Ranking and filtering groups recommendations

Patent metadata

Field	Value
Publication number	US-10210261-B2
Application number	US-201414308536-A
Country	US
Kind code	B2
Filing date	Jun 18, 2014
Priority date	Jun 18, 2014
Publication date	Feb 19, 2019
Grant date	Feb 19, 2019

Abstract

In one embodiment, a set of user groups of a social-networking system may be accessed. A first subset of the user groups may be determined for a particular user of the social-networking system, based on one or more filtering criteria. A number of recommendation-source processes may be applied to the first subset to determine a number of second subsets of the first set. Each recommendation-source process may represent a particular recommendation source. The second subsets may be combined into a list of user groups. The list of user groups may be ranked, and sent to the particular user.

First claim

What is claimed is:

1. A method comprising:
by one or more computing devices, accessing a set of user groups of a social-networking system;
by one or more computing devices, determining a first subset of the user groups for a particular user of the social-networking system based on a first set of filtering criteria, wherein:
the first set of filtering criteria comprises one or more group-specific filters regarding characteristics of individual groups, and one or more user-specific filters associated with characteristics of the particular user; and
a particular user group is added to the first subset if the particular user group passes the one or more group-specific filters through one or more of a weak AND (WAND) operator and a strong OR (SOR) operator as weighted search constraints and passes the one or more user-specific filters, wherein to pass through one or more of the WAND and SOR operators includes at least passing a first group-specific filter or second group-specific filter by a first predetermined amount, and wherein to pass through the one or more of the WAND and SOR operators includes at most failing the first group-specific filter or the second group-specific filter by a second predetermined amount;
by one or more computing devices, applying a plurality of recommendation-source processes to the first subset to determine a plurality of second subsets of the first subset, wherein:
each of the recommendation-source processes respectively determines a particular second subset of user groups by comparing the first subset to a second set of filtering criteria that is specific to each recommendation-source process;
each of the recommendation-source processes represents a particular recommendation source; and
each of the recommendation-source processes determines a predetermined number of user groups for its respective second subset;
by one or more computing devices, combining the second subsets into a list of user groups;
by one or more computing devices, ranking the list of user groups; and
by one or more computing devices, sending the list of user groups as ranked to the particular user,
wherein sending the list of user groups further comprises sending an explanation for each user group of the list of user groups why a particular group is recommended to the particular user.

2. The method of claim 1 , wherein:
the social-networking system is associated with a plurality of users;
the social-networking system comprises a graph that comprises a plurality of nodes and a plurality of edges connecting the nodes;
some of the nodes each correspond to one of the user groups; and
some of the nodes each correspond to one of the users.

3. The method of claim 1 , wherein comparing the first subset to the second set of filtering criteria comprises comparing the first subset with:
a user group joined by friends of the particular user;
a user group joined by users sharing common membership in another user group with the particular user;
a user group joined by users determined to have a common interest with the particular user;
a user group joined by users having a location or age in common with the particular user;
a user group joined by users having profile information in common with the particular user;
a user group associated with a topic that is associated with the particular user;
a user group having a co-interaction association with another user group associated with the particular user;
a user group having a co-interaction association with a page associated with the particular user; or
a user group viewed by the particular user.

4. The method of claim 1 , wherein the one or more group-specific filters are based at least in part on:
a most recent activity for a user group;
a number of members of a user group;
a most recent writing activity for a user group;
a rate of activities in a user group; or
a membership type of a user group.

5. The method of claim 4 , wherein the rate of activities in the user group comprises:
a rate of posts in the user group over a predetermined period of time;
a rate of comments or likes per post in the user group over a predetermined period of time; or
a rate of views in the user group over a predetermined period of time.

6. The method of claim 4 , wherein the membership type of the user group comprises:
an open user group;
a closed user group;
a secret user group;
a disabled user group; or
an empty user group.

7. The method of claim 4 , wherein the most recent writing activity for a user group comprises a date of a most recent post shared to the user group by a member of the user group.

8. The method of claim 4 , wherein the most recent activity for a user group comprises:
a date of a most recent post, like, or comment in the user group; or
a predetermined number of posts, likes, or comments in the user group over a predetermined period of time.

9. The method of claim 1 , wherein at least one recommendation source is updated in response to an action on the social-networking system by the particular user, the action comprising:
a new group joined by the particular user;
a view of a group by the particular user on the online social network;
a post shared to the online social network by the particular user;
a second user being added as a friend by the particular user;
a like or comment shared by the particular user; or
a profile update for the particular user.

10. The method of claim 1 , wherein a particular user group is excluded from the first subset based on a language filter.

11. The method of claim 1 , wherein a particular user group is excluded from the first subset based on at least one previous list of user groups comprising the particular user group being sent to a user of the social-networking system.

12. The method of claim 11 , further wherein the particular user group is excluded from the first subset based on one or more actions of the user of the social-networking system in response to receiving the at least one previous list of user groups.

13. The method of claim 1 , wherein a particular user group is excluded from the first subset based on previous activities of the particular user in association with the particular user group.

14. The method of claim 1 , wherein a particular user group is excluded from the first subset based on one or more administrative-approval factors for the particular user group, wherein the administrative-approval factors comprise:
a number of users waiting to be admitted into the particular user group;
a number of group join requests by the particular user awaiting administrative action; or
a rejection rate of new members for the particular user group.

15. The method of claim 1 , wherein ranking the list of user groups is based at least in part on:
a probability that the particular user will interact with a particular user group;
a social affinity or coefficient of a particular user group with respect to the particular user;
or
a ranking score for a particular user group with respect to the particular user, the ranking score comprising one or more feature scores, wherein the feature scores are based at least in part on one or more characteristics of the user group.

16. The method of claim 15 , wherein the ranking score is calculated through a machine-learning process.

17. The method of claim 1 , wherein combining the second subsets further comprises:
de-duplicating one or more identical groups in each of the second subsets; and
requesting, in response to de-duplicating one or more identical groups, additional user groups fr

Assignees

• Facebook Inc

Inventors

• Ju Jerry
• Green Bradley
• Fiore Andrew Rocco Tresolini
• Atallah Antoine Joseph
• Sorg Jonathan Daniel

Classifications

• G06Q10/40 Primary

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

• G06F16/9535 Primary

  Search customisation based on user profiles and personalisation · CPC title

• G06Q50/01

  Physics · mapped topic

• G06F17/3053

  Physics · mapped topic

• G06F17/30867 Primary

  Physics · mapped topic