# US 2017/0257653 A1 · Shot structure of online video as a predictor of success

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2017/0257653 A1
- Document ID: US20170257653A1
- Status: Published application
- Kind: A1
- Application number: 15/057,455
- Filed: 2016-03-01
- Published: 2017-09-07
- Earliest claimed: 2016-03-01
- Assignees: Disney Enterprises, Inc.
- Applicants: Disney Enterprises, Inc.
- Inventors: Miquel Á. FARRÉ GUIU; Seth FREY; Aljosa Aleksej Andrej SMOLIC; Michael R. CLEMENTS; Pablo Beltran SANCHIDRIAN
- Web page: https://patentgrep.com/patent/US20170257653A1

## CPC

- H04N 21/251: {Learning process for intelligent management, e.g. learning user preferences for recommending movies (details of learning user preferences for the retrieval of video data in a video database G06F16/739; computer systems using learning methods G06N3/08)}
- G06F 18/22: Matching criteria, e.g. proximity measures
- G06F 18/231: Hierarchical techniques, i.e. dividing or merging pattern sets so as to obtain a dendrogram
- G06N 5/04: Inference or reasoning models
- G06N 20/00: Machine learning
- G06V 10/7625: {Hierarchical techniques, i.e. dividing or merging patterns to obtain a tree-like representation; Dendograms}
- G06V 20/41: {Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items (segmenting video sequences G06V20/49)}
- H04N 21/23418: {involving operations for analysing video streams, e.g. detecting features or characteristics (television picture signal circuitry for scene change detection H04N5/147; filtering for image enhancement G06T5/00; methods or arrangements for recognising scenes G06V20/00; arrangements characterised by components specially adapted for monitoring, identification or recognition of video in broadcast systems H04H60/59)}
- H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

Systems, methods, and computer program products to perform an operation comprising receiving a plurality of superclusters that includes at least one of a plurality of shot clusters, wherein each of the shot clusters includes at least one of a plurality of video shots, and wherein each video shot includes one or more video frames, and computing an expected value for a metric based on the plurality of superclusters.

## Claims

### Claim 1

A method, comprising:
receiving a plurality of video shots of a video, wherein each video shot includes one or more frames of the video;
generating a hierarchy based on the plurality of video shots, wherein the hierarchy comprises a plurality of superclusters that include at least one of a plurality of shot clusters and describes relationships between the plurality of video shots, wherein each of the shot clusters includes at least one of the plurality of video shots;
computing a value for a statistic for the video based on the hierarchy; and
computing, by operation of a computer processor, an expected value for a metric based on the statistic.

### Claim 2

(canceled)

### Claim 3 (depends on 1)

The method of claim 1, wherein the statistic comprises a ratio of a count of the plurality of superclusters having less than four of the plurality of shot clusters to a total count of the plurality of superclusters, wherein the expected value for the metric is positively correlated to the ratio.

### Claim 4 (depends on 1)

The method of claim 1, wherein statistic comprises a number of iterations of a clustering algorithm applied to the plurality of video shots resulting in one of the plurality of superclusters including each of the plurality of shot clusters, wherein the expected value for the metric is negatively correlated to the number of iterations, wherein a first iteration of the clustering algorithm applies a more restrictive clustering threshold than a second iteration of the clustering algorithm.

### Claim 5 (depends on 4)

The method of claim 4, wherein the clustering algorithm comprises:
defining the clustering threshold for a current iteration;
generating the plurality of shot clusters based on the clustering threshold; and
assigning each shot cluster in a set of shot clusters, of the plurality of shot clusters, to a respective supercluster of the plurality of the plurality of superclusters, wherein each shot cluster in the set of shot clusters remains in the respective supercluster in a subsequent iteration of the clustering algorithm.

### Claim 6 (depends on 1)

The method of claim 1, further comprising prior to computing the expected value:
generating the plurality of video shots from the video;
generating the plurality of shot clusters based on similarities between the video shots, wherein the similarities between the video shots include at least one of showing a same person, showing a same place, and a same low level feature including at least one of a similar color and a similar low level geometry descriptor; and
generating the plurality of superclusters based on interactions between the plurality of shot clusters.

### Claim 7 (depends on 1)

The method of claim 1, wherein the expected value for the metric is computed responsive to a user request to return the expected value for the metric, wherein a predictive algorithm is configured to compute the expected value for the metric, wherein the metric comprises at least one of: (i) a number of views the video will receive, (ii) a number of likes the video will receive, (iii) a duration of the video a user will watch, (iv) a number of users that will watch a second video promoted by the video, (v) a likelihood that a user will click a link to an advertisement in the video, (vi) a likelihood that a user will stop watching the video at each second of the video, and (vii) an expected amount of advertising revenue the video will generate.

### Claim 8

A computer program product, comprising:
a non-transitory computer-readable storage medium having computer-readable program code embodied therewith, the computer-readable program code executable by a processor to perform an operation comprising:
receiving a plurality of video shots of a video, wherein each video shot includes one or more frames of the video;
generating a hierarchy based on the plurality of video shots, wherein the hierarchy comprises a plurality of superclusters that include at least one of a plurality of shot clusters and describes relationships between the plurality of video shots, wherein each of the shot clusters includes at least one of the plurality of video shots;
computing a value for a statistic for the video based on the hierarchy; and
computing an expected value for a metric based on the statistic.

### Claim 9

(canceled)

### Claim 10 (depends on 8)

The computer program product of claim 8, wherein the statistic comprises a ratio of a count of the plurality of superclusters having less than four of the plurality of shot clusters to a total count of the plurality of superclusters, wherein the expected value for the metric is positively correlated to the ratio.

### Claim 11 (depends on 8)

The computer program product of claim 8, wherein statistic comprises a number of iterations of a clustering algorithm applied to the plurality of video shots resulting in one of the plurality of superclusters including each of the plurality of shot clusters, wherein the expected value for the metric is negatively correlated to the number of iterations, wherein a first iteration of the clustering algorithm applies a more restrictive clustering threshold than a second iteration of the clustering algorithm.

### Claim 12 (depends on 11)

The computer program product of claim 11, wherein the clustering algorithm comprises:
defining the clustering threshold for a current iteration;
generating the plurality of shot clusters based on the clustering threshold; and
assigning each shot cluster in a set of shot clusters, of the plurality of shot clusters, to a respective supercluster of the plurality of the plurality of superclusters, wherein each shot cluster in the set of shot clusters remains in the respective supercluster in a subsequent iteration of the clustering algorithm.

### Claim 13 (depends on 8)

The computer program product of claim 8, further comprising prior to computing the expected value:
generating the plurality of video shots from the video;
generating the plurality of shot clusters based on similarities between the video shots, wherein the similarities between the video shots include at least one of showing a same person, showing a same place, and a same low level feature including at least one of a similar color and a similar low level geometry descriptor; and
generating the plurality of superclusters based on interactions between the plurality of shot clusters.

### Claim 14 (depends on 8)

The computer program product of claim 8, wherein the expected value for the metric is computed responsive to a user request to return the expected value for the metric, wherein a predictive algorithm is configured to compute the expected value for the metric, wherein the metric comprises at least one of: (i) a number of views the video will receive, (ii) a number of likes the video will receive, (iii) a duration of the video a user will watch, (iv) a number of users that will watch a second video promoted by the video, (v) a likelihood that a user will click a link to an advertisement in the video, (vi) a likelihood that a user will stop watching the video at each second of the video, and (vii) an expected amount of advertising revenue the video will generate.

### Claim 15

A system, comprising:
a processor; and
a memory containing a program which when executed by the processor performs an operation comprising:
receiving a plurality of video shots of a video, wherein each video shot includes one or more frames of the video;
generating a hierarchy based on the plurality of video shots, wherein the hierarchy comprises a plurality of superclusters that include at least one of a plurality of shot clusters and describes relationships between the plurality of video shots, wherein each of the shot clusters includes at least one of the plurality of video shots;
computing a value for a statistic for the video based on the hierarchy; and
computing an expected value for a metric based on the statistic.

### Claim 16

(canceled)

### Claim 17 (depends on 15)

The system of claim 15, wherein the statistic comprises a ratio of a count of the plurality of superclusters having less than four of the plurality of shot clusters to a total count of the plurality of superclusters, wherein the expected value for the metric is positively correlated to the ratio.

### Claim 18 (depends on 15)

The system of claim 15, wherein statistic comprises a number of iterations of a clustering algorithm applied to the plurality of video shots resulting in one of the plurality of superclusters including each of the plurality of shot clusters, wherein the expected value for the metric is negatively correlated to the number of iterations, wherein a first iteration of the clustering algorithm applies a more restrictive clustering threshold than a second iteration of the clustering algorithm.

### Claim 19 (depends on 18)

The system of claim 18, wherein the clustering algorithm comprises:
defining the clustering threshold for a current iteration;
generating the plurality of shot clusters based on the clustering threshold; and
assigning each shot cluster in a set of shot clusters, of the plurality of shot clusters, to a respective supercluster of the plurality of the plurality of superclusters, wherein each shot cluster in the set of shot clusters remains in the respective supercluster in a subsequent iteration of the clustering algorithm.

### Claim 20 (depends on 15)

The system of claim 15, further comprising prior to computing the expected value:
generating the plurality of video shots from the video;
generating the plurality of shot clusters based on similarities between the video shots, wherein the similarities between the video shots include at least one of showing a same person, showing a same place, and a same low level feature including at least one of a similar color and a similar low level geometry descriptor; and
generating the plurality of superclusters based on interactions between the plurality of shot clusters.
