# US 2020/0404351 A1 · Video analytics system

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

- Number: US 2020/0404351 A1
- Document ID: US20200404351A1
- Status: Published application
- Kind: A1
- Application number: 17/009,659
- Filed: 2020-09-01
- Published: 2020-12-24
- Earliest claimed: 2018-01-19
- Assignees: Mux, Inc.
- Applicants: Mux, Inc.
- Inventors: Adam Brown; Jonathan Dahl; Steven Hellernan; Justin Sanford; Matthew Ward; Scott Kidder; Benjamin Dodson; Alex Diel
- Web page: https://patentgrep.com/patent/US20200404351A1

## CPC

- H04N 21/2407: {Monitoring of transmitted content, e.g. distribution time, number of downloads (arrangements for monitoring programmes for broadcast or for distribution combined with broadcast H04H20/14)}
- H04N 21/4425: Monitoring of client processing errors or hardware failure
- H04N 21/466: Learning process for intelligent management, e.g. learning user preferences for recommending movies {(services using the results of monitoring in broadcast systems H04H60/61)}
- H04N 21/64723: {Monitoring of network processes or resources, e.g. monitoring of network load (traffic related reporting in data switching networks H04L43/062)}
- H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION

## Abstract

A video monitoring system can include multiple collectors to receive video beacon data from multiple video monitoring interface modules. At least one beacon stream is connected to receive data from multiple collectors. A processing module receives the beacon stream and provides a real-time event stream used for real-time data analysis and a video view stream used for long-term data analysis.

## Claims

### Claim 1

A method for monitoring playback of a video over a network comprising:
for each video player in a set of video players streaming the video during a first time period:
receiving a set of beacon data from the video player, the set of beacon data:
generated by a video interface module of the video player; and
comprising a set of tracked data metrics characterizing the video;
classifying the set of beacon data as deriving from a user view based on the set of tracked data metrics and a crawler model; and
in response to classifying the set of beacon data as deriving from the user view, aggregating the set of beacon data into a beacon stream;
generating a real-time event stream based on the set of tracked data metrics of each set of beacon data in the beacon stream;
calculating a summary metric based on the real-time event stream;
comparing the summary metric to historical data for the summary metric;
detecting a statistical anomaly based on the historical data for the summary metric; and
in response to detecting the statistical anomaly, serving an alert indicating the summary metric and the statistical anomaly at an interface module.

### Claim 2 (depends on 1)

The method of claim 1:
wherein receiving the set of beacon data comprises receiving the set of beacon data comprising the set of tracked data metrics characterizing the video and comprising an operating frequency of a computational device executing the video player; and
wherein classifying the set of beacon data as deriving from the user view based on the set of tracked data metrics and the crawler model comprises classifying the set of beacon data as deriving from the user view based on the operating frequency of the computational device executing the video player and the crawler model.

### Claim 3 (depends on 1)

The method of claim 1:
wherein receiving the set of beacon data comprises receiving the set of beacon data comprising the set of tracked data metrics characterizing the video and comprising an identifier of a computational device executing the video player; and
wherein classifying the set of beacon data as deriving from the user view based on the set of tracked data metrics and the crawler model comprises classifying the set of beacon data as deriving from the user view based on the identifier of the computational device executing the video player and the crawler model.

### Claim 4 (depends on 1)

The method of claim 1:
wherein calculating the summary metric based on the real-time event stream comprises calculating an error rate based on the real-time event stream;
wherein comparing the summary metric to historical data for the summary metric comprises comparing the error rate to historical data for the error rate;
wherein detecting the statistical anomaly based on the historical data for the summary metric comprises detecting the statistical anomaly based on the historical data for the error rate; and
wherein, in response to detecting the statistical anomaly, serving the alert indicating the summary metric and the statistical anomaly at the interface module comprises, in response to detecting the statistical anomaly, serving the alert indicating the error rate and the statistical anomaly at the interface module.

### Claim 5 (depends on 1)

The method of claim 1:
wherein calculating the summary metric based on the real-time event stream comprises calculating a number of rebuffering events based on the real-time event stream;
wherein comparing the summary metric to historical data for the summary metric comprises comparing the number of rebuffering events to historical data for the number of rebuffering events;
wherein detecting the statistical anomaly based on the historical data for the summary metric comprises detecting the statistical anomaly based on the historical data for the number of rebuffering events; and
wherein, in response to detecting the statistical anomaly, serving the alert indicating the summary metric and the statistical anomaly at the interface module comprises, in response to detecting the statistical anomaly, serving the alert indicating the number of rebuffering events and the statistical anomaly at the interface module.

### Claim 6 (depends on 1)

The method of claim 1:
wherein calculating the summary metric based on the real-time event stream comprises calculating an average time-to-first-frame based on the real-time event stream;
wherein comparing the summary metric to historical data for the summary metric comprises comparing the average time-to-first-frame to historical data for the average time-to-first-frame;
wherein detecting the statistical anomaly based on the historical data for the summary metric comprises detecting the statistical anomaly based on the historical data for the average time-to-first-frame; and
wherein, in response to detecting the statistical anomaly, serving the alert indicating the summary metric and the statistical anomaly at the interface module comprises, in response to detecting the statistical anomaly, serving the alert indicating the average time-to-first-frame and the statistical anomaly at the interface module.

### Claim 7 (depends on 1)

The method of claim 1, wherein classifying the set of beacon data as deriving from the user view based on the set of tracked data metrics and the crawler model comprises:
calculating an identifying signature for crawlers based on an historical dataset; and
classifying the set of beacon data as deriving from the user view based on the set of tracked data metrics, the identifying signature, and the crawler model.

### Claim 8 (depends on 1)

The method of claim 1:
wherein detecting the statistical anomaly based on the historical data for the summary metric comprises detecting the statistical anomaly based on historical data for the summary metric and for a customer providing the video; and
wherein, in response to detecting the statistical anomaly, serving the alert indicating the summary metric and the statistical anomaly at the interface module comprises, in response to detecting the statistical anomaly, serving the alert indicating the summary metric and the statistical anomaly at the interface module of the customer providing the video.

### Claim 9 (depends on 1)

The method of claim 1, wherein detecting the statistical anomaly based on the historical data for the summary metric comprises detecting the statistical anomaly based on the historical data for the summary metric and a machine learning module.

### Claim 10 (depends on 9)

The method of claim 9, wherein detecting the statistical anomaly based on the historical data for the summary metric and the machine learning module comprises:
predicting a summary metric value based on the historical data for the summary metric via the machine learning module;
calculating a deviation between the summary metric and the summary metric value; and
in response to the deviation exceeding a threshold deviation, detecting the statistical anomaly.

### Claim 11 (depends on 9)

The method of claim 9, wherein detecting the statistical anomaly based on the historical data for the summary metric and the machine learning module comprises:
via the machine learning module, executing a binary classifier based on the summary metric, the historical data for the summary metric, the real-time event stream, and a historical dataset of real-time event streams; and
detecting the statistical anomaly based on the binary classifier.

### Claim 12

A method for monitoring playback of a video over a network comprising:
for each video player in a set of video players streaming the video during a first time period:
receiving a set of beacon data from the video player, the set of beacon data:
generated by a video interface module of the video player; and
comprising a set of tracked data metrics characterizing the video;
classifying the set of beacon data as deriving from a user view based on the set of tracked data metrics and a crawler model;
in response to classifying the set of beacon data as deriving from the user view, aggregating the set of beacon data into a beacon stream;
generating a real-time event stream based on the set of tracked data metrics of each set of beacon data in the beacon stream;
calculating a summary metric based on the real-time event stream; and
in response to the summary metric exceeding a threshold, serving an alert at an interface module.

### Claim 13 (depends on 12)

The method of claim 12, further comprising calculating the threshold based on historical data via a machine learning module.

### Claim 14 (depends on 12)

The method of claim 12, wherein, in response to the summary metric exceeding the threshold, serving the alert at the interface module comprises, in response to the summary metric exceeding a preset threshold, serving the alert at the interface module.

### Claim 15

A method for monitoring playback of a video over a network comprising, for a video player streaming the video and prior to completion of the video by the video player:
at a first time, receiving a first set of beacon data from the video player, the first set of beacon data:
generated by a video interface module of the video player; and
comprising a first set of tracked data metrics characterizing the video at the first time;
classifying the first set of beacon data as deriving from a user view based on the first set of tracked data metrics and a crawler model; and
in response to classifying the first set of beacon data as deriving from the user view:
generating a real-time event stream based on the first set of tracked data metrics;
calculating a summary metric based on the real-time event stream; and
serving a first alert at an interface module based on the summary metric.

### Claim 16 (depends on 15)

The method of claim 15, further comprising, for the video player streaming the video and prior to completion of the video by the video player:
at a second time succeeding the first time by a time interval, receiving a second set of beacon data from the video player, the second set of beacon data:
generated by the video interface module of the video player; and
comprising a second set of tracked data metrics characterizing the video at the second time;
classifying the second set of beacon data as deriving from a user view based on the second set of tracked data metrics and the crawler model; and
in response to classifying the second set of beacon data as deriving from the user view:
aggregating the second set of tracked data metric into the real-time event stream;
recalculating the summary metric based on the real-time event stream; and
serving a second alert at the interface module based on the summary metric.

### Claim 17 (depends on 16)

The method of claim 16, wherein, at the second time succeeding the first time by the time interval, receiving the second set of beacon data from the video player comprises, at the second time succeeding the first time by the time interval less than one minute in duration, receiving the second set of beacon data from the video player.

### Claim 18 (depends on 16)

The method of claim 16, further comprising setting the time interval based on a player type of the video player.

### Claim 19 (depends on 18)

The method of claim 18, wherein setting the time interval based on the player type of the video player comprises, in response to identifying the player type of the video player as a mobile media player type, setting the time interval at less than five seconds.

### Claim 20 (depends on 18)

The method of claim 18, wherein setting the time interval based on the player type of the video player comprises, in response to identifying the player type of the video player as a set-top box player type, setting the time interval at less than thirty seconds.
