Methods and apparatus to correct for deterioration of a demographic model to associate demographic information with media impression information
Methods and apparatus to correct for deterioration of a demographic model to associate demographic information with media impression information are disclosed. An example method includes estimating first and second ages of audience members based on demographic information; estimating a third age of an audience member who is not included in the audience members; applying a window function to the second ages to determine a distribution of ages based on the third age; multiplying window values by the first ages to determine corrected first age components; dividing a total of the corrected first age components by a sum of the window values to determine an estimated age of the audience member at a first time; and determining the corrected age of the audience member at a second time based on the estimated age of the audience member at the first time and a time difference between the first and second times.
1. An audience measurement system for correcting age-based misattribution errors, comprising:
a processor; and
a non-transitory computer-readable storage medium, having stored thereon program instructions that, upon execution by the processor, cause performance of a set of operations comprising:
receiving, from a database proprietor, anonymized subscriber demographic data of a plurality of subscribers registered with the database proprietor, wherein the anonymized subscriber demographic data comprises corresponding age characteristics of the plurality of subscribers and is based on network communications received by the database proprietor from client devices of the plurality of subscribers;
storing, in a database, the anonymized subscriber demographic data;
receiving, from a client device, impression data after a media item is accessed on the client device, wherein the impression data comprises an impression request and a user identifier;
generating a demographic impression of a plurality of demographic impressions by attributing anonymized subscriber demographic data of a subscriber of the plurality of subscribers to the impression data,
wherein the demographic impression comprises an impression date and the user identifier, and
wherein the impression date and the user identifier are matched to the anonymized subscriber demographic data of the subscriber of the plurality of subscribers;
retrieving, based on the user identifier, a predicted age probability density function,
wherein the predicted age probability density function defines, for each of a plurality of age ranges, a respective probability that an actual age of the subscriber is within that age range;
selecting, using the predicted age probability density function, an age range of the plurality of age ranges that corresponds to the age characteristic of the subscriber;
calculating an aging factor based on the selected age range;
applying the aging factor to update the demographic impression to correct the age characteristic of the subscriber; and
outputting, using the updated demographic impression, media ratings information about the media item, wherein the updated demographic impression comprises a corrected age characteristic of the subscriber.
2. The audience measurement system of claim 1 , wherein the predicted age probability density function is stored at a first point in time.
3. The audience measuremnt system of claim 2 , wherein retrieving the predicted age probability density function occurs at a second point in time; and wherein the second point in time occurs after the first point in time.
4. The audience measurement system of claim 2 , further comprising:
storing, in another database, the user identifier, a predicted age probability density function identifier associated with the predicted age probability density function, and a prediction date, wherein the prediction date corresponds to the first point in time when the predicted age probability density function was originally predicted.
5. The audience measurement system of claim 4 , wherein the user identifier, the predicted age probability density function identifier, and the prediction date are stored in a same data structure in the other database; and wherein locating the user identifier from the another database also retrieves the predicted age probability density function identifier and the prediction date.
6. The audience measurement system of claim 5 , further comprises:
updating, based on applying the aging factor, the predicted age probability density function to generate an age-corrected probability density function; and
associating an age-corrected probability density function identifier with the age-corrected probability density function.
7. The audience measurement system of claim 6 , further comprising:
storing the age-corrected probability density function identifier by replacing the predicted age probability density function identifier with the age-corrected probability density function identifier asssociated with the user identifier.
8. A non-transitory computer-readable storage medium, having stored thereon program instructions that, upon execution by a processor, cause performance of a set of operations comprising:
receiving, from a database proprietor, anonymized subscriber demographic data of a plurality of subscribers registered with the database proprietor, wherein the anonymized subscriber demographic data comprises corresponding age characteristics of the plurality of subscribers and is based on network communications received by the database proprietor from client devices of the plurality of subscribers;
storing, in a database, the anonymized subscriber demographic data;
receiving, from a client device, impression data after a media item is accessed on a client device, wherein the impression data comprises an impression request and a user identifier;
generating a demographic impression of a plurality of demographic impressions by attributing anonymized subscriber demographic data of a subscriber of the plurality of subscribers to the impression data,
wherein the demographic impression comprises an impression date and the user identifier, and
wherein the impression date and the user identifier are matched to the anonymized subscriber demographic data of the subscriber of the plurality of subscribers;
retrieving, based on the user identifier, a predicted age probability density function,
wherein the predicted age probability density function defines, for each of a plurality of age ranges, a respective probability that an actual age of the subscribers is within that age range;
selecting, using the predicted age probability density function, an age range of the plurality of age ranges that corresponds to the age characteristic of the subscriber;
calculating an aging factor based on the selected age range;
applying the aging factor to update the demographic impression to correct the age characteristic of the subscriber; and
outputting, using the updated demographic impression, media ratings information about the media item, wherein the updated demographic impression comprises a corrected age characteristic of the subscriber.
9. The non-transitory computer-readable storage medium of claim 8 , wherein the predicted age probability density function is stored at a first point in time.
10. The non-transitory computer-readable storage medium of claim 9 , wherein retrieving the predicted age probability density function occurs at a second point in time; and wherein the second point in time occurs after the first point in time.
11. The non-transitory computer-readable storage medium of claim 9 , the set of operations further comprising:
storing, in another database, the user identifier, a predicted age probability density function identifiers associated with the predicted age probability density function, and a prediction date, wherein the prediction date corresponds to the first point in time when the predicted age probability density function was originally predicted.
12. The non-transitory computer-readable storage medium of claim 11 , wherein the user identifier, the predicted age probability density function identifier, and the prediction date are stored in a same data structure in the other database; and wherein locating the user identifier from the another database also retrieves the predicted age probability density function identifier and the prediction date.
13. The non-transitory computer-readable storage medium of claim 12 , the set of operations further comprises:
updating, based on applying the aging factor, the predicted age probability density function to generate an age-corrected probability density function; and
associating an age-corrected probability density function identifier with the age-corrected probability density function.
14. The non-transitory computer-readable storage medium of claim 13 , the set of operations further comprising:
storing the age-corrected probability density function identifier by replacing the predicted age probablility density function identifier with the age-corrected probability density function identifier associated with the user identifier.
15. A method for correcting age-based misattribution errors, the method omprising:
receiving, by a server of an audience measurement system, a network communication receiving, from a database proprietor, anonymized subscriber demographic data of a plurality of subscribers registered with the database proprietor, wherein the anonymized subscriber demographic data comprises corresponding age characteristics of the plurality of subscribers and is based on network communications received by the database proprietor from client devices of the plurality of subscribers;
storing, in a database, the anonymized subscriber demoghaphic data;
receiving, from a client device, immpression data after a media item is accessed on the client device, wherein the impression data comprises an impression request and a user identifier;
generating a demographic impression of a plurality of demographic impressions by attributing anonymized subscriber demographic data of a subscriber of the plurality of subscribers to the impression data,
wherein the demographic impression comprises an impression date and the user identifier, and
wherein the impression date and the user identifier are matched to the anonymized subscriber demographic data of the subscriber of the plurality of subscribers;
retrieving, based on the user identifier, a predicted age probability density function,
wherein the predicted age probability density function defines, for each of a plurality of age ranges, a respective probability that an actual age of the subscriber is within that age range;
selecting, using the predicted age probability density function, an age range of the plurality of age ranges that corresponds to the age charasteristic of the subscriber;
caluclating an aging factor based on the selected age range;
applying the aging factor to update the demographic impression to correct the age characteristic of the subscriber; and
outputting, using the updated demographic impression, media ratings information about the media item, wherein the updated demographic impression comprises a corrected age characteristic of the subscriber.
16. The method of claim 15 , wherein the predicted age probability density function is stored at a first point in time; and wherein retrieving the predicted age probability density function occurs at a secon d point in time; and wherein the second point in time occurs after the first point in time.
17. The method of claim 16 , further comprising:
storing, in another database, the user identifier, a predicted age probability density function identifier associated with the predicted age probability density function, and a prediction date, wherein the prediction date corresponds to the first point in time when the predicted age probability density function was originally predicted.
18. The method of claim 17 , wherein the user identifier, the predicted age probability density function identifier, and the prediction date are stored in a same data structure in the other database; and wherein locating the user identifier from the another database also retrieves the predicted age probability density function identifier and the prediction date.
19. The method of claim 18 , further comprises:
updating, based on applying the aging factor, the predicted age probability density function to generate an age-corrected probability density function; and
associating an age-corrected probability density function identifier with the age-corrected probability density function.
20. The method of claim 19 , further comprising:
storing the age-corrected probability density function identifier by replacing the predicted age probability density function identifier with the age-corrected probability density function identifier associated with the user identifier.