Granted Patent B2

US 12,367,431 · App. 18/646,433 · Granted Jul 22, 2025

Inventors: Jason Brewer (Mountain View, CA); Rodrigo B. Farnham (Los Angeles, CA); David B. Lue (Santa Monica, CA); Nicholas J. Stucky-Mack (Los Angeles, CA)

Assignee: SNAP INC.

G06N20/20G06N7/01G06N20/00H04L67/14H04L67/535

Abstract

Claims (42)

1. A method comprising:

in response to receiving a content request associated with aggregated content to be presented on a device of a user, identifying a candidate content item, the candidate content item having a bid value;

automatically generating, using a machine learning model, a relevancy value for the candidate content item, the relevancy value indicating whether the candidate content item is likely to be skipped by the user, the relevancy value being associated with a select value and a skip value, the select value indicating a likelihood that the user will select the candidate content item, and the skip value indicating a likelihood that the user will bypass the candidate content item;

generating, for the candidate content item, a combined value by adjusting the bid value for the candidate content item using the relevancy value generated for the candidate content item;

automatically selecting the candidate content item from a plurality of candidate content items based on the combined value generated for the candidate content item;

automatically integrating the candidate content item into at least one placeholder area among one or more pre-selected content items as part of the aggregated content; and

causing presentation of the aggregated content on the device of the user.

2. The method of claim 1 , wherein the device of the user generates the content request, and the one or more pre-selected content items are selected prior to the generation of the content request.

3. The method of claim 1 , wherein the selection of the candidate content item from the plurality of candidate content items comprises selecting multiple candidate content items from the plurality of candidate content items based on respective combined values generated for the multiple candidate content items, the multiple candidate content items comprise a first content item and a second content item, the at least one placeholder area comprises a first placeholder area and a second placeholder area, and the method further comprises:

detecting that the combined value generated for the first content item is higher than the combined value generated for the second content item; and

in response to detecting that the combined value generated for the first content item is higher than the combined value generated for the second content item, automatically integrating the first content item into the first placeholder area and the second content item into the second placeholder area.

4. The method of claim 3 , wherein the first placeholder area is between a first pair of pre-selected content items of the one or more pre-selected content items and the second placeholder area is between a second pair of pre-selected content items of the one or more pre-selected content items.

5. The method of claim 3 , wherein the aggregated content comprises an ephemeral message content collection, and the first content item is positioned so as to appear before the second content item in the ephemeral message content collection.

6. The method of claim 1 , wherein the aggregated content comprises an ephemeral message content collection.

7. The method of claim 6 , wherein the candidate content item is integrated into the ephemeral message content collection for presentation between a first pre-selected content item of the one or more pre-selected content items and a second pre-selected content item of the one or more pre-selected content items.

8. The method of claim 1 , wherein adjusting the bid value for the candidate content item comprises automatically boosting or attenuating the bid value based on whether the relevancy value is positive or negative for the candidate content item with respect to the user.

9. The method of claim 1 , wherein the content request is generated in response to the user navigating, within an application, to a page configured to receive the aggregated content.

10. The method of claim 9 , wherein the generating of the combined value, the selection of the candidate content item, and the integration of the candidate content item are performed while the user is on the page.

11. The method of claim 1 , wherein the user is enabled to skip the candidate content item by performing a first device input action through the device, the user is enabled to select the candidate content item by performing a second device input action through the device, and the device is configured to distinguish between the first device input action and the second device input action.

12. The method of claim 11 , wherein the selection of the candidate content item from the plurality of candidate content items comprises selecting multiple candidate content items from the plurality of candidate content items based on respective combined values generated for the multiple candidate content items, the aggregated content comprises a content collection in which respective ones of the multiple candidate content items and the one or more pre-selected content items are presented in sequence and navigable by performing the first device input action or the second device input action.

13. The method of claim 11 , wherein the first device input action comprises at least one of a tap gesture or a swipe gesture.

14. The method of claim 11 , wherein the second device input action at least one of a tap gesture or a swipe gesture.

15. The method of claim 1 , wherein causing the presentation of the aggregated content on the device of the user comprises transmitting the aggregated content to the device in response to the content request.

16. The method of claim 1 , further comprising storing the aggregated content including the candidate content item.

17. The method of claim 1 , wherein the machine learning model implements a random forest scheme.

18. The method of claim 1 , wherein the machine learning model comprises an ensemble classifier.

19. A system comprising:

one or more processors of a machine; and

a memory storing instructions that, when executed by the one or more processors, cause the machine to perform operations comprising:

in response to receiving a content request associated with aggregated content to be presented on a device of a user, identifying a candidate content item, the candidate content item having a bid value;

automatically generating, using a machine learning model, a relevancy value for the candidate content item, the relevancy value indicating whether the candidate content item is likely to be skipped by the user, the relevancy value being associated with a select value and a skip value, the select value indicating a likelihood that the user will select the candidate content item, and the skip value indicating a likelihood that the user will bypass the candidate content item;

generating, for the candidate content item, a combined value by adjusting the bid value for the candidate content item using the relevancy value generated for the candidate content item;

automatically selecting the candidate content item from a plurality of candidate content items based on the combined value generated for the candidate content item;

automatically integrating the candidate content item into at least one placeholder area among one or more pre-selected content items as part of the aggregated content; and

causing presentation of the aggregated content on the device of the user.

20. A non-transitory machine-readable storage device embodying instructions that, when executed by at least one machine, cause the at least one machine to perform operations comprising:

in response to receiving a content request associated with aggregated content to be presented on a device of a user, identifying a candidate content item, the candidate content item having a bid value;

automatically generating, using a machine learning model, a relevancy value for the candidate content item, the relevancy value indicating whether the candidate content item is likely to be skipped by the user, the relevancy value being associated with a select value and a skip value, the select value indicating a likelihood that the user will select the candidate content item, and the skip value indicating a likelihood that the user will bypass the candidate content item;

generating, for the candidate content item, a combined value by adjusting the bid value for the candidate content item using the relevancy value generated for the candidate content item;

automatically selecting the candidate content item from a plurality of candidate content items based on the combined value generated for the candidate content item;

automatically integrating the candidate content item into at least one placeholder area among one or more pre-selected content items as part of the aggregated content; and

causing presentation of the aggregated content on the device of the user.

Assignments (1)

Continuity (5)

References Cited (138)