Inferring user profile attributes from social information

• US 8,856,235 B2
• Filed: 07/18/2013
• Issued: 10/07/2014
• Est. Priority Date: 10/29/2010
• Status: Active Grant

• Current Assignee: Meta Platforms, Inc. (f/k/a Facebook, Inc.)
• Original Assignee: Meta Platforms, Inc. (f/k/a Facebook, Inc.)
• Inventors: Zhou, Ding, Moreels, Pierre
• Primary Examiner(s): Dalencourt, Yves
• Application Number:
  US13/945,615
• Publication Number:
  US 20130304658A1
• Time in Patent Office: 446 Days
• Field of Search: 709/203, 709/204, 705/14.66
• US Class Current: 709/204
• CPC Class Codes: G06N 5/04 Inference or reasoning models
• CPC Class Codes: G06Q 10/06398 Performance of employee with respect to a job function
• CPC Class Codes: G06Q 10/10 Office automation; Time management
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 50/01 Social networking

Abstract

User profile information for a user of a social networking system is inferred based on information about user profile of the user'"'"'s connections in the social networking system. The inferred user profile attributes may include age, gender, education, affiliations, location, and the like. To infer a value of a user profile attribute, the system may determine an aggregate value based on the attributes of the user'"'"'s connections. A confidence score may also be associated with the inferred attribute value. The set of connections analyzed to infer a user profile attribute may depend on the attribute, the types of connections, and the interactions between the user and the connections. The inferred attribute values may be used to update the user'"'"'s profile and to determine information relevant to the user to be presented to the user (e.g., targeting advertisements to the user based on the user'"'"'s inferred attributes).

Claims

1. A computer-implemented method comprising:

maintaining a user profile for each of a plurality of users of a social networking system, each user profile comprising a set of attributes;

selecting a user from the plurality of users;

identifying, by a computer, a plurality of interactions between the selected user and one or more of a set of users who are connected to the selected user in the social networking system;

inferring, by the computer, a value of an attribute of the user profile for the selected user based on one or more of;

topic analysis of the one or more identified interactions and sentiment analysis associated with one or more topics identified in the one or more identified interactions;

determining relevant information for the selected user based on the inferred user profile attribute; and

sending the relevant information to the selected user.

2. The computer-implemented method of claim 1, wherein inferring the value of the attribute of the user profile for the selected user comprises:

determining a frequency of occurrence of each value of the attribute for the set of users; and

selecting a value with a highest frequency of occurrence as the inferred value of the attribute for the selected user.

3. The computer-implemented method of claim 1, wherein inferring the value of the attribute of the user profile for the selected user comprises determining the inferred value of the attribute based on frequency of occurrence of values of the attribute of the set of users.

4. The computer-implemented method of claim 1, wherein inferring the value of the attribute of the user profile for the selected user comprises determining the inferred value of the attribute to be an average value of the attribute of the set of users.

5. The computer-implemented method of claim 1, wherein inferring the value of the attribute of the user profile for the selected user comprises determining the inferred value of the attribute to be a median value of the attribute of the set of users.

6. The computer-implemented method of claim 1, wherein inferring the value of the attribute of the user profile for the selected user comprises determining an aggregate value of the attribute by one or more of:

weighting the attribute for each user of the set of users based on a metric describing a closeness of the user with the selected user or weighting the attribute for each user of the set of users based on a type value of a connection between the user and the selected user.

7. The computer-implemented method of claim 1, wherein the set of users is determined to be a subset of users connected to the selected user, the subset determined based on a metric describing a closeness of the user with the selected user.

8. The computer-implemented method of claim 1, wherein the set of users is determined to be a subset of users connected to the selected user, the subset determined based on a type value of a connection between the user and the selected user.

9. The computer-implemented method of claim 1, wherein the set of users connected to the user comprises a cluster of users selected by performing cluster analysis based on the values of the attribute being inferred.

10. The computer-implemented method of claim 1, wherein the set of users connected to the user is determined by excluding users determined to be related to the selected user as a family member.

11. The computer-implemented method of claim 1, wherein the set of users is determined based on the attribute being inferred for the selected user.

12. The computer-implemented method of claim 1, wherein the relevant information for the user comprises an advertisement.

13. The computer-implemented method of claim 1, wherein the relevant information for the user comprises a recommendation to establish a connection with another user in the social networking system.

14. The computer-implemented method of claim 1, wherein the inferred attribute is a location of the selected user, inferred by one or more of:

determining a location value with the highest frequency of occurrence in the set of users or determining the location value based on sessions created by the selected user with the social networking system determining the location value based on internet protocol (IP) address of a client device used by the selected user to create sessions with the social networking system.

15. The computer-implemented method of claim 1, wherein the inferred attribute is a gender of the selected user, inferred by one or more of:

determining a gender value with the highest frequency of occurrence in the set of users or determining the gender value with the highest frequency of occurrence in a photo in which the selected user is tagged along with other users.

16. The computer-implemented method of claim 1, wherein the inferred attribute is an age of the selected user, and inferring the value of age of the selected user further comprises one or more of:

analyzing products associated with the user, performing cluster analysis of users connected to the selected user based on age values of the users connected to the selected user, or performing cluster analysis of users connected to the selected user based on their education.

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

determining whether the inferred value of the attribute of the selected user is consistent with at least one other user profile attribute of the user; and

responsive to determining that the inferred value is inconsistent with at least one other user profile attribute, rejecting the inferred user profile value.

18. The computer-implemented method of claim 1, wherein the inferred attribute is an age of the selected user, the method further comprising:

determining an inferred value of age of the selected user;

determining whether the inferred value of age of the selected user is below a minimum age value as determined based on information describing education of the selected user in the user profile for the selected user; and

rejecting the inferred age value if the inferred value of age of the selected user is determined to be below the minimum age value.

19. The computer-implemented method of claim 1, wherein the inferred attribute comprises one or more interests of the user.

20. The computer-implemented method of claim 1, wherein the inferred attribute is a language of the user, and the set of users comprise users connected to the selected user determined to be family members of the selected user.

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

selecting an attribute of the user profile for inferring a value responsive to the attribute missing a user provided value.

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

selecting an attribute of the user profile for inferring a value responsive to the attribute having a user provided value expected to be different from the inferred value.

21. A computer program product comprising a non-transitory computer-readable storage medium containing computer program code for:

maintaining a user profile for each of a plurality of users of a social networking system, each user profile comprising a set of attributes;

selecting a user from the plurality of users;

identifying a plurality of interactions between the selected user and one or more of a set of users who are connected to the selected user in the social networking system;

inferring a value of an attribute of the user profile for the selected user based on one or more of;

topic analysis of the one or more identified interactions and sentiment analysis associated with one or more topics identified in the one or more identified interactions;

determining relevant information for the selected user based on the inferred user profile attribute; and

sending the relevant information to the selected user.