Systems and methods for providing content

Patent metadata

Field	Value
Publication number	US-10699216-B2
Application number	US-201615354592-A
Country	US
Kind code	B2
Filing date	Nov 17, 2016
Priority date	Nov 17, 2016
Publication date	Jun 30, 2020
Grant date	Jun 30, 2020

Abstract

Systems, methods, and non-transitory computer-readable media can generate a set of candidate content items from a plurality of content items that are available in the social networking system for a first user. A corresponding score for each of the candidate content items can be generated based at least in part on one or more social affinity coefficients corresponding to the first user and a respective second user associated with a candidate content item, wherein a social affinity coefficient provides a quantitative measurement of the strength of a relationship between two users. A first set of content items from the set of candidate content items can be determined based at least in part on the respective scores, wherein content items in the first set are included in a content feed provided to the first user.

First claim

What is claimed is:

1. A computer-implemented method comprising:
generating, by a social networking system, a set of candidate content items from a plurality of content items that are available in the social networking system for a first user;
generating, by the social networking system, a set of training examples to train a machine learning model, wherein generating the set of training examples includes:
generating, by the social networking system, a first training example that corresponds to a content item for which the first user selected a like option, wherein the first training example includes at least a first social affinity coefficient corresponding to the first user and an author of the content item as a feature, the first social affinity coefficient being measured by the social networking system,
wherein the first training example also includes a second social affinity coefficient corresponding to the first user and the author of the content item as a feature, the second social affinity coefficient being measured by a second social networking system that is different from the social networking system;

generating, by the social networking system, using the trained machine learning model, a corresponding score for each of the candidate content items based at least in part on one or more social affinity coefficients corresponding to the first user and a respective second user associated with a candidate content item, wherein a social affinity coefficient provides a quantitative measurement of the strength of a relationship between two users; and
recommending, by the social networking system, a first set of content items from the set of candidate content items based at least in part on the respective scores, wherein content items in the first set are included in a content feed provided to the first user.

2. The computer-implemented method of claim 1 , wherein the generating the corresponding score for each of the candidate content items further comprises:
providing, by the social networking system, a set of inputs to a trained machine learning model, the set of inputs including at least information describing a content item to be scored and a first social affinity coefficient corresponding to the first user and an author of the content item as measured by the social networking system; and
obtaining, by the social networking system, a likelihood of the first user selecting an option to like the content item as output from the trained machine learning model, wherein the score for the content item is based at least in part on the likelihood.

3. The computer-implemented method of claim 2 , wherein the set of inputs also includes a second social affinity coefficient corresponding to the first user and the author of the content item as measured by a second social networking system.

4. The computer-implemented method of claim 1 , wherein the generating the corresponding score for each of the candidate content items further comprises:
training, by the social networking system, a machine learning model for predicting a likelihood of whether the first user will select an option to like a content item, wherein the score for the content item is based at least in part on the likelihood.

5. The computer-implemented method of claim 4 , wherein the training the machine learning model further comprises:
generating, by the social networking system, a set of training examples for training the model, wherein at least some of the training examples in the set include one or more social affinity coefficients that correspond to the first user and another user as features; and
causing, by the social networking system, the model to be trained using the set of training examples.

6. The computer-implemented method of claim 5 , wherein the generating the set of training examples further comprises:
generating, by the social networking system, a first training example that corresponds to a content item for which a second user being followed by the first user selected a like option, wherein the first training example includes at least a first social affinity coefficient corresponding to the first user and the second user as a feature.

7. The computer-implemented method of claim 5 , wherein the generating the set of training examples further comprises:
generating, by the social networking system, a first training example that corresponds to a content item for which a second user being followed by the first user selected a like option in both the social networking system and a second social networking system, wherein the first training example includes at least a combined social affinity coefficient corresponding to the first user and the second user as a feature.

8. The computer-implemented method of claim 7 , wherein the combined social affinity coefficient is determined based at least in part on i) a first social affinity coefficient corresponding to the first user and the second user as measured by the social networking system and ii) a second social affinity coefficient corresponding to the first user and the second user as measured by the second social networking system.

9. A system comprising:
at least one processor; and
a memory storing instructions that, when executed by the at least one processor, cause the system to perform:
generating a set of candidate content items from a plurality of content items that are available in the social networking system for a first user;
generating a set of training examples to train a machine learning model, wherein generating the set of training examples includes:
generating a first training example that corresponds to a content item for which the first user selected a like option, wherein the first training example includes at least a first social affinity coefficient corresponding to the first user and an author of the content item as a feature, the first social affinity coefficient being measured by the social networking system,
wherein the first training example also includes a second social affinity coefficient corresponding to the first user and the author of the content item as a feature, the second social affinity coefficient being measured by a second social networking system that is different from the social networking system;

generating, using the trained machine learning model, a corresponding score for each of the candidate content items based at least in part on one or more social affinity coefficients corresponding to the first user and a respective second user associated with a candidate content item, wherein a social affinity coefficient provides a quantitative measurement of the strength of a relationship between two users; and
recommending a first set of content items from the set of candidate content items based at least in part on the respective scores, wherein content items in the first set are included in a content feed provided to the first user.

10. The system of claim 9 , wherein the generating the corresponding score for each of the candidate content items further causes the system to perform:
providing a set of inputs to a trained machine learning model, the set of inputs including at least information describing a content item to be scored and a first social affinity coefficient corresponding to the first user and an author of the content item as measured by the social networking system; and
obtaining a likelihood of the first user selecting an option to like the content item as output from the trained machine learning model, wherein the score for the content item is based at least in part on the likelihood.

11. The system of claim 10 , wherein the set of inputs also includes a second social affinity coefficient corresponding

Assignees

• Facebook Inc

Inventors

• Dimson Thomas Frederick
• Gordon Taylor
• Huang Fei
• Kalis Jan
• Moore Justin T
• Backstrom Lars Seren

Classifications

• G06Q10/40

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

• H04L51/212

  using filtering or selective blocking · CPC title

• H04L51/52

  for supporting social networking services · CPC title

• H04L51/214

  using selective forwarding · CPC title

• G06Q10/42

  Determination of affinities or common interests between users · CPC title