Ranking Objects by Social Relevance

• US 20130097180A1
• Filed: 10/18/2011
• Published: 04/18/2013
• Est. Priority Date: 10/18/2011
• Status: Active Grant

• Current Assignee: Meta Platforms, Inc. (f/k/a Facebook, Inc.)
• Original Assignee: Meta Platforms, Inc. (f/k/a Facebook, Inc.)
• Inventors: Tseng, Erick
• Granted Patent: US 8,977,611 B2
• Time in Patent Office: Days
• US Class Current: 707/748
• CPC Class Codes: G06F 16/38 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 16/9536 Search customisation based on social or collaborative filtering
• CPC Class Codes: G06F 16/955 using information identifiers, e.g. uniform resource locators [URL]

Abstract

In one embodiment, a system includes one or more computing systems that implement a social networking environment and are operable to access stored information including a plurality of nodes including a first set of user nodes that each correspond to a respective user and a second set of concept nodes that each correspond to a respective concept. The system may generate a match coefficient for the user and concept, representing the degree of relevance of a particular concept node to a particular user node.

Claims

1. A method comprising, by one or more computing systems:

accessing, in a social graph, a user node representing a particular user, the user node connected to a plurality of attribute nodes; and

for each of a plurality of object nodes in the social graph, each object node connected to a plurality of attribute nodes, calculating a match coefficient between the user node and an individual object node by accessing a set of all attribute nodes connected to both the user node and the object node.

2. The method of claim 1, calculating a match coefficient comprising:

calculating, for each attribute node found in the set, a first coefficient between the user node and the attribute node, and a second coefficient between the object node and the attribute node; and

obtaining the sum-product of all the first and second coefficients.

3. The method of claim 2, wherein the first coefficient decreases as the degree of separation between the user node and the attribute node increases, and the second coefficient decreases as the degree of separation between the object node and the attribute node increases.

4. The method of claim 2, wherein each of the second coefficients is multiplied by a feedback weight for its associated attribute node and object node pair.

5. The method of claim 4, wherein the feedback weight is calculated via a machine learning algorithm.

6. The method of claim 5, wherein the feedback weight for each attribute node and object node pair is calculated by:

monitoring all user node interaction with the object node;

for each user node interaction with the object node;

determining whether the attribute node is also connected to the user node;

as a result of a positive determination;

incrementing the feedback weight; and

as a result of a negative determination, decrementing the feedback weight.

7. The method of claim 6, wherein the type of user node interaction with the object node determines the amount by which the feedback weights are incremented or decremented.

8. The method of claim 1, wherein attribute nodes connected to the particular user node comprises affinities calculated from the social graph.

9. The method of claim 8, wherein the social graph comprises information from the particular user'"'"'s profile.

10. The method of claim 8, wherein the social graph comprises the user'"'"'s actions on a social network.

11. The method of claim 8, wherein the social graph comprises the user'"'"'s actions on Internet-connected nodes external to a social network.

12. The method of claim 8, wherein the social graph comprises common keywords in the user'"'"'s messages.

13. The method of claim 6, wherein the attribute nodes comprise metadata tags.

14. The method of claim 13, wherein the metadata tags are received from a node administrator associated with a particular object node.

15. The method of claim 14, wherein the received metadata tags are initially assigned a predetermined feedback weight.

16. The method of claim 1, further comprising promoting nodes with a match coefficient exceeding a predetermined threshold to the particular user.

17. The method of claim 1, further comprising ranking all nodes by descending match coefficient for display to the particular user.

18. The method of claim 1, wherein attribute nodes comprise concept nodes.

19. The method of claim 18, wherein object nodes comprise concept nodes.

20. The method of claim 18, wherein concept nodes are dynamically generated by the social graph.