1 · Ahlm et al. · 2016 [cited by applicant]
US 20170171580A1 · Hirsch et al. · 2017 [cited by applicant]
US 20170278114A1 · Renders · 2017 [cited by applicant]
US 20190080009A1 · Basu et al. · 2019 [cited by applicant]
US 20200051117A1 · Mitchell · 2020 [cited by examiner]
US 20200204862A1 · Miller · 2020 [cited by applicant]
US 20200226167A1 · Chen et al. · 2020 [cited by applicant]
US 20200226168A1 · Chen et al. · 2020 [cited by applicant]
US 20200234359A1 · Sarma · 2020 [cited by examiner]
US 20200409983A1 · Miller · 2020 [cited by applicant]
US 20210390129A1 · Miller · 2021 [cited by applicant]
EP 2816511A1 · 2014 [cited by examiner]
WO WO03063023A2 · 2003 [cited by examiner]
Mahmood T, Mujtaba G, Venturini A. Dynamic personalization in conversational recommender systems. Information Systems and e-Business Management. May 2014;12(2):213-38. (Year: 2014). [cited by examiner]