Continuously evolving symmetrical object profiles for online advertisement targeting

• US 9,721,263 B2
• Filed: 10/26/2012
• Issued: 08/01/2017
• Est. Priority Date: 10/26/2012
• Status: Active Grant

• Current Assignee: NBC Universal Media LLC (Comcast Corporation)
• Original Assignee: NBC Universal Media LLC (Comcast Corporation)
• Inventors: Fort, Pavel A
• Primary Examiner(s): Jeanty, Romain
• Application Number:
  US13/661,736
• Publication Number:
  US 20140122165A1
• Time in Patent Office: 1,740 Days
• Field of Search: 705 729
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

A system is described that implements symmetrical object profiles across one or more objects, wherein an object profile is influenced by other object profiles with which it interacts. In particular, the system includes a configuration for a first object profile that is associated with a first object, wherein the first object profile comprises a set of attributes having a first set of valuations. The system includes a configuration for a second object profile that is associated with a second object, wherein the second object profile comprises said set of attributes having a second set of valuations. The system includes a profile updater for managing the first object profile and the second object profile, wherein in a transaction involving the first and second objects, corresponding object profiles are updated based on valuations in the first and second set of valuations.

Claims

1. A system for generating updated symmetrical object profiles over a communications network, said system comprising:

a memory resident on a host computer system and configured to store a plurality of object profiles for a plurality of objects, wherein a first object profile from said plurality of object profiles comprises a consumer profile associated with a first object and comprising a first set of attributes having a first set of valuations, wherein said memory is configured to store a second object profile from said plurality of object profiles, wherein said second object profile comprises an asset profile associated with a second object and comprising a second set of attributes having a second set of valuations; and

a processor on said host computer system coupled to said memory and configured to detect a first online interaction between said first object and said second object over a communications network, wherein said processor is configured to calculate updated attribute valuations for said consumer profile and said asset profile based on said first online interaction, wherein said processor is configured to determine symmetry between said first object and a third object by comparing the updated attribute valuations for said first object profile with attribute valuations associated with a third object profile for said third object, wherein said symmetry indicates a high likelihood that said first object will interact with said third object during a second online transaction.

2. The system of claim 1, wherein said first object profile is updated based on said second set of valuations for said second object profile, and said second object profile is updated based on said first set of valuations for said first object profile.

3. The system of claim 2, wherein an attribute of said first object profile is updated by averaging attribute values of said attribute for objects in a history of transactions involving said first object over a total number of transactions.

4. The system of claim 3, wherein said history of transactions includes an initial seeding of said first object profile.

5. The system of claim 1, wherein said first and second sets of attributes comprises:

a category comprising one or more category attributes, wherein valuations for said one or more category attributes total 100 percent.

6. The system of claim 1, wherein said first object comprises a group of objects each of which is associated with a corresponding object profile, and said first object profile is based on objects profiles of said group of objects.

7. The system of claim 1, wherein said processor is configured to calculate attribute valuations for said consumer profile and said asset profile based on a history of online interactions.

8. A computer-implemented method for generating updated symmetrical object profiles over a communications network, said method comprising:

storing, in memory resident on a host computer system, a plurality of object profiles for a plurality of objects, wherein a first object profile from said plurality of object profiles is a consumer profile and is associated with a first object and comprises a first set of attributes and a second object profile from said plurality of object profiles is an asset profile and is associated with a second object and comprises a second set of attributes;

detecting a first online interaction between said first object and said second object over a communications network;

calculating updated attribute valuations for said consumer profile and said asset profile based on said first online interaction; and

determining symmetry between said first object and a third object by comparing the updated attribute valuations for said first object profile with attribute valuations associated with a third object profile for said third object, wherein said symmetry indicates a high likelihood that said first object will interact with said third object during a second online transaction.

9. The method of claim 8, further comprising:

updating a first attribute of said first object profile by averaging valuations for said first attribute associated with a history of transactions involving said first object that is based on a total number of transactions in said history of transactions.

10. The method of claim 9, further comprising:

seeding said first attribute with an initial valuation as a first transaction in said history of transactions.

11. The method of claim 8, further comprising:

configuring a group of one or more category attributes in said first and second sets, wherein valuations for said group of one or more category attributes total 100 percent for a category in any given object profile.

12. The method of claim 11, wherein valuations for said group of one or more category attributes across said category are unique between a first object and a second object.

13. The method of claim 8, further comprising:

configuring said first object as a group of one or more objects, wherein each object in said group is associated with a corresponding object profile; and

configuring said first object profile based on object profiles in said group of one or more objects.

14. The method of claim 13, wherein said configuring said first object profile further comprises:

for an attribute in said first object profile, averaging valuations of corresponding first attributes of object profiles in said group of one or more objects that is based on the number of object profiles.

15. The method of claim 8, wherein said calculating further comprises calculating valuations for said consumer profile and said asset profile based on a history of online interactions.

16. A non-transitory computer-readable medium having computer executable instructions for performing a computer-implemented method for updating symmetrical object profiles over a communications network, said method comprising:

storing, in memory resident on a host computer system, a plurality of object profiles for a plurality of objects, wherein a first object profile from said plurality of object profiles is a consumer profile and is associated with a first object and comprises a first set of attributes and a second object profile from said plurality of object profiles is an asset profile and is associated with a second object and comprises a second set of attributes;

detecting a first online interaction between said first object and said second object over a communications network;

calculating updated attribute valuations for said consumer profile and said asset profile based on said first online interaction;

determining symmetry between said first object and a third object by comparing the updated attribute valuations for said first object profile with attribute valuations associated with a third object profile for said third object, wherein said symmetry indicates a high likelihood that said first object will interact with said third object during a second online transaction.

17. The computer-readable medium of claim 16, wherein said updating each of said first and second object profiles in said method comprises:

updating a first attribute of said first object profile by averaging valuations for said first attribute associated with a history of transactions involving said first object that is based on a total number of transactions in said history of transactions.

18. The computer-readable medium of claim 17, wherein said method further comprises:

seeding said first attribute with an initial valuation as a first transaction in said history of transactions.

19. The computer-readable medium of claim 16, wherein said method further comprises:

configuring a group of one or more category attributes in said first and second sets, wherein valuations for said group of one or more category attributes total 100 percent for a category in any given object profile.

20. The computer-readable medium of claim 16, wherein said plurality of object profiles comprises vectors for said plurality of objects.