Adaptive ranking of news feed in social networking systems

Patent metadata

Field	Value
Publication number	US-9286575-B2
Application number	US-201414286803-A
Country	US
Kind code	B2
Filing date	May 23, 2014
Priority date	Jul 29, 2011
Publication date	Mar 15, 2016
Grant date	Mar 15, 2016

Abstract

Machine learning models are used for ranking news feed stories presented to users of a social networking system. The social networking system divides its users into different sets, for example, based on demographic characteristics of the users and generates one model for each set of users. The models are periodically retrained. The news feed ranking model may rank news feeds for a user based on information describing other users connected to the user in the social networking system. Information describing other users connected to the user includes interactions of the other users with objects associated with news feed stories. These interactions include commenting on a news feed story, liking a news feed story, or retrieving information, for example, images, videos associated with a news feed story.

First claim

What is claimed is:

1. A computer-implemented method comprising:
determining, by a computer, a plurality of demographic groups of users of a social networking system, the determining comprising, for each demographic group of users, selecting as plurality of users based on demographic characteristics of the users;
for each demographic group of users, generating a model configured to rank news feed stories for presentation to users from the demographic group, the model configured to receive as input, one or more user attributes describing a viewing user and ranking newsfeed stories for the viewing user based on the one or more user attributes, the generating of the model comprising:
selecting a set of features for the demographic group based on the characteristics of the users of the demographic group;
training the model the training utilizing training sets obtained from the demographic group of users, the model comprising the selected set of features;

identifying stories for presentation to a user belonging to a demographic group;
providing one or more attributes describing the user as input to the model; and
ranking the stories identified for presentation to the user using the model and sending the stories for presentation to the user based on the ranking.

2. The method of claim 1 , wherein a first model for a first demographic group of users is trained using a first set of features and a second model for a second demographic group of users is trained using a second set of features.

3. The method of claim 1 , wherein selecting the users for a demographic group is based on demographic characteristics of other users connected to the user in the social networking system.

4. The method of claim 1 , further comprising:
for each demographic group of users, periodically retraining the model for the demographic group of users.

5. The method of claim 4 , further comprising:
determining a rate at which the model for a demographic group is periodically retrained based on interactions of users of the demographic group with the news feed stories.

6. The method of claim 4 , further comprising:
determining a rate at which the model for a demographic group is periodically retrained based on interactions of users of the demographic group with objects associated with newsfeed stories, the objects represented in the social networking system.

7. The method of claim 1 , wherein the demographic characteristics for determining a demographic group include one or more of: age of users, ethnicity of users, gender of users, and income of users.

8. The method of claim 1 , wherein the features used for the model comprise interactions of users of the demographic group with news feed stories presented to the users of the demographic group.

9. The method of claim 8 , wherein an interaction with a news feed story comprises one of selecting a link in the news feed story, commenting on the news feed story, liking the news feed story, or hiding the news feed story.

10. The method of claim 1 , wherein the features used for the model comprise interactions of users of the demographic group with objects associated with news feed stories presented to the users of the demographic group.

11. The method of claim 10 , wherein an object associated with a news feed story comprises one of an image, a video, a comment posted by a user, or a user profile.

12. A computer-implemented method comprising:
determining, by a computer, a plurality of demographic groups of users of a social networking system, the determining comprising, for each demographic group of users, selecting as plurality of users based on demographic characteristics of the users;
for each demographic group of users, generating a model configured to rank news feed stories for presentation to users from the demographic group, the model configured to receive as input, one or more user attributes describing a viewing user and ranking newsfeed stories for the viewing user based on the one or more user attributes, the generating of the model comprising:
selecting a set of features for the demographic group based on the characteristics of the users of the demographic group, wherein features of the set comprise interactions of users of the demographic group with news feed stories presented to the users;
training the model, the training utilizing training sets obtained from the demographic group of users, the model comprising the selected set of features; and
storing information describing the model.

13. The method of claim 12 , wherein a first model for a first demographic group of users is trained using a first set of features and a second model for a second demographic group of users is trained using a second set of features.

14. The method of claim 12 , wherein a user is selected for a demographic group based on demographic characteristics of other users connected to the user in the social networking system.

15. The method of claim 12 , further comprising:
for each demographic group of users, periodically retraining the model at a rate determined using the interactions of users of the demographic group with the news feed stories.

16. The method of claim 12 , wherein the demographic characteristics for determining a demographic group include one or more of: age of users, ethnicity of users, gender of users, and income of users.

17. A non-transitory computer-readable storage medium storing computer-executable code for ranking news feed stories of a social networking system, the code comprising instructions for:
determining, by a computer, a plurality of demographic groups of users of a social networking system, the determining comprising, for each demographic group of users, selecting as plurality of users based on demographic characteristics of the users;
for each demographic group of users, generating a model configured to rank news feed stories for presentation to users from the demographic group, the model configured to receive as input, one or more user attributes describing a viewing user and ranking newsfeed stories for the viewing user based on the one or more user attributes, the generating of the model comprising;
selecting a set of features for the demographic group based on the characteristics of the users of the demographic group;
training the model, the training utilizing training sets obtained from the demographic group of users, the model comprising the selected set of features;

identifying stories for presentation to a user belonging to a demographic group;
providing one or more attributes describing the user as input to the model; and
ranking the stories identified for presentation to the user using the model and sending the stories for presentation to the user based on the ranking.

18. The non-transitory computer-readable storage medium of claim 17 , wherein a first model corresponding to a first demographic group of users is trained using a first set of features and a second model based on a second demographic group of users is trained using a second set of features.

19. The non-transitory computer-readable storage medium of claim 17 , wherein the code further comprises instructions for:
for each demographic group of users, periodically retraining the model at a rate determined using interactions of users of the demographic group with the news feed stories.

20. The non-transitory computer-readable storage medium of claim 17 , wherein the demographic characteristics for determining a demographic group include one or more of: age of users, ethnicity of users,

Assignees

• Facebook Inc

Inventors

• Gubin Max
• Kao Wayne
• Vickrey David
• Maykov Alexey

Classifications

• G06Q10/40

  Business processes related to social networking or social networking services · CPC title

• G06Q10/06393

  Score-carding, benchmarking or key performance indicator [KPI] analysis · CPC title

• G06N99/005 Primary

  Physics · mapped topic

• G06F17/30867

  Physics · mapped topic

• G06Q50/01

  Physics · mapped topic