Methods and apparatus to correct errors in audience measurements for media accessed using over-the-top devices

• US 10,045,082 B2
• Filed: 12/13/2015
• Issued: 08/07/2018
• Est. Priority Date: 07/02/2015
• Status: Active Grant

• Current Assignee: The Nielsen Company LLC (MOOD Media Corporation)
• Original Assignee: The Nielsen Company LLC (MOOD Media Corporation)
• Inventors: Nagaraja Rao, Kumar, Toker Yildiz, Kamer, Haskell, Jennifer, Ion, Cristina, Zhang, Mimi
• Primary Examiner(s): Huerta, Alexander Q
• Application Number:
  US14/967,355
• Publication Number:
  US 20170006342A1
• Time in Patent Office: 968 Days
• CPC Class Codes: G06Q 30/00 Commerce
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0241 Advertisements
• CPC Class Codes: G06Q 30/0251 Targeted advertisements
• CPC Class Codes: G06Q 50/01 Social networking
• CPC Class Codes: H04N 21/252 Processing of multiple end-users' preferences to derive collaborative data
• CPC Class Codes: H04N 21/25883 being end-user demographical data, e.g. age, family status or address arrangements for identifying locations of users in broadcast systems H04H60/52
• CPC Class Codes: H04N 21/25891 being end-user preferences retrieval of video data in a video database based on user preferences G06F16/739; arrangements for recognizing users' preferences H04H60/46; user profiles in network data switching protocols H04L67/306; processing of user preferences or user profiles in wireless networks H04W8/18
• CPC Class Codes: H04N 21/44222 Analytics of user selections, e.g. selection of programs or purchase activity monitoring of user selections in data processing systems G06F11/34; arrangements for monitoring the user's behaviour or opinions in broadcast systems H04H60/33
• CPC Class Codes: H04N 21/44226 on social networks
• CPC Class Codes: H04N 21/4532 involving end-user characteristics, e.g. viewer profile, preferences monitoring of user activities for profile generation for accessing a video database G06F16/739; user profiles in network data switching protocols H04L67/306; processing of user preferences or user profiles in wireless networks H04W8/18
• CPC Class Codes: H04N 21/4661 Deriving a combined profile for a plurality of end-users of the same client, e.g. for family members within a home user profiles in network data switching protocols H04L67/306
• CPC Class Codes: H04N 21/4663 involving probabilistic networks, e.g. Bayesian networks
• CPC Class Codes: H04N 21/4667 Processing of monitored end-user data, e.g. trend analysis based on the log file of viewer selections
• CPC Class Codes: H04N 21/812 involving advertisement data advertising per se G06Q30/02

Abstract

Methods and apparatus to correct errors in measuring audiences of over-the-top media are disclosed. In some examples, the methods and apparatus identify a first set of data from a first data source, the first set of data different from a second set of data from a second data source. In some examples, the methods and apparatus generate a third set of data based on the second set of data from the second data source. In some examples, the methods and apparatus generate a model based on a difference between the first set of data and the third set of data. In some examples, the methods and apparatus apply the model to the first set of data. In some examples, the methods and apparatus assign viewership to an impression associated with the first set of data by imputing viewership associated with the second set of data to the first set of data.

Claims

1. A method comprising:

identifying, by executing an instruction via a processor, a first set of impression data received from a computer at a first data source, the first set of impression data having matched demographic data from users registered with both an over-the-top (OTT) device and a database proprietor, the first set of impression data different from a second set of data from a second data source, the computer producing a misattribution error in the first set of impression data, the misattribution error based on a demographic data error in the first set of impression data, the demographic data error based on a difference between reported demographic data in the first set of impression data and actual demographic data corresponding to the first set of impression data;

generating, by executing an instruction via the processor, a third set of data based on the second set of data from the second data source;

generating, by executing an instruction via the processor, an independent binary model based on a difference between the first set of impression data and the third set of data;

correcting the demographic data error in the first set of impression data by applying, by executing an instruction via the processor, the independent binary model to the first set of impression data to generate corrected demographic data; and

correcting the misattribution error produced by the computer by assigning, by executing an instruction via the processor, viewership to an impression associated with the first set of impression data using the corrected demographic data.

2. The method as defined in claim 1, wherein the generating of the third set of data includes:

identifying a demographic bucket;

determining whether the demographic bucket exists in a household based on the independent binary model; and

determining the number of members in the household associated with the demographic bucket based on a demographic distribution associated with the second set of data.

3. The method as defined in claim 2, wherein the independent binary model is a machine learning algorithm generated based on at least one of a household size, an age, a gender, a person status, an income, an education, or an ethnicity.

4. The method as defined in claim 3, wherein the machine learning algorithm is at least one of a classification and regression tree, a log it function, a conditional inference tree, a random forest, a neural network, or a bootstrap aggregate decision tree.

5. The method as defined in claim 1, wherein the first set of impression data from the first data source has a fourth set of data missing, further including linearly scaling the first set of impression data to accommodate for the missing fourth set of data.

6. A method comprising:

identifying, by executing an instruction via a processor, a first set of impression data received from a computer at a first data source, the first set of impression data different from a second set of data from a second data source, the computer producing a misattribution error in the first set of impression data, the misattribution error based on a demographic data error in the first set of impression data, the demographic data error based on a difference between reported demographic data in the first set of impression data and actual demographic data corresponding to the first set of impression data;

generating, via the processor, a third set of data based on the second set of data from the second data source;

generating, via the processor, an independent binary model based on a difference between the first set of impression data and the third set of data; and

correcting the demographic data error in the first set of impression data by applying, via the processor, the independent binary model to the first set of impression data to generate corrected demographic data; and

correcting the misattribution error produced by the computer by assigning, via the processor, viewership to an impression associated with the first set of impression data using the corrected demographic data, the assigning of the viewership to the impression includes;

identifying viewing history associated with the second set of data;

determining a first time associated with a first demographic viewing a media presentation in a first household associated with the second set of data;

determining a second time associated with the first demographic and a second demographic viewing the media presentation in the household;

determining a first probability that the first demographic viewed the media presentation by dividing the first time by the second time;

identifying a first person in the first household associated with the second set of data having a second probability similar to the first probability; and

imputing a viewing history of the first person to a second person in a second household associated with the first set of impression data.

7. The method as defined in claim 6, wherein the identifying of the first person in the first household associated with the second set of data having the second probability similar to the first probability includes:

identifying a third person in the first household associated with the second set of data;

ranking the first and third person based on a first highest probability;

identifying a fourth person in the second household associated with the first set of impression data;

ranking the second and fourth person based on a second highest probability; and

identifying the first person has the second probability similar to the first probability when the first person has the first highest probability and the second person has the second highest probability.

8. An apparatus comprising:

a demographic corrector to;

identify a first set of impression data received from a computer at a first data source, the first set of impression data having matched demographic data from users registered with both an over-the-top (OTT) device and a database proprietor, the first set of impression data different from a second set of data from a second data source, the computer producing a misattribution error in the first set of impression data, the misattribution error based on a demographic data error in the first set of impression data, the demographic data error based on a difference between reported demographic data in the first set of impression data and actual demographic data corresponding to the first set of impression data;

generate a third set of data based on the second set of data from the second data source;

generate an independent binary model based on a difference between the first set of impression data and the third set of data; and

correct the demographic data error in the first set of impression data by applying the independent binary model to the first set of impression data to generate corrected demographic data; and

a viewership assigner to correct the misattribution error produced by the computer by assigning viewership to an impression associated with the first set of impression data using the corrected demographic data, in which at least one of the demographic corrector or the viewership assigner is a logic circuit.

9. The apparatus as defined in claim 8, wherein to generate the third set of data, the demographic corrector is to:

identify a demographic bucket;

determine whether the demographic bucket exists in a household based on the independent binary model; and

determine the number of members in the household associated with the demographic bucket based on a demographic distribution associated with the second set of data.

10. The apparatus as defined in claim 9, wherein the independent binary model is a machine learning algorithm generated based on at least one of a household size, an age, a gender, a person status, an income, an education, or an ethnicity.

11. The apparatus as defined in claim 10, wherein the machine learning algorithm is at least one of a classification and regression tree, a log it function, a conditional inference tree, a random forest, a neural network, or a bootstrap aggregate decision tree.

12. The apparatus as defined in claim 8, wherein the first set of impression data from the first data source has a fourth set of data missing from the first set of impression data, further including a demographic predictor to linearly scale the first set of impression data to accommodate for the missing fourth set of data.

13. An apparatus comprising:

a demographic corrector to;

identify a first set of impression data received from a computer at a first data source, the first set of impression data different from a second set of data from a second data source, the computer producing a misattribution error in the first set of impression data, the misattribution error based on a demographic data error in the first set of impression data, the demographic data error based on a difference between reported demographic data in the first set of impression data and actual demographic data corresponding to the first set of impression data;

generate a third set of data based on the second set of data from the second data source;

generate a model based on a difference between the first set of impression data and the third set of data; and

correct the demographic data error in the first set of impression data by applying the model to the first set of impression data to generate corrected demographic data; and

a viewership assigner to correct the misattribution error produced by the computer by assigning viewership to an impression associated with the first set of impression data using the corrected demographic data, the viewership assigner is to assign viewership to the impression by;

identifying viewing history associated with the second set of data;

determining a first time associated with a first demographic viewing a media presentation in a first household associated with the second set of data;

determining a second time associated with the first demographic and a second demographic viewing the media presentation in the household;

determining a first probability that the first demographic viewed the media presentation by dividing the first time by the second time;

identifying a first person in the first household associated with the second set of data having a second probability similar to the first probability; and

imputing a viewing history of the first person to a second person in a second household associated with the first set of impression data, at least one of the demographic corrector or the viewership assigner is a logic circuit.

14. The apparatus as defined in claim 13, wherein to identify the first person in the first household associated with the second set of data having the second probability similar to the first probability, the viewership assigner is to:

identify a third person in the first household associated with the second set of data;

rank the first and third person based on a first highest probability;

identify a fourth person in the second household associated with the first set of impression data;

rank the second and fourth person based on a second highest probability; and

identify the first person has the second probability similar to the first probability when the first person has the first highest probability and the second person has the second highest probability.

15. A tangible computer readable storage medium comprising instructions that, when executed, cause a machine to at least:

identify a first set of impression data received from a computer at a first data source, the first set of impression data having matched demographic data from users registered with both an over-the-top (OTT) device and a database proprietor, the first set of impression data different from a second set of data from a second data source, the computer producing a misattribution error in the first set of impression data, the misattribution error based on a demographic data error in the first set of impression data, the demographic data error based on a difference between reported demographic data in the first set of impression data and actual demographic data corresponding to the first set of impression data;

generate a third set of data based on the second set of data from the second data source;

generate an independent binary model based on a difference between the first set of impression data and the third set of data;

correct the demographic data error in the first set of impression data by applying the independent binary model to the first set of impression data to generate corrected demographic data; and

correct the misattribution error produced by the computer by assigning viewership to an impression associated with the first set of impression data using the corrected demographic data.

16. The tangible computer readable storage medium as defined in claim 15, wherein the instructions, when executed, further cause the machine to:

identify a demographic bucket;

determine whether the demographic bucket exists in a household based on the independent binary model; and

determine the number of members in the household associated with the demographic bucket based on a demographic distribution associated with the second set of data.

17. The tangible computer readable storage medium as defined in claim 16, wherein the independent binary model is a machine learning algorithm generated based on at least one of a household size, an age, a gender, a person status, an income, an education, or an ethnicity.

18. The tangible computer readable storage medium as defined in claim 15, wherein the first set of impression data from the first data source has a fourth set of data missing from the first set of impression data, the instructions, when executed, to further cause the machine to linearly scale the first set of impression data to accommodate for the missing fourth set of data.

19. A tangible computer readable storage medium comprising instructions that, when executed, cause a machine to at least:

identify a first set of impression data received from a computer at a first data source, the first set of impression data different from a second set of data from a second data source, the computer producing a misattribution error in the first set of impression data, the misattribution error based on a demographic data error in the first set of impression data, the demographic data error based on a difference between reported demographic data in the first set of impression data and actual demographic data corresponding to the first set of impression data;

generate a third set of data based on the second set of data from the second data source;

generate an independent binary model based on a difference between the first set of impression data and the third set of data;

correct the demographic data error in the first set of impression data by applying the independent binary model to the first set of impression data to generate corrected demographic data;

correct the misattribution error produced by the computer by assigning viewership to an impression associated with the first set of impression data using the corrected demographic data;

identify viewing history associated with the second set of data;

determine a first time associated with a first demographic viewing a media presentation in a first household associated with the second set of data;

determine a second time associated with the first demographic and a second demographic viewing the media presentation in the household;

determine a first probability that the first demographic viewed the media presentation by dividing the first time by the second time;

identify a first person in the first household associated with the second set of data having a second probability similar to the first probability; and

impute a viewing history of the first person to a second person in a second household associated with the first set of impression data.

20. The tangible computer readable storage medium as defined in claim 19, wherein the instructions, when executed, further cause the machine to:

identify a third person in the first household associated with the second set of data;

rank the first and third person based on a first highest probability;

identify a fourth person in the second household associated with the first set of impression data;

rank the second and fourth person based on a second highest probability; and

identify the first person has the second probability similar to the first probability when the first person has the first highest probability and the second person has the second highest probability.