Providing content items based on user affinity in a social network environment

• US 9,544,382 B2
• Filed: 10/28/2015
• Issued: 01/10/2017
• Est. Priority Date: 08/11/2006
• Status: Active Grant

• Current Assignee: Meta Platforms, Inc. (f/k/a Facebook, Inc.)
• Original Assignee: Meta Platforms, Inc. (f/k/a Facebook, Inc.)
• Inventors: Bosworth, Andrew G., Cox, Chris
• Primary Examiner(s): DAILEY, THOMAS J
• Application Number:
  US14/925,770
• Publication Number:
  US 20160065681A1
• Time in Patent Office: 440 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06Q 30/0269 based on user profile or attribute
• CPC Class Codes: G06Q 30/08 Auctions
• CPC Class Codes: G06Q 50/01 Social networking
• CPC Class Codes: H04L 51/52 for supporting social networking services
• CPC Class Codes: H04L 67/306 User profiles
• CPC Class Codes: H04L 67/535 Tracking the activity of the user network monitoring arrangements H04L43/00; recording of computer activity G06F11/34
• CPC Class Codes: H04L 67/63 Routing a service request depending on the request content or context

Abstract

A system and method for measuring user affinity in a social network environment is provided. One or more activities performed by a user associated with a social network environment are monitored. A relationship associated with the one or more activities is identified. An affinity for one or more objects associated with the social network environment is then determined based on the one or more activities and the relationship.

Claims

1. A computer-implemented method comprising:

receiving a plurality of candidate content items for a viewing user, each content item associated with an action performed in a social networking system and each content item identifying a respective user who is connected to the viewing user in the social networking system;

defining, for the viewing user, a first plurality of weights for a corresponding plurality of action classes, a weight for an action class indicative of a significance to the viewing user of information about actions of that action class;

determining, by a process running on the social networking system, an affinity of the viewing user for each of the plurality of candidate content items, by;

identifying a respective action associated with the content item,

determining a first numeric rating for the respective action associated with the content item based on a respective relevance to the viewing user of the respective action,

determining an action class of the plurality of action classes for the respective action, and a corresponding first weight, from among the first plurality of weights, associated with the determined action class, and

computing the affinity for the content item based on a first weighted measure of the first numeric rating for the respective action and the first weight associated with the determined action class, the weighted measure obtained by scaling the first numeric rating by the determined first weight;

comparing the viewing user'"'"'s determined affinities across the candidate content items; and

selecting, for display to the viewing user, one or more of the candidate content items based on the comparison.

2. The method of claim 1, further comprising:

defining, for the viewing user, a second plurality of weights for a corresponding plurality of user groups, a weight for a user group indicative of a significance to the viewing user of information about users of that user group,

wherein determining the affinity of the viewing user for each of the plurality of candidate content items further comprises;

identifying a respective user associated with the content item,

determining a second numeric rating for the respective user associated with the content item based on a respective relevance to the viewing user of the respective user,

determining a user group of the plurality of user groups for the respective user, and a corresponding second weight, from among the second plurality of weights, associated with the determined user group,

computing a second weighted measure of the second numeric rating for the respective user and the second weight associated with the determined user group, the weighted measure obtained by scaling the second numeric rating by the determined second weight, and

computing the affinity for the content item by combining the first weighted measure and the second weighted measure.

3. The method of claim 1, wherein:

the determined action class comprises actions performed by social networking system users and described in the candidate content items, the action class includes user interactions with media items, user interactions with social networking content, or user interactions with other social networking system users; and

the first weight is defined based on types of actions, a number of prior accesses, a frequency of prior accesses, or a recency of prior accesses of information about actions in the determined action class.

4. The method of claim 2, wherein:

the determined user group includes social networking system users identified by the candidate content items, the user group including social networking system users sharing a connection or a predefined type of relationship with the viewing user; and

the second weight is defined based on a nature or a strength of the viewing user'"'"'s relationships with users of the determined user group.

5. The method of claim 2, wherein the first weight trends inversely with the second weight based on a relative significance to the viewing user of information about the determined action class versus information about the determined user group.

6. The method of claim 2, wherein determining the affinity of the viewing user for each of the plurality of candidate content items further comprises:

adjusting the first and second weights individually for each candidate content item; and

scaling the first numeric rating based on the adjusted first weight and scaling the second numeric rating based on the adjusted second weight.

7. The method of claim 2, further comprising:

observing the viewing user'"'"'s interactions with the selected content items; and

adjusting the first and second weights based on the observed interactions.

8. The method of claim 1, wherein:

the first numeric rating for a respective action associated with the content item is determined based on a type of action, a number of prior accesses, a frequency of prior accesses, or a recency of prior accesses of information about the respective action associated with the content item.

9. The method of claim 2, wherein:

the second numeric rating for the respective user associated with the content item is determined based on a nature or a strength of the viewing user'"'"'s relationship with the respective user identified by the content item.

10. The method of claim 2, wherein:

the first numeric rating for a respective action associated with the content item is determined based on historical interactions of the viewing user with information about the respective action, the first numeric rating adjusted based on a first decay factor, the first decay factor measured based on a time-varying trend of the historical interactions with information about the respective action over time; and

the second numeric rating for a respective user identified by the content item is determined based on historical interactions of the viewing user with the respective user identified by the content item, the second numeric rating adjusted based on a second decay factor, the second decay factor measured based on a time-varying trend of the historical interactions with the respective user identified by the content item.

11. The method of claim 2, wherein:

each of the plurality of content items corresponds to an image in a photo album shared by a user connected to the viewing user in the social networking system;

the respective action corresponds to an action graphically illustrated in the image or an action textually described in a comment or in caption associated with the image;

the respective user corresponds to a user who is displayed in the image or a user who is tagged in the image; and

the selecting of one or more of the candidate content items corresponds to selecting one or more images from the photo album for display, to the viewing user, as cover images of the photo album.

12. The method of claim 2, wherein:

each of the plurality of content items corresponds to a candidate social story about a user connected to the viewing user in the social networking system;

the respective action corresponds to an action that is textually described in the social story;

the respective user corresponds to an additional user identified in the social story as having performed the action; and

the selecting of one or more of the candidate content items corresponds to selecting one or more candidate social stories about the user connected to the viewing user, for display in a newsfeed to the viewing user.

13. The method of claim 2, wherein:

each of the plurality of content items corresponds to an item of information about an event associated with a user connected to the viewing user in the social networking system;

the respective action corresponds to an action that is textually described in the item of event information;

the respective user corresponds to an additional user identified in the item of event information as being associated with the respective action; and

the selecting of one or more of the candidate content items corresponds to selecting one or more items of information about the event, for display in a calendar view to the viewing user.

14. A computer program product embodied on a non-transitory computer readable storage medium including instructions, which when executed, cause a processor to:

receive a plurality of candidate content items for a viewing user, each content item associated with an action performed in a social networking system and each content item identifying a respective user who is connected to the viewing user in the social networking system;

define, for the viewing user, a first plurality of weights for a corresponding plurality of action classes, a weight for an action class indicative of a significance to the viewing user of information about actions of that action class;

determine, by the process, an affinity of the viewing user for each of the plurality of candidate content items, by;

identifying a respective action associated with the content item,

determining a first numeric rating for the respective action associated with the content item based on a respective relevance to the viewing user of the respective action,

determining an action class of the plurality of action classes for the respective action, and a corresponding first weight, from among the first plurality of weights, associated with the determined action class, and

computing the affinity for the content item based on a first weighted measure of the first numeric rating for the respective action and the first weight associated with the determined action class, the weighted measure obtained by scaling the first numeric rating by the determined first weight;

compare the viewing user'"'"'s determined affinities across the candidate content items; and

select, for display to the viewing user, one or more of the candidate content items based on the comparison.

15. The computer program product of claim 14, further comprising instructions that cause the processor to:

define, for the viewing user, a second plurality of weights for a corresponding plurality of user groups, a weight for a user group indicative of a significance to the viewing user of information about users of that user group,

wherein the instructions that cause the processor to determine the affinity of the viewing user for each of the plurality of candidate content items further comprise instructions that cause the processor to;

identify a respective user associated with the content item,

determine a second numeric rating for the respective user associated with the content item based on a respective relevance to the viewing user of the respective user,

determine a user group of the plurality of user groups for the respective user, and a corresponding second weight, from among the second plurality of weights, associated with the determined user group,

compute a second weighted measure of the second numeric rating for the respective user and the second weight associated with the determined user group, the weighted measure obtained by scaling the second numeric rating by the determined second weight, and

compute the affinity for the content item by combining the first weighted measure and the second weighted measure.

16. The computer program product of claim 14, wherein:

the determined action class comprises actions performed by social networking system users and described in the candidate content items, the action class includes user interactions with media items, user interactions with social networking content, or user interactions with other social networking system users; and

the first weight is defined based on types of actions, a number of prior accesses, a frequency of prior accesses, or a recency of prior accesses of information about actions in the determined action class.

17. The computer program product of claim 15, wherein:

the determined user group includes social networking system users identified by the candidate content items, the user group including social networking system users sharing a connection or a predefined type of relationship with the viewing user; and

the second weight is defined based on a nature or a strength of the viewing user'"'"'s relationships with users of the determined user group.

18. The computer program product of claim 15, wherein the first weight trends inversely with the second weight based on a relative significance to the viewing user of information about the determined action class versus information about the determined user group.

19. The computer program product of claim 15, further comprising instructions that cause the processor to:

observe the viewing user'"'"'s interactions with the selected content items; and

adjust the first and second weights based on the observed interactions.

20. The computer program product of claim 15, wherein:

the first numeric rating for a respective action associated with the content item is determined based on a type of action, a number of prior accesses, a frequency of prior accesses, or a recency of prior accesses of information about the respective action associated with the content item; and

the second numeric rating for the respective user associated with the content item is determined based on a nature or a strength of the viewing user'"'"'s relationship with the respective user identified by the content item.