Methods and apparatus to detect and rectify false set top box tuning data
Methods, apparatus, systems, and articles of manufacture are disclosed to rectify false set top box tuning data. Disclosed examples methods include identifying in return path data a first group of set top boxes classified as likely to exhibit machine events in tuning data of the return path data more frequently than a second group of set top boxes represented in the return path data. Additionally, in some examples, the method includes determining whether the first group of set top boxes includes a machine event based on a pattern of the tuning data in the return path data for respective ones of the first group of set top boxes and improving an accuracy of return path data by rectifying the machine event.
1. An apparatus to remove false tuning data, the apparatus comprising:
a set top box identifier to identify in return path data a first group of set top boxes classified as likely to exhibit machine events in tuning data of the return path data more frequently than a second group of set top boxes represented in the return path data;
a pattern evaluator to determine whether the first group of set top boxes includes a machine event based on a pattern of the tuning data in the return path data for respective ones of the first group of set top boxes, the pattern based on comparison of (i) the tuning data in the return path data for respective ones of a third group of set top boxes identified as also associated with corresponding meters that are to report panelist exposure data, and (ii) the panelist exposure data obtained from the meters, the meters separate from the third group of set top boxes; and
a false tuning event rectifier to improve an accuracy of return path data by rectifying the machine event.
2. The apparatus of claim 1 , wherein the pattern evaluator is to evaluate the pattern of the tuning data to determine a difference between a first percentage of the first group of set top boxes having tuning events in the return path data during a first time interval and a second percentage of the second group of set top boxes having tuning events in the return path data during the first time interval satisfying a threshold.
3. The apparatus of claim 2 , wherein the false tuning event rectifier rectifies the machine event by removing first tuning data associated with the first time interval in the return path data corresponding to the first percentage of the first group of set top boxes.
4. The apparatus of claim 1 , wherein the pattern of the tuning data includes a feature activated machine event producing false tuning data.
5. The apparatus of claim 4 , wherein the false tuning event rectifier is to rectify the machine event by identifying and removing the false tuning data while maintaining veridical tuning data in the return path data.
6. The apparatus of claim 4 , wherein the false tuning event rectifier is to identify the feature activated machine event based on a prerequisite of the feature activated machine event, the prerequisite determined based on the comparison of the return path data to the panelist exposure data obtained from the meters.
7. The apparatus of claim 6 , wherein panelist exposure data is classified as veridical.
8. A method to remove false tuning data, the method comprising:
identifying, by executing an instruction with a processor, in return path data a first group of set top boxes classified as likely to exhibit machine events in tuning data of the return path data more frequently than a second group of set top boxes represented in the return path data;
determining, by executing an instruction with the processor, whether the first group of set top boxes includes a machine event based on a pattern of the tuning data in the return path data of the first group of set top boxes, the pattern based on comparison of (i) the tuning data in the return path data for respective ones of a third group of set top boxes identified as also associated with corresponding meters that are to report panelist exposure data, and (ii) the panelist exposure data obtained from the meters, the meters separate from the third group of set top boxes; and
improving an accuracy of return path data by rectifying the machine event.
9. The method of claim 8 , wherein the determining of whether the first group of set top boxes includes the machine event based on the pattern of the tuning data includes determining a difference between a first percentage of the first group of set top boxes having tuning events in the return path data during a first time interval and a second percentage of the second group of set top boxes having tuning events in the return path data during the first interval.
10. The method of claim 9 , wherein the rectifying of the machine event includes removing first tuning data associated with the first time interval in the return path data corresponding to the first percentage of the first group of set top boxes.
11. The method of claim 8 , wherein the pattern of the tuning data includes a feature activated machine event producing false tuning data.
12. The method of claim 11 , wherein the rectifying of the machine event includes identifying and removing the false tuning data while maintaining veridical tuning data in the return path data.
13. The method of claim 11 , wherein the feature activated machine event is identified based on a prerequisite of the feature activated machine event, the prerequisite determined by the comparing of the return path data to the panelist exposure data obtained from the meters.
14. The method of claim 13 , wherein panelist exposure data is classified as veridical.
15. A non-transitory computer readable medium comprising computer readable instructions which, when executed, cause a processor to at least:
identify in return path data a first group of set top boxes classified as likely to exhibit machine events in tuning data of the return path data more frequently than a second group of set top boxes represented in the return path data;
determine whether the first group of set top boxes includes a machine event based on a pattern of the tuning data in the return path data for respective ones of the first group of set top boxes, the pattern based on comparison of (i) the tuning data in the return path data for respective ones of a third group of set top boxes identified as also associated with corresponding meters that are to report panelist exposure data, and (ii) the panelist exposure data obtained from the meters, the meters separate from the third group of set top boxes; and
improve an accuracy of return path data by rectifying the machine event.
16. The non-transitory computer readable medium of claim 15 , wherein to determine whether the first group of set top boxes includes the machine event based on the pattern of the tuning data, the instructions, when executed, cause the processor to determine a difference between a first percentage of the first group of set top boxes having tuning events in the return path data during a first time interval and a second percentage of the second group of set top boxes having tuning events in the return path data during the first interval satisfying a threshold.
17. The non-transitory computer readable medium of claim 16 , wherein to rectify the machine event, the instructions when executed cause the processor to remove first tuning data associated with first time interval in the return path data corresponding to the first percentage of the first group of set top boxes.
18. The non-transitory computer readable medium of claim 15 , wherein the pattern of the tuning data includes a feature activated machine event producing false tuning data.
19. The non-transitory computer readable medium of claim 18 , wherein to rectify the machine event, the instructions, when executed, cause the processor to identify and remove the false tuning data while maintaining veridical tuning data in the return path data.
20. The non-transitory computer readable medium of claim 18 , wherein the instructions, when executed, cause the processor to identify the feature activated machine event based on a prerequisite of the feature activated machine event, the prerequisite determined based on the comparison of the return path data to the panelist exposure data obtained from the meters.