Identification of target audience for content delivery in social networks by quantifying semantic relations and crowdsourcing

• US 9,946,798 B2
• Filed: 06/18/2015
• Issued: 04/17/2018
• Est. Priority Date: 06/18/2015
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Bhatia, Nimesh, Hashemian, Seyyed Vahid, Le Hors, Arnaud J., Lu, Cindy H., Tran, Thai Q.
• Primary Examiner(s): Leroux, Etienne
• Application Number:
  US14/742,893
• Publication Number:
  US 20160371271A1
• Time in Patent Office: 1,034 Days
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation

Abstract

A mechanism is provided in a data processing system for content delivery. The mechanism identifies a candidate user of a social networking service. The candidate user has an associated profile including at least one concept of interest. The mechanism determines a probability that the candidate user is interested in an item of content based on a semantic similarity of the at least one concept of interest and at least one concept tag associated with the item of content using a weighted semantic graph. Responsive to the probability exceeding a probability threshold, the mechanism delivers the item of content to the candidate user. Responsive to receiving feedback comprising at least one action taken by the candidate user with respect to the item of content, the mechanism adjusts weights in the weighted semantic graph.

Claims

1. A method, in a data processing system, for content delivery, the method comprising:

identifying a candidate user of a social networking service, wherein the candidate user has an associated profile including a plurality of user-stated concepts of interest;

determining a probability that the candidate user is interested in an item of content based on a semantic similarity of the plurality of user-stated concepts of interest and a plurality of concept tags associated with the item of content using a weighted semantic graph, wherein the weighted semantic graph comprises nodes representing concepts and edges representing relationships between concepts of connected nodes and wherein each edge is weighted with a semantic similarity value, wherein determining the probability that the candidate user is interested in the item of content comprises;

responsive to determining the plurality of user-stated concepts of interest and the plurality of concept tags do not intersect, determining a path from a given user-stated concept of interest and a given concept tag in the weighted semantic graph; and

determining a weight of the path from the given user-stated concept of interest and the given concept tag; and

responsive to the probability exceeding a probability threshold, delivering the item of content to a client data processing system of the candidate user.

2. The method of claim 1, further comprising:

responsive to receiving feedback comprising an action taken by the candidate user with respect to the item of content, adjusting at least one semantic similarity value of a relationship between the given user-stated concept of interest and the given concept tag in the weighted semantic graph.

3. The method of claim 2, further comprising:

responsive to adjusting a semantic value of a given edge between two concepts in the weighted semantic graph, incrementing an adjustment count associated with the given edge.

4. The method of claim 2, wherein adjusting the at least one semantic similarity value of a relationship between the user-stated concept of interest and the concept tag in the weighted semantic graph comprises:

determining an adjustment rate, AR, for the at least one semantic similarity value based on the action; and

determining an adjustment increment for a given edge in the relationship between the given user-stated concept of interest and the given concept tag as;

5. The method of claim 4, wherein adjusting the at least one semantic similarity value of a relationship between the given user-stated concept of interest and the given concept tag in the weighted semantic graph further comprises:

determining a new semantic similarity value for the given edge as;

6. The method of claim 1, wherein the weighted semantic graph comprises a Resource Description Framework (RDF) graph.

7. The method of claim 1, wherein the plurality concepts of interest, the plurality of concept tags associated with the item of content, and concepts in the weighted semantic graph comprise Uniform Resource Identifiers (URIs).

8. The method of claim 1, wherein the path from the given user-stated concept of interest and the given concept tag comprises a plurality of edges and wherein determining the weight of the path comprises multiplying semantic similarity values of the plurality of edges.

9. The method of claim 1, wherein determining the probability that the candidate user is interested in the item of content further comprises:

responsive to identifying a plurality of paths from the given user-stated concept of interest and the given concept tag, selecting a path within the plurality of paths having a highest weight.

10. A computer program product comprising a computer readable storage medium having a computer readable program stored therein, wherein the computer readable program, when executed on a computing device, causes the computing device to:

identify a candidate user of a social networking service, wherein the candidate user has an associated profile including a plurality of user-stated concepts of interest;

determine a probability that the candidate user is interested in an item of content based on a semantic similarity of the plurality of user-stated concepts of interest and a plurality of concept tags associated with the item of content using a weighted semantic graph, wherein the weighted semantic graph comprises nodes representing concepts and edges representing relationships between concepts of connected nodes and wherein each edge is weighted with a semantic similarity value, wherein determining the probability that the candidate user is interested in the item of content comprises;

responsive to determining the plurality of user-stated concepts of interest and the plurality of concept tags do not intersect, determining a path from a given user-stated concept of interest and a given concept tag in the weighted semantic graph; and

determining a weight of the path from the given user-stated concept of interest and the given concept tag; and

responsive to the probability exceeding a probability threshold, deliver the item of content to a client data processing system of the candidate user.

11. The computer program product of claim 10, wherein the computer readable program further causes the computing device to:

responsive to receiving feedback comprising an action taken by the candidate user with respect to the item of content, adjust at least one semantic similarity value of a relationship between the given user-stated concept of interest and the given concept tag in the weighted semantic graph.

12. The computer program product of claim 11, wherein the computer readable program further causes the computing device to:

responsive to adjusting a semantic value of a given edge between two concepts in the weighted semantic graph, increment an adjustment count associated with the given edge.

13. The computer program product of claim 11, wherein adjusting the at least one semantic similarity value of a relationship between the given user-stated concept of interest and the given concept tag in the weighted semantic graph comprises:

determining an adjustment rate, AR, for the at least one semantic similarity value based on the action; and

determining an adjustment increment for a given edge in the relationship between the given user-stated concept of interest and the given concept tag as;

14. The computer program product of claim 13, wherein adjusting the at least one semantic similarity value of a relationship between the given user-stated concept of interest and the given concept tag in the weighted semantic graph further comprises:

determining a new semantic similarity value for the given edge as;

15. The computer program product of claim 10, wherein the weighted semantic graph comprises a Resource Description Framework (RDF) graph.

16. The computer program product of claim 10, wherein the plurality of concepts of interest, the plurality of concept tags associated with the item of content, and concepts in the weighted semantic graph comprise Uniform Resource Identifiers (URIs).

17. The computer program product of claim 10, wherein the path from the given user-stated concept of interest and the given concept tag comprises a plurality of edges and wherein determining the weight of the path comprises multiplying semantic similarity values of the plurality of edges.

18. The computer program product of claim 10, wherein determining the probability that the candidate user is interested in the item of content further comprises:

responsive to identifying a plurality of paths from the given user-stated concept of interest and the given concept tag, selecting a path within the plurality of paths having a highest weight.

19. An apparatus comprising:

a processor; and

a memory coupled to the processor, wherein the memory comprises instructions which, when executed by the processor, cause the processor to;

identify a candidate user of a social networking service, wherein the candidate user has an associated profile including a plurality of user-stated concepts of interest;

determine a probability that the candidate user is interested in an item of content based on a semantic similarity of the plurality of user-stated concepts of interest and a plurality of concept tags associated with the item of content using a weighted semantic graph, wherein the weighted semantic graph comprises nodes representing concepts and edges representing relationships between concepts of connected nodes and wherein each edge is weighted with a semantic similarity value, wherein determining the probability that the candidate user is interested in the item of content comprises;

responsive to determining the plurality of user-stated concepts of interest and the plurality of concept tags do not intersect, determining a path from a given user-stated concept of interest and a given concept tag in the weighted semantic graph; and

determining a weight of the path from the given user-stated concept of interest and the given concept tag; and

responsive to the probability exceeding a probability threshold, deliver the item of content to a client data processing system of the candidate user.

20. The apparatus of claim 19, wherein determining the probability that the candidate user is interested in the item of content further comprises:

responsive to identifying a plurality of paths from the given user-stated concept of interest and the given concept tag, selecting a path within the plurality of paths having a highest weight.