Ranking of news feeds of content including consideration of specific content clicks by users

Patent metadata

Field	Value
Publication number	US-10733254-B2
Application number	US-201514964815-A
Country	US
Kind code	B2
Filing date	Dec 10, 2015
Priority date	Dec 10, 2015
Publication date	Aug 4, 2020
Grant date	Aug 4, 2020

Abstract

An online system, such as a social networking system, monitors user interactions with news feed stories of the social networking system and divides the user interactions into non-content clicks and content clicks. The non-content clicks indicate a user's interest in news feed stories based on user actions such as comments on, likes, shares, and hides the news feed stories. The content clicks indicate a user's interest in news feed stories based on user actions on different specific portions of multimedia content (e.g., videos) in the news feed stories such as playing, fast forwarding. The social networking system trains a model based on the monitored user interactions with news feed stories and uses the trained model to rank news feed stories for presentation to a user. The ranks of news feed stories for a user are determined based on a likelihood that the user would find the story interesting.

First claim

The invention claimed is:

1. A computer-implemented method comprising:
providing, for display to a user, a plurality of news feed stories in a first news feed of a social networking system, including a video story having a plurality of user interface controls that affect an interface of the video story;
receiving interactions by the user with two or more of the user interface controls within the video story;
logging each of the interactions by the user with the two or more user interface controls within the video story as distinct content clicks by the user indicating interest of the user in the video story and logging other types of interactions with the news feed stories as non-content clicks;
training, based on training data, a model configured to rank a different set of news feed stories of the social networking system for presentation to the user in a second news feed based on aggregated scores for the news feed stories, the training data for the user comprising the logged content clicks of the user interactions with the two or more user interface controls within the video story, the logged non-content clicks, and features associated with interactions by the user with news feed stories, wherein the training comprises:
assigning a weight and a probability to each feature associated with an interaction by the user with a news feed story, wherein some of the features are content clicks that are interactions with two or more user interface controls within a video story, the assigning comprising:
determining whether an interaction by the user with a news feed story is a content click interaction;
assigning a weight and a probability to the interaction by the user on a video in the news feed story;
generating a score for the interaction by the user on the video in the news feed story; and
for each subsequent user interaction on the video in the news feed story, generating a higher score for the user interaction than the score for the previous user interaction on the video in the news feed story;

generating a score for each feature associated with the interaction based on the weight and probability associated with the feature; and
aggregating scores for the features associated with the interaction to generate an aggregated score for the user interaction;

receiving a request by the user for a second news feed in the social networking system;
ranking the different set of news feed stories, including video stories, identified as candidates for presentation to the user of the social networking system using the model trained based on the logged content and non-content clicks, the ranking occurring based on the aggregated scores generated by the model such that video stories in the set that are similar to the video story are ranked higher among the news feed stories based on the content clicks than video stories in the set that are not similar to the video story; and
providing for display to the user a plurality of news feed stories of the different set in an order in the second news feed according to the ranking determined using the model trained based on the logged content and non-content clicks.

2. The method of claim 1 , wherein a distinct content click by the user within a video story captures an action of the user on a specific user interface control of the video story.

3. The method of claim 2 , wherein the video story comprises a video.

4. The method of claim 2 , wherein a distinct content click by the user within the video story indicates a level of interest of the user in the video story.

5. The method of claim 2 , wherein the distinct content click by the user comprises at least one of:
playing the video;
fast forwarding the video;
rewinding the video;
muting sound of the video;
turning on sound of the video;
expanding a display showing the video to full screen;
showing the video in high definition mode;
opening configuration settings of the video; and
hiding display of the video.

6. The method of claim 1 , further comprising receiving a plurality of non-content clicks by the user in response to the user interacting with the video story, wherein a non-content click by the user with the video story comprises at least one of:
commenting on the video story;
liking the video story;
selecting a link in the video story; and
hiding the video story.

7. The method of claim 1 , wherein the training data further comprises features associated with attributes of the users and features associated with attributes of the news feed stories.

8. The method of claim 1 , further comprising:
monitoring interactions by the user with the news feed stories presented to the user; and
determining whether to retrain the model based on the monitoring.

9. A non-transitory computer-readable storage medium storing executable computer program instructions, the computer program instructions comprising code for:
providing, for display to a user, a plurality of news feed stories in a first news feed of a social networking system, including a video story having a plurality of user interface controls that affect an interface of the video story;
receiving interactions by the user with two or more of the user interface controls within the video story;
logging each of the interactions by the user with the two or more user interface controls within the video story as distinct content clicks by the user indicating interest of the user in the video story and logging other types of interactions with the news feed stories as non-content clicks;
training, based on training data, a model configured to rank a different set of news feed stories of the social networking system for presentation to the user in a second news feed based on aggregated scores for the news feed stories, the training data for the user comprising the logged content clicks of the user interactions with the two or more user interface controls within the video story, the logged non-content clicks, and features associated with interactions by the user with news feed stories, wherein the training comprises:
assigning a weight and a probability to each feature associated with an interaction by the user with a news feed story, wherein some of the features are content clicks that are interactions with two or more user interface controls within a video story, the assigning comprising:
determining whether an interaction by the user with a news feed story is a content click interaction;
assigning a weight and a probability to the interaction by the user on a video in the news feed story;
generating a score for the interaction by the user on the video in the news feed story; and
for each subsequent user interaction on the video in the news feed story, generating a higher score for the user interaction than the score for the previous user interaction on the video in the news feed story;

generating a score for each feature associated with the interaction based on the weight and probability associated with the feature; and
aggregating scores for the features associated with the interaction to generate an aggregated score for the user interaction;

receiving a request by the user for a second news feed in the social networking system;
ranking the different set of news feed stories, including video stories, identified as candidates for presentation to the user of the social networking system using the model trained based on the logged content and non-content clicks, the ranking occurring based on the aggregated scores generated by the model such that video stories in the set that are similar to the video story are ranked higher among the news feed stories based on the content clicks than video stories in the set that are not similar to the video story; and
providing for

Assignees

• Facebook Inc

Inventors

• Marra Gregory Matthew
• Vickrey David
• Tas Mahmud Sami
• Zhuo Yue

Classifications

• G06Q10/40

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

• H04L51/214 Primary

  using selective forwarding · CPC title

• H04L51/52

  for supporting social networking services · CPC title

• H04L67/535

  Tracking the activity of the user (network monitoring arrangements H04L43/00; recording of computer activity G06F11/34) · CPC title

• G06F16/24578

  using ranking · CPC title