Estimating demographic compositions of television audiences

• US 8,984,547 B2
• Filed: 04/11/2011
• Issued: 03/17/2015
• Est. Priority Date: 04/11/2011
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Lambert, Diane, Zigmond, Daniel J.
• Primary Examiner(s): IDOWU, OLUGBENGA O
• Application Number:
  US13/083,663
• Publication Number:
  US 20120260278A1
• Time in Patent Office: 1,436 Days
• Field of Search: 725/9, 725/10, 725/12
• US Class Current: 725/9
• CPC Class Codes: H04H 60/45 for identifying users
• CPC Class Codes: H04H 60/66 for using the result on distributors' side
• CPC Class Codes: 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
• 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

Abstract

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for estimating demographic compositions of television audiences. In an aspect, audience demographics are estimated from viewing device log records and household demographic data describing demographic segments to which members of the households belong.

Claims

1. A system, comprising:

a data processing apparatus; and

software stored on a computer storage apparatus and comprising instructions executable by the data processing apparatus and upon such execution cause the data processing apparatus to perform operations comprising;

accessing viewer model data describing, for each of a plurality of demographic segments and for each of plurality of channels at a plurality of time blocks, a probability that a member of the demographic segment was viewing the channel at the time block, the viewer model data being generated using data for one person households, each one person household having only one member;

accessing household demographic data describing, for each of a plurality of households, one or more members of the household and, for each member of the household, one of the demographic segments to which the member belongs;

generating household model data from the viewer model data and the household demographic data, the household model data describing, for each of the households and for each demographic segment of the one or more members of the household, an expected number of viewers belonging to the demographic segment for each of the channels at each of the time blocks, wherein generating household model data from the viewer model data and the household demographic data comprises, for each channel at each time block, and for each household and for each demographic segment of members of the household;

determining a probability that at least one member of the household was viewing the channel at the time block; and

determining a ratio of the probability that one or more members belonging to the demographic segment was viewing the channel at the time block to the probability that at least one member of the household was viewing the channel at the time block; and

generating audience model data from the household model data, the audience model data describing, for each demographic segment, an estimated fraction of an audience belonging to the demographic segment for each of the channels at each of the time blocks.

2. The system of claim 1, wherein the instructions cause the data processing apparatus to perform operations comprising:

accessing channel tune data describing, for each of a plurality of viewing devices, channel tunes for viewing device, each channel tune specifying a channel to which the viewing device tuned to and a time that the viewing device tuned to the channel, each of the viewing devices being associated with a corresponding household; and

generating viewer model data from the channel tune data and the household demographic data.

3. The system of claim 2, wherein generating viewer model data from the channel tune data and the household demographic data comprises:

for each of the channels and each of the time blocks, and for each demographic segments;

determining a first number of one person households tuned to the channel at the time block, each of the number of one person households having a member belonging to the demographic;

determining a total number of one person households, each of the total number of one person households having a member belonging to the demographic; and

determining the probability that the member of the demographic segment was viewing the channel at the time block based on a ratio of the first number to the total number.

4. The system of claim 2, wherein determining the probability that at least one member of the household belonging to the demographic segment was viewing the channel at the time block comprises:

for each viewing device associated with the household for which channel tune data corresponding to the channel and time block exist, determining a respective probability that at least member of the household was watching the channel at the time block; and

generating the probability that at least one member of the household was viewing the channel at the time block from the respective probabilities.

5. The system of claim 4, wherein the probability that at least one member of the household was viewing the channel at the time block from the respective probabilities comprises generating a product of the respective probabilities.

6. The system of claim 1, wherein generating audience model data from the household model data comprises, for each channel at each time block, for each demographic segment:

determining a ratio of an estimated number of viewers of the channel at the time block belonging to the demographic segment to a total number of viewers of the channel at the time block.

7. The system of claim 6, wherein determining the ratio of the estimated number of viewers of the channel at the time block belonging to the demographic segment to the total number of viewers of the channel at the time block comprises:

summing the expected number of viewers belonging to the demographic segment for the channel at the time block to generate a first sum;

summing the expected number of viewers for the channel at the time block to generate a second sum; and

dividing the first sum by the second sum.

8. A computer implemented method, comprising:

accessing viewer model data describing, for each of a plurality of demographic segments and for each of plurality of channels at a plurality of time blocks, a probability that a member of the demographic segment was viewing the channel at the time block, the viewer model data being generated using data for one person households, each one person household having only one member;

accessing household demographic data describing, for each of a plurality of households, one or more members of the household and, for each member of the household, one of the demographic segments to which the member belongs;

generating, in a data processing apparatus, household model data from the viewer model data and the household demographic data, the household model data describing, for each of the households and for each demographic segment of the one or more members of the household, an expected number of viewers belonging to the demographic segment for each of the channels at each of the time blocks, wherein generating household model data from the viewer model data and the household demographic data comprises, for each channel at each time block, and for each household and for each demographic segment of members of the household;

determining a probability that at least one member of the household was viewing the channel at the time block; and

determining a ratio of the probability that one or more members belonging to the demographic segment was viewing the channel at the time block to the probability that at least one member of the household was viewing the channel at the time block; and

generating, in the data processing apparatus, audience model data from the household model data, the audience model data describing, for each demographic segment, an estimated fraction of an audience belonging to the demographic segment for each of the channels at each of the time blocks.

9. The method of claim 8, wherein the instructions cause the data processing apparatus to perform operations comprising:

accessing channel tune data describing, for each of a plurality of viewing devices, channel tunes for viewing device, each channel tune specifying a channel to which the viewing device tuned to and a time that the viewing device tuned to the channel, each of the viewing devices being associated with a corresponding household; and

generating viewer model data from the channel tune data and the household demographic data.

10. The method of claim 9, wherein generating viewer model data from the channel tune data and the household demographic data comprises:

for each of the channels and each of the time blocks, and for each demographic segments;

determining a first number of one person households tuned to the channel at the time block, each of the number of one person households having a member belonging to the demographic;

determining a total number of one person households, each of the total number of one person households having a member belonging to the demographic; and

determining the probability that the member of the demographic segment was viewing the channel at the time block based on a ratio of the first number to the total number.

11. The method of claim 9 wherein determining the probability that at least one member of the household belonging to the demographic segment was viewing the channel at the time block comprises:

for each viewing device associated with the household for which channel tune data corresponding to the channel and time block exist, determining a respective probability that at least member of the household was watching the channel at the time block; and

generating the probability that at least one member of the household was viewing the channel at the time block from the respective probabilities.

12. The method of claim 11, wherein the probability that at least one member of the household was viewing the channel at the time block from the respective probabilities comprises generating a product of the respective probabilities.

13. The method of claim 8, wherein generating audience model data from the household model data comprises, for each channel at each time block, for each demographic segment:

determining a ratio of an estimated number of viewers of the channel at the time block belonging to the demographic segment to a total number of viewers of the channel at the time block.

14. The method of claim 13, wherein determining the ratio of the estimated number of viewers of the channel at the time block belonging to the demographic segment to the total number of viewers of the channel at the time block comprises;

summing the expected number of viewers belonging to the demographic segment for the channel at the time block to generate a first sum;

summing the expected number of viewers for the channel at the time block to generate a second sum; and

dividing the first sum by the second sum.

15. A system, comprising:

a data storage device storing viewer model data and household demographic data, the viewer model data describing, for each of a plurality of demographic segments and for each of plurality of channels at a plurality of time block, the viewer model data being generated using data for one person households, each one person household having only one member, and the viewer model data defining a probability that a member of the demographic segment was viewing the channel at the time block, and the household demographic data describing, for each of a plurality of households, one or more members of the household and, for each member of the household, one of the demographic segments to which the member belongs;

means for generating household model data from the viewer model data and the household demographic data, the household model data describing, for each of the households and for each demographic segment of the one or more members of the household, an expected number of viewers belonging to the demographic segment for each of the channels at each of the time blocks, wherein generating household model data from the viewer model data and the household demographic data comprises, for each channel at each time block, and for each household and for each demographic segment of members of the household;

determining a probability that at least one member of the household was viewing the channel at the time block; and

determining a ratio of the probability that one or more members belonging to the demographic segment was viewing the channel at the time block to the probability that at least one member of the household was viewing the channel at the time block; and

means for generating audience model data from the household model data, the audience model data describing, for each demographic segment, an estimated fraction of an audience belonging to the demographic segment for each of the channels at each of the time blocks.

16. The system of claim 15, wherein:

the data storage device stores channel tune data describing, for each of a plurality of viewing devices, channel tunes for viewing device, each channel tune specifying a channel to which the viewing device tuned to and a time that the viewing device tuned to the channel, each of the viewing devices being associated with a corresponding household; and

further comprising means for generating the viewer model data from the channel tune data and the household demographic data.

17. A non-transitory computer readable medium storing instructions that, when executed by one or more data processing apparatus, cause the one or more data processing apparatus to perform operations comprising:

accessing viewer model data describing, for each of a plurality of demographic segments and for each of plurality of channels at a plurality of time blocks, a probability that a member of the demographic segment was viewing the channel at the time block, the viewer model data being generated using data for one person households, each one person household having only one member;

accessing household demographic data describing, for each of a plurality of households, one or more members of the household and, for each member of the household, one of the demographic segments to which the member belongs;

generating household model data from the viewer model data and the household demographic data, the household model data describing, for each of the households and for each demographic segment of the one or more members of the household, an expected number of viewers belonging to the demographic segment for each of the channels at each of the time blocks, wherein generating household model data from the viewer model data and the household demographic data comprises, for each channel at each time block, and for each household and for each demographic segment of members of the household;

determining a probability that at least one member of the household was viewing the channel at the time block; and

determining a ratio of the probability that one or more members belonging to the demographic segment was viewing the channel at the time block to the probability that at least one member of the household was viewing the channel at the time block; and

generating audience model data from the household model data, the audience model data describing, for each demographic segment, an estimated fraction of an audience belonging to the demographic segment for each of the channels at each of the time blocks.

18. The computer readable medium of claim 17, wherein the operations further comprise:

accessing channel tune data describing, for each of a plurality of viewing devices, channel tunes for viewing device, each channel tune specifying a channel to which the viewing device tuned to and a time that the viewing device tuned to the channel, each of the viewing devices being associated with a corresponding household; and

generating viewer model data from the channel tune data and the household demographic data.