Granted Patent B1

US 8,874,652 · App. 14/211,375 · Granted Oct 28, 2014

Abstract

Claims (52)

1. A computer-implemented method comprising:

accessing panel data representing accesses by a first set of users with multiple types of media platforms to media content associated with multiple media entities, the first set of users being included in a sampled panel of users;

based on the accessed panel data, determining, for each of multiple media entities, an overlap in the accessed panel data that represents users in the first set of users who have accessed media content associated with the media entity with more than one of the multiple types of media platforms;

accessing census data representing accesses by a second set of users with the multiple types of media platforms to media content associated with the multiple media entities, wherein the second set of users includes a greater number of users than the first set of users;

based on the accessed panel data, the determined overlap in the accessed panel data, and the accessed census data, deriving, by one or more processors, an overlap function that estimates an overlap in the accessed census data that represents users in the second set of user who have accessed media content associated with a media entity with more than one of the multiple types of media platforms; and

applying, by the one or more processors, the derived overlap function to census data associated with a first media entity to estimate an overlap in the census data associated with the first media entity that represents users who have accessed media content associated with the first media entity with more than one of the multiple types of media platforms.

2. A computer-implemented method comprising:

accessing panel data representing accesses by a first set of users with multiple types of media platforms to media content associated with multiple media entities, the first set of users being included in a sampled panel of users;

based on the accessed panel data, determining, for each of multiple media entities, a panel overlap in the accessed panel data that represents users in the first set of users who have accessed media content associated with the media entity with more than one of the multiple types of media platforms;

accessing census data representing accesses by a second set of users with the multiple types of media platforms to media content associated with the multiple media entities, wherein the second set of users includes a greater number of users than the first set of users; and

based on the accessed panel data, the determined overlap in the accessed panel data, and the accessed census data, estimating, for each of a sub-set of the multiple media entites, a census overlap in the accessed census data, the census overlap representing users in the second set of users who have accessed media content associated with a media entity with more than one of the multiple types of media platforms;

based on the census overlap estimated for the subset of media entities and the accessed census data, training a model to generate an overlap function, the overlap function configured to estimate an overlap in the accessed census data that represents users in the second set of user who have accessed media content associated with a media entity in the subset of media entities with more than one of the multiple types of media platforms; and

applying, by the one or more processors, the derived overlap function to census data associated with a first media entity to estimate an overlap in the census data associated with the first media entity, the estimated overlap in the census data represents users who have accessed media content associated with the first media entity with more than one of the multiple types of media platforms.

3. The computer-implemented method of claim 2 , wherein estimating, for each of a sub-set of the multiple media entites, a census overlap in the accessed census data comprises applying a Maximum Likelihood Estimate (MLE) method to the accessed panel data, the determined overlap in the accessed panel data, and the accessed census data associated with the sub-set of the multiple media entites.

4. The computer-implemented method of claim 2 , wherein training a model to generate an overlap function comprises training a quantile regression model to the census overlap estimated for the subset of media entities and features derived from the accessed census data.

5. The computer-implemented method of claim 4 , wherein the features derived from the accessed census data comprise platform reach, frequency on platform, relative size of platforms, and dictionary differences between related media entities on each platform.

6. The computer-implemented method of claim 2 , wherein:

the accessed panel data represents accesses by the first set of users with three types of media platforms; and

training a model to generate an overlap function comprises generating an overlap function for each unique pair of the three types of media platforms, wherein each generated overlap function estimates overlap in the census data that represent users who have accessed media content associated with the first media entity with both media platforms in a unique pair of the three types of media platforms.

7. The computer-implemented method of claim 6 , further comprising combining outputs from each of the generated overlap functions for each unique pair of the three types of media platforms to estimate a de-duplicated audience for a media entity across the three types of media platforms.

8. A system comprising:

a first set of media platforms configured to access media content associated with one or more media entities, the first set of media platforms being of multiple different types and being associated with a first set of users who access the media content with the media platforms;

a second set of media platforms configured to access media content associated with one or more media entities, the second set of media platforms being of multiple different types and being associated with a second set of users who access the media content with the media platforms;

one or processors;

one or more storage devices that store instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising:

accessing panel data representing accesses by the first set of users with the first set of media platforms to media content associated with multiple media entities, the first set of users being included in a sampled panel of users;

based on the accessed panel data, determining, for each of multiple media entities, a panel overlap in the accessed panel data that represents users in the first set of users who have accessed media content associated with the media entity with more than one of the multiple types of media platforms;

accessing census data representing accesses by the second set of users with the second set of media platforms to media content associated with the multiple media entities, wherein the second set of users includes a greater number of users than the first set of users; and

based on the accessed panel data, the determined overlap in the accessed panel data, and the accessed census data, estimating, for each of a sub-set of the multiple media entites, a census overlap in the accessed census data, the census overlap representing users in the second set of users who have accessed media content associated with a media entity with more than one of the multiple types of media platforms;

based on the census overlap estimated for the subset of media entities and the accessed census data, training a model to generate an overlap function, the overlap function configured to estimate an overlap in the accessed census data that represents users in the second set of user who have accessed media content associated with a media entity in the subset of media entities with more than one of the multiple types of media platforms; and

applying, by the one or more processors, the derived overlap function to census data associated with a first media entity to estimate an overlap in the census data associated with the first media entity, the estimated overlap in the census data represents users who have accessed media content associated with the first media entity with more than one of the multiple types of media platforms.

9. The system of claim 8 , wherein estimating, for each of a sub-set of the multiple media entites, a census overlap in the accessed census data comprises applying a Maximum Likelihood Estimate (MLE) method to the accessed panel data, the determined overlap in the accessed panel data, and the accessed census data associated with the sub-set of the multiple media entites.

10. The system of claim 8 , wherein training a model to generate an overlap function comprises training a quantile regression model to the census overlap estimated for the subset of media entities and features derived from the accessed census data.

11. The system of claim 10 , wherein the features derived from the accessed census data comprise platform reach, frequency on platform, relative size of platforms, and dictionary differences between related media entities on each platform.

12. The system of claim 8 , wherein:

the accessed panel data represents accesses by the first set of users with three types of media platforms; and

training a model to generate an overlap function comprises generating an overlap function for each unique pair of the three types of media platforms, wherein each generated overlap function estimates overlap in the census data that represent users who have accessed media content associated with the first media entity with both media platforms in a unique pair of the three types of media platforms.

13. The system of claim 12 , wherein the one or more storage devices further store instructions that, when executed by the one or more processors, cause the one or more processors to perform the operation of: combining outputs from each of the generated overlap functions for each unique pair of the three types of media platforms to estimate a de-duplicated audience for a media entity across the three types of media platforms.

14. A non-transitory computer readable medium storing instructions that, when executed by one or more computing devices, cause the one or more computing devices to perform operations comprising:

accessing panel data representing accesses by a first set of users with multiple types of media platforms to media content associated with multiple media entities, the first set of users being included in a sampled panel of users;

based on the accessed panel data, determining, for each of multiple media entities, a panel overlap in the accessed panel data that represents users in the first set of users who have accessed media content associated with the media entity with more than one of the multiple types of media platforms;

accessing census data representing accesses by a second set of users with the multiple types of media platforms to media content associated with the multiple media entities, wherein the second set of users includes a greater number of users than the first set of users; and

based on the accessed panel data, the determined overlap in the accessed panel data, and the accessed census data, estimating, for each of a sub-set of the multiple media entites, a census overlap in the accessed census data, the census overlap representing users in the second set of users who have accessed media content associated with a media entity with more than one of the multiple types of media platforms;

based on the census overlap estimated for the subset of media entities and the accessed census data, training a model to generate an overlap function, the overlap function configured to estimate an overlap in the accessed census data that represents users in the second set of user who have accessed media content associated with a media entity in the subset of media entities with more than one of the multiple types of media platforms; and

applying, by the one or more processors, the derived overlap function to census data associated with a first media entity to estimate an overlap in the census data associated with the first media entity, the estimated overlap in the census data represents users who have accessed media content associated with the first media entity with more than one of the multiple types of media platforms.

15. The non-transitory computer readable medium of claim 14 , wherein estimating, for each of a sub-set of the multiple media entites, a census overlap in the accessed census data comprises applying a Maximum Likelihood Estimate (MLE) method to the accessed panel data, the determined overlap in the accessed panel data, and the accessed census data associated with the sub-set of the multiple media entites.

16. The non-transitory computer readable medium of claim 14 , wherein training a model to generate an overlap function comprises training a quantile regression model to the census overlap estimated for the subset of media entities and features derived from the accessed census data.

17. The non-transitory computer readable medium of claim 16 , wherein the features derived from the accessed census data comprise platform reach, frequency on platform, relative size of platforms, and dictionary differences between related media entities on each platform.

18. The non-transitory computer readable medium of claim 14 , wherein:

the accessed panel data represents accesses by the first set of users with three types of media platforms; and

training a model to generate an overlap function comprises generating an overlap function for each unique pair of the three types of media platforms, wherein each generated overlap function estimates overlap in the census data that represent users who have accessed media content associated with the first media entity with both media platforms in a unique pair of the three types of media platforms.

19. The non-transitory computer readable medium of claim 18 , wherein the stored instructions further comprise instructions that, when executed by the one or more computing devices, cause the one or more computing devices to perform the operation of: combining outputs from each of the generated overlap functions for each unique pair of the three types of media platforms to estimate a de-duplicated audience for a media entity across the three types of media platforms.

Assignments (7)