Predictive content placement on a managed services system

• US 9,807,457 B1
• Filed: 07/22/2014
• Issued: 10/31/2017
• Est. Priority Date: 03/04/2011
• Status: Active Grant

• Current Assignee: CSC Holdings LLC (Altice USA, Inc.)
• Original Assignee: CSC Holdings LLC (Altice USA, Inc.)
• Inventors: Neill, Richard
• Primary Examiner(s): Shang, Annan
• Application Number:
  US14/338,002
• Time in Patent Office: 1,197 Days
• Field of Search: 725 34, 725 93-116, 725 12- 14, 709217-232
• CPC Class Codes: H04N 21/252 Processing of multiple end-users' preferences to derive collaborative data
• CPC Class Codes: H04N 21/44213 Monitoring of end-user related data arrangements for monitoring the users' behaviour or opinions in broadcast systems H04H60/33
• CPC Class Codes: H04N 21/4532 involving end-user characteristics, e.g. viewer profile, preferences monitoring of user activities for profile generation for accessing a video database G06F16/739; user profiles in network data switching protocols H04L67/306; processing of user preferences or user profiles in wireless networks H04W8/18
• CPC Class Codes: H04N 21/812 involving advertisement data advertising per se G06Q30/02

Abstract

A distributed stochastic learning agent analyzes viewing and/or interactive service behavior patterns of users of a managed services system. The agent may operate on embedded and/or distributed devices such as set-top boxes, portable video devices, and interactive consumer electronic devices. Content may be provided with services such as video and/or interactive applications at a future time with maximum likelihood that a subscriber will be viewing a video or utilizing an interactive service at that future time. For example, user impressions can be maximized for content such as advertisements, and content may be scheduled in real-time to maximize viewership from across all video and/or interactive services.

Claims

1. A computer-implemented method, comprising:

receiving, from a viewing device, user data associated with user activity of a first user;

generating, using the user data, an algorithm, and a plurality of operating parameters, a first estimator that is a maximum likelihood state vector, wherein the first estimator predicts future viewing activity of the first user;

receiving a stochastic recommendation for content from a server, the stochastic recommendation being based on the first estimator and a second estimator, wherein the second estimator is generated using user activity of a second user that is different from the first user; and

presenting the content on the viewing device based on the stochastic recommendation.

2. The computer-implemented method of claim 1, wherein the stochastic recommendation is associated with selective targeting of the content.

3. The computer-implemented method of claim 1, wherein the algorithm is based on one or more of:

a Bayesian network, a Hidden Markov Model (HMM), a belief network, an inference engine, a decision tree engine, a machine learning algorithm, a linear regression analysis, a statistical regression analysis, a cluster analysis, data mining, a channel behavior transition matrix, or a support vector machine (SVM).

4. The computer-implemented method of claim 1, further comprising:

deter mining performance of the algorithm that is used to generate the first estimator; and

swapping the algorithm with a second algorithm based on the performance of the algorithm.

5. The computer-implemented method of claim 4, further comprising:

generating a playlist using the stochastic recommendation; and

wherein the determining the performance of the algorithm further comprises monitoring user feedback to the playlist.

6. The computer-implemented method of claim 1, wherein the stochastic recommendation is based on an aggregation of the first estimator and the second estimator.

7. The computer-implemented method of claim 1, wherein the stochastic recommendation is for content to be displayed on a group of viewing devices, wherein the group of viewing devices includes the viewing device.

8. The computer-implemented method of claim 1, further comprising:

transmitting the first estimator to a server; and

in response to the transmitting, receiving, from the server, an application targeted to the first user, wherein the application is based, at least in part, on the first estimator aggregated with a plurality of estimators from a plurality of viewing devices.

9. The computer-implemented method of claim 8, wherein the application includes an advertisement or a playlist.

10. The computer-implemented method of claim 1, wherein the second estimator represents a summary state vector of an estimation of predicted future behavior of the second user.

11. A system, comprising:

one or more memories;

one or more processors coupled to the one or more memories; and

a client-agent stored in the one or more memories and executing on the one or more processors and configured to;

receive, from a viewing device, user data associated with user activity of a first user;

generate, using the user data, an algorithm, and a plurality of operating parameters, a first estimator that is a maximum likelihood state vector, wherein the first estimator predicts future viewing activity of the first user;

transmit the first estimator to a server for determining a stochastic recommendation for content;

receive the stochastic recommendation from the server, the stochastic recommendation being based on the first estimator and a second estimator, wherein the second estimator is generated using user activity of a second user that is different from the first user; and

present the content on the viewing device based on the stochastic recommendation.

12. The system of claim 11, wherein the stochastic recommendation is associated with selective targeting of the content.

13. The system of claim 11, wherein the algorithm is based on one or more of:

a Bayesian network, a Hidden Markov Model (HMM), a belief network, an inference engine, a decision tree engine, a machine learning algorithm, a linear regression analysis, a statistical regression analysis, a cluster analysis, data mining, a channel behavior transition matrix, or a support vector machine (SVM).

14. The system of claim 11, wherein the client-agent is further configured to:

determine performance of the algorithm that is used to generate the first estimator; and

swap the algorithm with a second algorithm based on the performance.

15. The system of claim 14, wherein the client-agent is further configured to:

generate a playlist using the stochastic recommendation; and

wherein the client-agent is further configured to monitor user feedback for the playlist to determine the performance of the algorithm.

16. The system of claim 11, wherein the stochastic recommendation is determined based on an aggregation of the first estimator and the second estimator.

17. The system of claim 16, wherein the viewing device is within a group of viewing devices;

and

wherein the stochastic recommendation is determined for the group of viewing devices based on the aggregation of the first estimator and the second estimator.

18. The system of claim 11, wherein the client-agent is further configured to:

transmit the first estimator to the server; and

in response to the transmission, receive, from the server, an application targeted to the user, wherein the application is based, at least in part, on the first estimator aggregated with a plurality of estimators from a plurality of viewing devices.

19. The system of claim 18, wherein the application includes an advertisement or a playlist.

20. The system of claim 11, wherein the second estimator represents a summary state vector of an estimation of predicted future behavior of the second, user.

21. A non-transitory computer-readable medium having instructions stored thereon, that when executed by one or more processors cause the one or more processors to perform operations, the operations comprising:

receiving, from a viewing device, user data associated with user activity of a first user;

generating, using the user data, an algorithm, and a plurality of operating parameters, a first estimator that is a maximum likelihood state vector, wherein the first estimator predicts future viewing activity of the first user;

receiving a stochastic recommendation for content from a server, the stochastic recommendation being based on the first estimator and a second estimator, wherein the second estimator is generated using user activity of a second user that is different from the first user; and

presenting the content on the viewing device based on the stochastic recommendation.

22. The non-transitory computer-readable medium of claim 21, wherein the stochastic recommendation is based on an aggregation of the first estimator and the second estimator.

23. The non-transitory computer-readable medium of claim 21, wherein the second estimator represents a summary state vector of an estimation of predicted future behavior of the second user.

24. A computer-implemented method, comprising:

receiving, from a first viewing device, a first estimator that is a first maximum likelihood state vector, wherein the first estimator predicts future viewing activity of the first user, the first estimator being generated using first user data associated with user activity of a first user, an algorithm, and a plurality of operating parameters;

receiving, from a second user device, a second estimator that is a second maximum likelihood state vector, wherein the second estimator predicts future viewing activity of the second user, the second estimator being generated using second user data associated with user activity of a second user, the algorithm, and the plurality of operating parameters;

determining, by a server, a stochastic recommendation for content based on the first estimator and the second estimator; and

presenting the content on the first viewing device based on the stochastic recommendation.

25. A non-transitory computer-readable medium having instructions stored thereon, that when executed by one or more processors cause the one or more processors to perform operations, the operations comprising:

receiving, from a first user device, a first estimator that is a first maximum likelihood state vector, wherein the first estimator predicts future viewing activity of the first user, the first estimator being generated using first user data associated with user activity of a first user, an algorithm, and a plurality of operating parameters;

receiving, from a second user device, a second estimator that is a second maximum likelihood state vector, wherein the second estimator predicts future viewing activity of the second user, the second estimator being generated using second user data associated with user activity of a second user, the algorithm, and the plurality of operating parameters;

determining, by a server, a stochastic recommendation for content based on the first estimator and the second estimator; and

presenting the content on the first user device based on the stochastic recommendation.