Methods and apparatus to correct for deterioration of a demographic model to associate demographic information with media impression information

Abstract

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.

What is claimed is: 
     
         1 . A method to correct for deterioration of a demographic model to associate demographic information with media impression information, comprising:
 collecting, at a processor at an audience measurement entity, messages indicating first impressions of a media item delivered to devices via the Internet, the messages identifying the media item presented at the devices;   receiving, at the processor at the audience measurement entity, first demographic information describing first numbers of impressions of the media item and first numbers of audience members attributed to respective demographic groups by a database proprietor, the first numbers of the impressions and the first numbers of audience members corresponding to the first impressions of the media;   estimating first ages of the audience members based on the first demographic information, the first ages corresponding to a first time;   estimating second ages of the audience members based on the first demographic information, the second ages corresponding to a second time after the first time;   estimating a third age of an audience member who is not included in the audience members from the database proprietor, the third age corresponding to the second time;   estimating a corrected age of the audience member at the second time by:
 applying a window function to the second ages to determine a distribution of ages based on the third age of the audience member as a mean of the distribution; 
 multiplying window values of the second ages by respective ones of the first ages to determine corrected first age components; 
 summing the corrected first age components and 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 the first time; and 
 determining the corrected age of the audience member at the 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; and 
 determining ratings information for the media by attributing impressions and audience counts to the media using the corrected age of the audience member instead of the third age. 
   
     
     
         2 . The method as defined in  claim 1 , wherein the window function includes a probability density function based on a Gaussian distribution. 
     
     
         3 . The method as defined in  claim 1 , wherein the estimating of the first ages includes determining a predicted age probability density function. 
     
     
         4 . The method as defined in  claim 3 , wherein the estimating of the second ages includes applying an aging factor to an age bucket in the predicted age probability density function. 
     
     
         5 . The method as defined in  claim 1 , further including selecting the audience members from a larger set of audience members based on the second ages being within an age range, the age range being based on the third age. 
     
     
         6 . The method as defined in  claim 1 , further including transmitting audience measurement entity identifiers to the devices in response to at least a portion of the messages, the estimating of the third age being in response to determining, based on the audience member not being associated with an audience measurement entity identifier, that the audience member has not been previously identified. 
     
     
         7 . The method as defined in  claim 6 , further including sending re-direct messages in response to at least a portion of the messages, the re-direct messages to cause at least a portion of the devices to send third messages to the database proprietor, the first demographic information being received based on the third messages. 
     
     
         8 . An apparatus to correct for deterioration of a demographic model to associate demographic information with media impression information, comprising:
 a first impressions collector to collect messages indicating first impressions of a media item delivered to devices via the Internet, the messages identifying the media item presented at the devices;   a second impressions collector to receive first demographic information describing first numbers of impressions of the media item and first numbers of audience members attributed to respective demographic groups by a database proprietor, the first numbers of the impressions and the first numbers of audience members corresponding to the first impressions of the media;   an age predictor to:
 estimate first ages of the audience members based on the first demographic information, the first ages corresponding to a first time; 
 estimate second ages of the audience members based on the first demographic information, the second ages corresponding to a second time after the first time; 
 estimate a third age of an audience member who is not included in the audience members from the database proprietor, the third age corresponding to the second time; 
   a model corrector to:
 apply a window function to the second ages to determine a distribution of ages based on the third age of the audience member as a mean of the distribution; 
 multiply window values of the second ages by respective ones of the first ages to determine corrected first age components; 
 divide a sum of the corrected first age components by a sum of the window values to determine an estimated age of the audience member at the first time; and 
 determine the corrected age of the audience member at the 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; and 
   a ratings determiner to determine ratings information for the media by attributing impressions and audience counts to the media using the corrected age of the audience member instead of the third age.   
     
     
         9 . The apparatus as defined in  claim 8 , wherein the model corrector is to apply the window function by applying a probability density function based on a Gaussian distribution. 
     
     
         10 . The apparatus as defined in  claim 8 , wherein the age predictor is to estimate the first ages by determining a predicted age probability density function. 
     
     
         11 . The apparatus as defined in  claim 10 , wherein the age predictor is to estimate the second ages by applying an aging factor to an age bucket in the predicted age probability density function. 
     
     
         12 . The apparatus as defined in  claim 8 , wherein the model corrector is to select the audience members from a larger set of audience members based on the second ages being within an age range, the age range being based on the third age. 
     
     
         13 . The apparatus as defined in  claim 8 , wherein the first impressions collector is to transmit audience measurement entity identifiers to the devices in response to at least a portion of the messages, the age predictor to estimate of the third age in response to determining, based on the audience member not being associated with an audience measurement entity identifier, that the audience member has not been previously identified. 
     
     
         14 . The apparatus as defined in  claim 8 , wherein the first impressions collector is to send re-direct messages in response to at least a portion of the messages, the re-direct messages to cause at least a portion of the devices to send third messages to the database proprietor, the first demographic information being received based on the third messages. 
     
     
         15 . A tangible computer readable storage medium comprising computer readable instructions which, when executed, cause a processor to at least:
 collect messages indicating first impressions of a media item delivered to devices via the Internet, the messages identifying the media item presented at the devices;   access first demographic information describing first numbers of impressions of the media item and first numbers of audience members attributed to respective demographic groups by a database proprietor, the first numbers of the impressions and the first numbers of audience members corresponding to the first impressions of the media;   estimate first ages of the audience members based on the first demographic information, the first ages corresponding to a first time;   estimate second ages of the audience members based on the first demographic information, the second ages corresponding to a second time after the first time;   estimate a third age of an audience member who is not included in the audience members from the database proprietor, the third age corresponding to the second time;   estimate a corrected age of the audience member at the second time by:
 applying a window function to the second ages to determine a distribution of ages based on the third age of the audience member as a mean of the distribution; 
 multiply window values of the second ages by respective ones of the first ages to determine corrected first age components; 
 summing the corrected first age components and 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 the first time; and 
 determining the corrected age of the audience member at the 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; and 
   determine ratings information for the media by attributing impressions and audience counts to the media using the corrected age of the audience member instead of the third age.   
     
     
         16 . The tangible computer readable storage medium as defined in  claim 15 , wherein the window function includes a probability density function based on a Gaussian distribution. 
     
     
         17 . The tangible computer readable storage medium as defined in  claim 15 , wherein the instructions are to cause the processor to estimate the first ages by determining a predicted age probability density function. 
     
     
         18 . The tangible computer readable storage medium as defined in  claim 17 , wherein the instructions are to cause the processor to estimate the second ages by applying an aging factor to an age bucket in the predicted age probability density function. 
     
     
         19 . The tangible computer readable storage medium as defined in  claim 15 , wherein the instructions are further to cause the processor to select the audience members from a larger set of audience members based on the second ages being within an age range, the age range being based on the third age. 
     
     
         20 . The tangible computer readable storage medium as defined in  claim 15 , wherein the instructions are further to cause the processor to transmit audience measurement entity identifiers to the devices in response to at least a portion of the messages, the instructions to cause the processor to estimate the third age in response to determining, based on the audience member not being associated with an audience measurement entity identifier, that the audience member has not been previously identified.