Determining textual content that is responsible for causing a viewing spike within a video in a digital medium environment

Patent metadata

Field	Value
Publication number	US-9774895-B2
Application number	US-201615006982-A
Country	US
Kind code	B2
Filing date	Jan 26, 2016
Priority date	Jan 26, 2016
Publication date	Sep 26, 2017
Grant date	Sep 26, 2017

Abstract

A digital medium environment is described to determine textual content that is responsible for causing a viewing spike within a video. Video analytics data associated with a video is queried. The video analytics data identifies a number of previous user viewings at various locations within the video. A viewing spike within the video is detected using the video analytics data. The viewing spike corresponds to an increase in the number of previous user viewings of the video that begins at a particular location within the video. Then, text of one or more video sources or video referral sources read by users prior to viewing the video from the particular location within the video is analyzed to identify textual content that is at least partially responsible for causing the viewing spike.

First claim

What is claimed is:

1. In a digital medium environment to determine textual content that is at least partially responsible for causing a viewing spike within a video, a method implemented by at least one computing device comprising:
querying, by the at least one computing device, video analytics data associated with the video, the video analytics data identifying for each of a plurality of locations within the video, a number of users that viewed the video at the location within the video;
detecting, by the at least one computing device and based on the video analytics data, a viewing spike within the video, the viewing spike corresponding to an increase in the number of users that viewed the video at a particular location in the video relative to a location in the video that precedes the particular location in the video;
analyzing, by the at least one computing device, text of at least one video source or video referral source read by users prior to viewing the video at the particular location within the video to identify textual content that is at least partially responsible for causing the viewing spike; and
automatically generating, by the at least one computing device, a short scene from the video that includes the particular location in the video corresponding to the viewing spike.

2. The method as described in claim 1 , wherein the textual content comprises user comments regarding the video at the video source or the video referral source.

3. The method as described in claim 1 , wherein the at least one video source comprises at least one of a web page or an application at which the video is located, and wherein the at least one video referral source comprises at least one of a web page or an application that includes a link to one of the video sources.

4. The method as described in claim 1 , wherein the identifying the textual content further comprises identifying textual content that contains a location that correlates to the particular location of the viewing spike.

5. The method as described in claim 1 , wherein the identifying the textual content further comprises identifying textual content that contains words or phrases that match corresponding words or phrases of a video transcript associated with the particular location of the viewing spike.

6. The method as described in claim 5 , wherein the analyzing further comprises:
querying user interaction data corresponding to user interactions with the at least one video source or video referral source;
determining, based at least in part on the user interaction data, at least one particular comment read by users at the video source or the video referral source prior to viewing the video from the particular location of the viewing spike; and
analyzing the at least one particular comment to determine whether the at least one particular comment is at least partially responsible for causing viewing spike.

7. The method as described in claim 6 , wherein the user interaction data comprises scroll tracking data that indicates the at least one particular comment read by users prior to viewing the video from the particular location of the viewing spike.

8. The method as described in claim 6 , wherein the user interaction data comprises eye tracking data that indicates the at least one particular comment read by users prior to viewing the video from the particular location of the viewing spike.

9. The method as described in claim 1 , further comprising extracting relevant tags from the textual content that is at least partially responsible for the viewing spike, and associating the relevant tags with the video to increase the number of viewings of the video.

10. In a digital medium environment to determine a contribution score for multiple instances of textual content that are each at least partially responsible for causing a viewing spike within a video, a method implemented by at least one computing device comprising:
detecting, by the at least one computing device, the viewing spike within the video, the viewing spike corresponding to an increase in a number of previous user viewings of the video that begin at a particular location within the video relative to a location in the video that precedes the particular location in the video;
analyzing, by the at least one computing device, text of at least one or more video source or video referral source read by users prior to viewing the video from the particular location within the video to identify multiple instances of textual content that are each at least partially responsible for causing the viewing spike;
for each particular instance of textual content, calculating by the at least one computing device a contribution score indicating the contribution of the particular instance of textual content to the viewing spike; and
extracting, by the at least one computing device, relevant tags from the textual content that is at least partially responsible for the viewing spike and associating the extracted tags with searchable data associated with the video to increase a likelihood that the video will appear in search engine results.

11. The method as described in claim 10 , wherein the calculating the contribution score for the particular instance of textual content comprises:
determining a number of viewings of the video, from the particular location of the viewing spike, that were caused by the particular instance of textual content;
determining a total number of viewings of the video from that particular location of the viewing spike; and
calculating the contribution score for the particular instance of textual content by dividing the number of viewings that were caused by the particular instance of textual content by the total number of viewings.

12. The method as described in claim 10 , further comprising generating a ranked list of the multiple instances of textual content that are at least partially responsible for causing the viewing spike, the multiple instance of the textual content ranked based at least in part on the contribution scores.

13. The method as described in claim 10 , wherein the contribution score is calculated based at least in part on emotion data that indicates emotional reactions by users after reading the particular instance of textual content.

14. In a digital medium environment to determine user comments that are at least partially responsible for causing a viewing spike within a video, a system implemented at least partially in hardware, the system comprising:
a viewing spike module implemented at least partially in hardware to query video analytics data associated with the video, the video analytics data identifying unique user viewings at various locations within the video;
an analytics analysis module to detect a viewing spike within the video using the video analytics data, the viewing spike corresponding to an increase in unique user viewings of the video that begin at a particular location within the video relative to a location in the video that precedes the particular location in the video;
a text analysis engine to analyze user comments read by users prior to viewing the video from the particular location within the video to identify at least one user comment that is at least partially responsible for causing the viewing spike; and
a communication module to extract a relevant tag from the identified at least one user comment and associate the extracted relevant tag with searchable data associated with the video to increase a likelihood that the video will appear in search engine results.

15. The system as described in claim 14 , further comprising a scoring module configur

Assignees

• Adobe Systems Inc

Inventors

• Chang Walter Wei-Tuh
• Swaminathan Viswanathan
• Soni Sachin
• Stringham Russell R
• Duggal Ashish
• Dhawan Anmol

Classifications

• H04L67/10

  in which an application is distributed across nodes in the network (software deployment G06F8/60; multiprogramming arrangements G06F9/46) · CPC title

• G06F17/30864

  Physics · mapped topic

• H04L67/02

  based on web technology, e.g. hypertext transfer protocol [HTTP] · CPC title

• H04N21/24 Primary

  Monitoring of processes or resources, e.g. monitoring of server load, available bandwidth, upstream requests {(monitoring of server performance or load G06F11/34; arrangements for observation, testing or troubleshooting for broadcast or for distribution combined with broadcast H04H20/12)} · CPC title

• H04N21/85

  Assembly of content; Generation of multimedia applications · CPC title