U.S. patent application number 16/035527 was filed with the patent office on 2019-04-18 for deterministic household assignment model.
The applicant listed for this patent is comScore, Inc.. Invention is credited to Josh Chasin, John Clougherty, Cameron Meierhoefer, Frank Pecjak, Michael J. Vinson.
Application Number | 20190116392 16/035527 |
Document ID | / |
Family ID | 66096259 |
Filed Date | 2019-04-18 |
United States Patent
Application |
20190116392 |
Kind Code |
A1 |
Vinson; Michael J. ; et
al. |
April 18, 2019 |
DETERMINISTIC HOUSEHOLD ASSIGNMENT MODEL
Abstract
Techniques for projecting household-level viewing events are
described herein. Population data may be accessed including classes
of a plurality of demographic attributes for households in a
market. Representative household units (RHUs) may be generated, and
the RHUs may be assigned a class for each of the demographic
attributes and a quota based on the demographic attributes of a
plurality of panelist households. Each of the panelist households
may be assigned to one of the RHUs based on at least one panelist
classes matching the classes for respective demographic attributes
of the RHU, and the number of matching panelist households assigned
to each of the RHU may be based on the quota. Panelist viewing data
representing viewing events associated with the panelist household
may be accessed. A report may be generated with the classes of the
RHUs and the panelist viewing data of the assigned panelist
households.
Inventors: |
Vinson; Michael J.;
(Piedmont, CA) ; Pecjak; Frank; (Fairfax, VA)
; Clougherty; John; (Vienna, VA) ; Chasin;
Josh; (New York, NY) ; Meierhoefer; Cameron;
(Washington, DC) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
comScore, Inc. |
Reston |
VA |
US |
|
|
Family ID: |
66096259 |
Appl. No.: |
16/035527 |
Filed: |
July 13, 2018 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
62571823 |
Oct 13, 2017 |
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
H04H 60/52 20130101;
H04N 21/25883 20130101; G06Q 30/0242 20130101; H04H 60/45 20130101;
G06Q 30/0201 20130101 |
International
Class: |
H04N 21/258 20060101
H04N021/258; H04H 60/52 20060101 H04H060/52; G06Q 30/02 20060101
G06Q030/02 |
Claims
1. A system, comprising: at least one processor; and at least one
memory storing instructions that, when executed, cause the at least
one processor to: access population data including classes for each
of first and second demographic attributes of households in a
market; generate an array of representative household units (RHUs)
including a first RHU, wherein the RHUs are each assigned a class
for each of the first and second demographic attributes; access a
panelist class of each of the first and second demographic
attributes for first and second panelist households; match the
panelist classes of the first and second demographic attributes for
the first panelist household to the respective classes of the first
and second demographic attributes for a first RHU; assign the first
panelist household to the first RHU; determine that the panelist
classes of the first and second demographic attributes for the
second panelist household do not match the respective classes of
the first and demographic attributes for any RHU; match the
panelist class of the first demographic attribute for the second
panelist household to the class of the first demographic attribute
for the first RHU; assign the second panelist household to the
first RHU; access panelist viewing data representing viewing events
associated with the first and second panelist households; and
generate a report including the classes of the first RHU and the
panelist viewing data of the first and second panelist
households.
2. The system of claim 1, wherein the first demographic attribute
includes include one or more of an income of the household, a
language spoken in the household, a number of members of the
household, and a number of children of the household.
3. The system of claim 1, wherein the second demographic attribute
includes one or more of an age of at least one member of the
household, a gender of at least one member of the household, a race
of at least one member of the household, an ethnicity of at least
one member of the household, and an education level of at least one
member of the household.
4. The system of claim 1, wherein the population data includes
classes for a third demographic attributes of the households in the
market, the RHUs are each assigned a class for the third
demographic attribute, a panelist class of each of the first,
second, and third demographic attributes are accessed for a third
panelist household, and the instructions, when executed, further
cause the at least one processor to: determine that the panelist
classes of the first, second, and third demographic attributes for
the third panelist household do not match the respective classes of
the first, second, and third demographic attributes for any RHU;
determine that the panelist classes of the first and second
demographic attributes for the third panelist household do not
match the respective classes of the first, second, and third
demographic attributes for any RHU; match the panelist class of the
first demographic attribute for the third panelist household to the
class of the first RHU for the first demographic attribute; and
assign the third panelist household to the first RHU.
5. The system of claim 5, wherein the third demographic attribute
includes a number of television sets.
6. The system of claim 1, wherein each viewing event in the
panelist viewing data includes an identification of a media, an
advertisement, a website, an app, a network, and/or a program
associated with the viewing event and a time duration that the
panelist household was exposed to the viewing event.
7. The system of claim 1, wherein the viewing event occurs on one
or more of a television, a mobile phone, a tablet, and a smart
watch.
8. The system of claim 1, wherein the instructions, when executed,
further cause the at least one processor to generate a quota based
on the number of households with the demographic attributes of the
RHU relative to the number of households in the market, wherein the
number of matching panelist households assigned to each RHU is
based on the quota.
9. The system of claim 9, wherein the instructions, when executed,
further cause the at least one processor to stop assigning
panelists households to an RHU based on the number of matching
panelist households meeting the quota of the RHU.
10. The system of claim 9, wherein the instructions, when executed,
further cause the at least one processor to duplicate the viewing
data of the at least one first panelist households for an RHU based
on the number of matching panelist households assigned to the RHU
being less than the quota after the plurality of panelist
households are assigned.
11. The system of claim 1, wherein the instructions, when executed,
further cause the at least one processor to determine that the
panelist households are active based on viewing data accessed from
a predetermined period of time, wherein only active panelist
households are assigned to the RHUs.
12. The system of claim 1, wherein the population data is received
from one or more of a credit bureau and a census bureau.
13. A computer-implemented process, comprising: accessing
population data including classes for each of first and second
demographic attributes of households in a market; generating an
array of representative household units (RHUs) including a first
RHU, wherein the RHUs are each assigned a class for each of the
first and second demographic attributes; accessing a panelist class
of each of the first and second demographic attributes for first
and second panelist households; matching the panelist classes of
the first and second demographic attributes for the first panelist
household to the respective classes of the first and second
demographic attributes for a first RHU; assigning the first
panelist household to the first RHU; determining that the panelist
classes of the first and second demographic attributes for the
second panelist household do not match the respective classes of
the first and demographic attributes for any RHU; matching the
panelist class of the first demographic attribute for the second
panelist household to the class of the first demographic attribute
for the first RHU; assigning the second panelist household to the
first RHU; accessing panelist viewing data representing viewing
events associated with the first and second panelist households;
and generating a report including the classes of the first RHU and
the panelist viewing data of the first and second panelist
households.
14. The computer-implemented process of claim 13, wherein the first
demographic attribute includes include one or more of an income of
the household, a language spoken in the household, a number of
members of the household, and a number of children of the
household.
15. The computer-implemented process of claim 13, wherein the
second demographic attribute includes one or more of an age of at
least one member of the household, a gender of at least one member
of the household, a race of at least one member of the household,
an ethnicity of at least one member of the household, and an
education level of at least one member of the household.
16. The computer-implemented process of claim 13, wherein the
population data includes classes for a third demographic attributes
of the households in the market, the RHUs are each assigned a class
for the third demographic attribute, a panelist class of each of
the first, second, and third demographic attributes are accessed
for a third panelist household, and the process further includes:
determining that the panelist classes of the first, second, and
third demographic attributes for the third panelist household do
not match the respective classes of the first, second, and third
demographic attributes for any RHU; determining that the panelist
classes of the first and second demographic attributes for the
third panelist household do not match the respective classes of the
first, second, and third demographic attributes for any RHU;
matching the panelist class of the first demographic attribute for
the third panelist household to the class of the first RHU for the
first demographic attribute; and assigning the third panelist
household to the first RHU.
17. A computer-readable medium comprising computer-executable
instructions which, when executed by at least one processor, cause
the at least one processor to: access population data including
classes for each of first and second demographic attributes of
households in a market; generate an array of representative
household units (RHUs) including a first RHU, wherein the RHUs are
each assigned a class for each of the first and second demographic
attributes; access a panelist class of each of the first and second
demographic attributes for first and second panelist households;
match the panelist classes of the first and second demographic
attributes for the first panelist household to the respective
classes of the first and second demographic attributes for a first
RHU; assign the first panelist household to the first RHU;
determine that the panelist classes of the first and second
demographic attributes for the second panelist household do not
match the respective classes of the first and demographic
attributes for any RHU; match the panelist class of the first
demographic attribute for the second panelist household to the
class of the first demographic attribute for the first RHU; assign
the second panelist household to the first RHU; access panelist
viewing data representing viewing events associated with the first
and second panelist households; and generate a report including the
classes of the first RHU and the panelist viewing data of the first
and second panelist households.
18. The computer-readable medium of claim 17, wherein the first
demographic attribute includes include one or more of an income of
the household, a language spoken in the household, a number of
members of the household, and a number of children of the
household.
19. The computer-readable medium of claim 17, wherein the second
demographic attribute includes one or more of an age of at least
one member of the household, a gender of at least one member of the
household, a race of at least one member of the household, an
ethnicity of at least one member of the household, and an education
level of at least one member of the household.
20. The computer-readable medium of claim 17, wherein the
population data includes classes for a third demographic attributes
of the households in the market, the RHUs are each assigned a class
for the third demographic attribute, a panelist class of each of
the first, second, and third demographic attributes are accessed
for a third panelist household, and the instructions, when
executed, further cause the at least one processor to: determine
that the panelist classes of the first, second, and third
demographic attributes for the third panelist household do not
match the respective classes of the first, second, and third
demographic attributes for any RHU; determine that the panelist
classes of the first and second demographic attributes for the
third panelist household do not match the respective classes of the
first, second, and third demographic attributes for any RHU; match
the panelist class of the first demographic attribute for the third
panelist household to the class of the first RHU for the first
demographic attribute; and assign the third panelist household to
the first RHU.
Description
CROSS REFERENCE TO RELATED APPLICATIONS
[0001] This application claims the benefit of U.S. Provisional
Patent App. No. 62/571,823, filed Oct. 13, 2017, the disclosure of
which is hereby incorporated by reference herein.
TECHNICAL FIELD
[0002] The present disclosure generally relates to systems and
methods for determining program viewership, and more particularly
to systems and methods for determining the demographics of viewers
of programs using deterministic household assignment.
BACKGROUND
[0003] Advertising relies on program and network viewership data in
order to determine the reach and impressions of targeted
advertisement. Advertisers are interested in numbers of viewers as
well as the demographics of viewers in order to effectively manage
advertising timing and content. Understanding audience viewing and
habits may be useful in supporting planning, buying, and selling
advertising.
[0004] Therefore, there is a need for improved systems and methods
for determining the demographics of viewers of content using
deterministic household assignment.
SUMMARY
[0005] Techniques for projecting household-level viewing events are
described herein. Initially, population data may be accessed
including classes of a plurality of demographic attributes for
households in a market. An array of representative household units
(RHUs) may be generated, and the RHUs may be assigned a class for
each of the demographic attributes and a quota based on the
demographic attributes of the population data. A panelist class may
be accessed for each of the demographic attributes of a plurality
of panelist households. Each of the panelist households may be
assigned to one of the RHUs based on at least one of the panelist
classes matching the classes for respective demographic attributes
of the RHU, and the number of matching panelist households assigned
to each of the RHU may be based on the quota. Panelist viewing data
representing viewing events associated with the panelist household
may be accessed. A report may be generated with the classes of the
RHUs and the panelist viewing data of the assigned panelist
households.
[0006] In some embodiments, assigning the panelist households to
one of the RHUs may be based on each of the panelist classes
matching the classes for the respective demographic attributes of
the RHU. In some embodiments, the panelist viewing data may include
an identification of a displayed media, advertisement, website,
app, network and/or program and a time duration of the viewing
event. In some embodiments, the viewing event may be displayed on
one or more of a television, a mobile phone, a tablet, a laptop
computer, a desktop computer, smart appliances, and a smart watch.
In some embodiments, the demographic attributes may include one or
more of a television stratum, a presence of a DVR, and a number of
television sets.
[0007] In some embodiments, the demographic attributes may include
one or more of an age of at least one member of the household, a
race of at least one member of the household, an ethnicity of at
least one member of the household, and an education level of at
least one member of the household. In some embodiments, the
demographic attributes may include one or more of an income of the
household, a language spoken in the household, a number of members
of the household, and a number of children of the household.
[0008] In some embodiments, the instructions, when executed, may
further cause the at least one processor to determine that the
panelist households are active based on viewing data accessed from
a predetermined period of time, wherein only active panelist
households are assigned to the RHUs. In some embodiments, the
instructions, when executed, may further cause the at least one
processor to generate the quota based on the number of households
with the demographic attributes of the RHU relative to the number
of households in the market. In some embodiments, the instructions,
when executed, may further cause the at least one processor to stop
assigning panelists households to an RHU based on the number of
matching panelist households meeting the quota of the RHU.
[0009] In some embodiments, the instructions, when executed, may
further cause the at least one processor to duplicate viewing data
of the panelists households for an RHU based on the number of
matching panelist households assigned to the RHU being less than
the quota after the plurality of panelist households are assigned.
In some embodiments, the population data may be received from one
or more of a credit bureau and a census bureau. In some
embodiments, the instructions, when executed, may further cause the
at least one processor to receive a known value of viewing data for
the market, and adapt the panelist viewing data for at least one of
the RHUs based on the known value of the viewing data. In some
embodiments, the instructions, when executed, may further cause the
at least one processor to receive second population data for at
least one second market, and scale the panelist viewing data for at
least one of the RHUs of the market based on a relative size of the
population data compared to the second population data.
[0010] The assignment of the panelist households may be based on
first and second demographic attributes. The RHUs may be assigned a
class for each of the first and second demographic attributes.
Panelist classes of the first and second demographic attributes for
a first panelist household may be matched to the respective classes
of the first and second demographic attributes for a first RHU. The
first panelist class may then be assigned to the first RHU. The
panelist classes of the first and second demographic attributes for
a second panelist household may be determined to not match the
respective classes of the first and second demographic attributes
for any RHU.
[0011] The panelist class of the first demographic attribute for
the second panelist household may be matched to the class of the
first demographic attribute for the first RHU. The second panelist
household may then be assigned to the first RHU. The report may be
generated including the classes of the first RHU and the panelist
viewing data of the first and second panelist households.
[0012] In some embodiments, the first demographic attribute may
include one or more of an income of the household, a language
spoken in the household, a number of members of the household, and
a number of children of the household. In some embodiments, the
second demographic attribute may include one or more of an age of
at least one member of the household, a gender of at least one
member of the household, a race of at least one member of the
household, and an education level of at least one member of the
household.
[0013] In some embodiments, the population data may include classes
for a third demographic attributes of the households in the market,
the RHUs may be each assigned a class for the third demographic
attribute. A panelist class for each of the first, second, and
third demographic attributes may be accessed for a third panelist
household. Accordingly, the instructions, when executed, may
further cause the at least one processor to: determine that the
panelist classes of the first, second, and third demographic
attributes for the third panelist household do not match the
respective classes of the first, second, and third demographic
attributes for any RHU; determine that the panelist classes of the
first and second demographic attributes for the third panelist
household do not match the respective classes of the first and
second demographic attributes for any RHU; match the panelist class
of the first demographic attribute for the third panelist household
to the class of the first RHU for the first demographic attribute;
and assign the third panelist household to the first RHU. In some
embodiments, the third demographic attribute may include a number
of television sets.
[0014] Embodiments of any of the described techniques may include a
method or process, an apparatus, a device, a machine, a system, or
instructions stored on a computer-readable storage device. The
details of particular embodiments are set forth in the accompanying
drawings and description below. Other features will be apparent
from the following description, including the drawings, and the
claims.
BRIEF DESCRIPTION OF THE DRAWINGS
[0015] FIG. 1 illustrates an exemplary system in which viewership
information may be collected and processed to determine and/or
estimate panelist viewing data and assign panelist households.
[0016] FIG. 2 illustrates an exemplary system in which
household-level viewing data can be used to project market-level
data through household assignment.
[0017] FIG. 3 is a flow chart illustrating an exemplary process for
generating a report with market-level data projected from
household-level viewing data.
[0018] FIG. 4 illustrates a schematic illustrating an exemplary
array of demographic attributes for Residential Household Units
(RHUs).
[0019] FIG. 5 illustrates a schematic illustrating quotas of a
functional cross-section of RHUs.
[0020] FIG. 6 is a flow chart illustrating an exemplary process for
assigning panelist households to RHUs.
DETAILED DESCRIPTION
[0021] The methodology involves creating an array of Representative
Household Units (RHUs), demographically and behaviorally balanced
to represent a geographic market. These RHUs may become the
recipient dataset into which disparate donor datasets are assigned.
The RHUs may be proportional to the overall geographic market
(e.g., a ratio of 1 per 10 households in the market) and assigned
demographic and/or behavior characteristics of the market at large.
The behavior and/or demographic characteristics of the market can
be established from population data received over a network from a
trusted third party, such as data from one or more census bureaus
and/or credit bureaus. The population data may provide invaluable
granular data on the general make-up of a market at scale, but
viewership data at the market-scale is often incomplete and/or
inaccurate.
[0022] Therefore, real data from different datasets at a smaller
scale (e.g., household-scale, person-scale, and/or device-scale)
may be incorporated. The real data may be obtained from content
viewed on devices, such as televisions (TVs), tablets, mobile
phones, and/or other electronic devices. The viewing data may be
accessed from "panelist households" who, in at least some cases,
have agreed to have their viewing behavior actively and/or
passively, directly monitored. For example, television viewership
of the panelist household may be measured by a set-top-box (STB)
logging viewing activity. Due to the direct access to the STB, the
viewing data may provide a rich dataset accurately detailing
viewing events of the members of the panelist households. The
viewing data and/or panelist households may also be maintained
current by providing a threshold of activity. For example, only
panelist households with viewing data within the past 30 or 60 days
may be assigned to the RHUs. The viewing data may also incorporate
other devices (e.g., an iPad) associated with the panelist
household by connecting to a household network router. The panelist
viewing data may further include viewing data of devices (e.g.
mobile phones) that are registered to a member of the panelist
household and accessed data through a cellular network. The
panelist viewing data may include return path data (RPD), which is
a passive data collection technique that collects any user/viewer
activity collected from a device defined by a start time and a
duration. The panelist viewing data provides a rich-set of data of
the media that real individuals are consuming. However, since the
panelist viewing data is mainly received from a self-selecting
population (panelists), the panelist viewing data itself does not
provide an indication of viewing data of an overall geographic
market.
[0023] Household data may also be maintained and include
demographic attributes for each of the panelist households and
associated members, such as household income, number of members of
the household, the gender of the members, the age of the members,
and strata of television access. The demographic attributes from
the panelist households may be mapped to the RHUs of the population
data to populate the RHUs with real panelist households. The
viewing data of the panelist households may be assigned to the RHUs
based on matching demographic attributes, such that actual viewing
data from a real household may be assigned to a demographically,
behaviorally matched RHU. Therefore, the viewing data of the
panelist households may be proportionally calibrated to accurately
represent the market population.
[0024] To maintain the demographic make-up of the market, each of
the RHUs may be assigned a quota based on the overall population of
the market. For example, if, according to the population data,
households of two members (Male aged 25-34 years old; Female aged
25-34 years old) have a higher population than households of one
member (Male aged 25-34 years old), RHUs of the first households
may be assigned a proportionally higher quota than the RHUs of the
second households. Assignment of the panelist households to the
RHUs may be repeated until the respective quota is reached and
stopped after the quota is reached, while the remaining RHUs may be
populated until the respective quota is reached.
[0025] Perfect matches between the panelist households to the RHUs
may be prioritized. However, some degree of inference and
ascription may be required during the assignment phase in cases
where either (1) there are not enough exact matches for the RHUs
and/or (2) the raw unit assigned is not behaviorally complete
(e.g., the donor household doesn't have as many set top box devices
as have been designated for the RHU; or if DVR records are absent).
In these cases, the match may be constrained to as many
demographics and attributes as available, and the best remaining
raw record may be assigned based on closest distance behaviorally.
Assignments made to a particular RHU may be maintained if the raw
input is still available (e.g., not attrited in the data set). This
maintains longitudinal consistency in behavioral profiles. The goal
of the assignment process is to achieve most of the target unique
visitation and additive totals directly without additional
allocations or adjustments (the target is 85% of market-network-day
hours for TV). When the data is insufficient for the RHUs, matching
of panelist households may be relaxed by removing certain
demographic attributes from the requirements of being assigned to
an RHU. The core demographic attributes may be prioritized to
maintain the integrity of the core demographic definition of the
RHU. For example, one or more device attributes (e.g., number of
devices) may be relaxed first, one or more member attributes (e.g.,
age of the members, race of the members, and/or education level of
the members) may be relaxed second, and one or more household
attributes (e.g., income of the household, language spoken in the
household, number of members, and/or number of children) may be
relaxed third. In some embodiments, if RHUs do not have sufficient
matches of panelist households to reach the quota, viewing data of
previously assigned panelist households may be duplicated to ensure
that the panelist viewing data of the RHUs is correctly
proportional.
[0026] After assignment of the panelist households to the RHUs,
viewing data received directly from the panelist households may be
readily assigned to the respective RHU. The result is a massive
respondent-level dataset that projects back to universe, matches
individual currency measures from component data sets. Reports may
be generated from the data and displayed to provide an accurate
measure of demographic-based viewing data for the overall market to
the content providers, advertisers, and others. The reports may
therefore be used to estimate the number of viewing people and/or
households of a particular demographic for a particular program,
advertisement, sporting event, and/or other content item.
[0027] The resultant dataset and reports may further be adapted
based on available known and trusted reported results from third
parties. For example, the system may compare the projected assigned
viewing data and additives to the individual platform targets
(e.g., available as either census totals or a combination of census
and enumeration) and may assign individual events into the
appropriate RHUs to hit the targets. This assures that the
projected results in the system match individual-currency reported
results. The added events may be actual events from the pool of
previously unassigned activity. The specific rules and targets for
triggering the adaptive process may be determined by the individual
platforms but generally an incremental event may be added to an RHU
that shows a high propensity for the type of event and has a gap in
activity that can accept the event. The result of the adaptive
process is an individual respondent level profile that is
empirically valid and the aggregation of those profiles achieves
the core platform targets (e.g., market-network-day hours for TV).
This adaptive process may be run daily and does not guarantee
longitudinal consistency of assignment of events across RHUs or raw
households (that is, a representative RHU is not guaranteed to get
an adaptive event every day, nor will it get an event from the same
source household/person/device every day).
[0028] FIG. 1 illustrates an example of a system 100 in which
viewership information may be collected and processed to determine
and/or estimate audience measurement data. The system 100 may
include a number of panelist households 101, such as the
illustrated panelist household assigned the Identification Number
1231. The panelist households 101 may include one or more panelist
devices 112 for viewing content by one or more members 102. The
panelist devices 112 may be embodied by and/or be connected with
any number of a television, a mobile phone, a tablet, a laptop
computer, a desktop computer, smart appliances, and/or a smart
watch. For example, the panelist device 112 may include a number of
different types of devices associated with the panelists household
101, such as a television in the household 101, a digital video
recorder (DVR) connected to the television, a set-top-box (STB)
associated connected to the television, and/or a home network
router.
[0029] The panelist devices 112 may record panelist viewing data
116 for viewing events displayed on the panelist devices 112 or an
associated display. The viewing event may indicate a media, an
advertisement, a website, an app, a network and/or a program
transmitted to the panelist device 112, and/or a time duration that
the panelist household 101 was exposed to the media, an
advertisement, a website, an app, a network and/or a program. The
panelist devices 112 may report the panelist viewing data 116 to a
usage collection server 114, and the panelist viewing data 116 may
be stored in a storage device 120. In addition to viewing events,
the panelist viewing data 116 may include data corresponding to the
panelist household 101, the panelist device 112, stream control
data, data representing content recorded by the panelist device
112, programs ordered on the panelist device 112 through an on
demand service, and/or data about when the panelist device 112 was
turned on or off. Other data about the status of the panelist
device 112 and user interaction with the panelist device 112 may
also be recorded and included in the panelist viewing data 116.
[0030] In some embodiments, the panelist devices 112 may include an
STB that transmits television programs to a display (e.g., a
television) from various stratum, such as over the air (OTA),
direct broadcast satellite (DBS), cable, and/or telephone companies
(telco). Thus, the panelist viewing data 116 may include tuning
data recorded by the STB indicating media, advertisements, website,
app, network and/or program being transmitted to the television and
a time duration. The panelist devices 112 may also include a
household network router of the panelist household 101 that
monitors access of a network (e.g., the Internet) by computers,
smart phones, and/or tablets in the household 101. The household
router may monitor viewing events of the household 101 and report
panelist viewing data 116 to the usage collection server 114. The
panelist devices 112 may further include portable devices (e.g.,
mobile phones) physically located outside of the panelist household
101. Such a panelist device 112 may be associated with the panelist
household 101 by being registered to one of the members 102, and
the panelist device 112 may monitor viewing events on the panelist
device 112 and report panelist viewing data 116 to the usage
collection server 114.
[0031] In some embodiments, the panelist devices 112 may,
additionally or alternatively, generate panelist viewing data 116
by monitoring media viewed by the member 102 while carrying the
panelist device 112. For example, the panelist device 112 may
include a microphone to capture and analyze ambient audio
information to determine a likelihood that the member 102 is
watching a particular television program. In some cases, the
panelist device 112 may extract encoded signals from the sound
information identifying the particular television program being
watched by the member 102. The panelist device 112 may also
identify the particular television program from the sound
information using other mechanisms, such as, for example, by
generating acoustic fingerprint from the sound information in
querying a storage device mapping known acoustic fingerprints to
television programs. In some embodiments, the panelist device 112
may monitor other types of information to determine a television
program being watched by the member 102, such as, for example,
video information, radio frequency (RF) signals, infrared (IR)
signals, or other information. The panelist viewing data 116
generated from the panelist devices 112 in this manner may be saved
associated with the panelist household 101 and/or member 102
associated with the panelist device 112.
[0032] The panelist devices 112 may produce panelist viewing data
116 representing viewing activity by the members 102. In some
embodiments, the panelist device 112 may provide panelist viewing
data 116 directly to the storage device 120. The panelist devices
112 may, additionally or alternatively, provide the panelist
viewing data 116 to a separate collection server or set of servers,
and the panelist viewing data 116 may be acquired by or otherwise
stored in the storage device 120. In some embodiments, the panelist
viewing data 116 may include information regarding television
viewing events, such as, for example, a television program being
watched, a television network, an entity operating the television
network, a start time and stop time for the television viewing
event, an identifier of the member 102 associated with the
television viewing event, and/or other information.
[0033] The members 102 may be associated with demographics, such as
age, gender, race, ethnicity, income, education level, and these
demographics may be collected and stored in the storage device 120
or another storage as panelist household data 110. In the example
illustrated in FIG. 1, the panelist household 101 includes four
members 102: an 18-year-old male, a 24-year-old female, a
35-year-old female, and a 46-year-old male. The specific age and/or
gender of the members 102 may be stored in panelist household data
110. The demographic attributes of the members 102 may,
additionally or alternatively, be associated with demographic
panelist classes. For example, each member 102 may be associated
with one of a panelist class for age (e.g., 18-24, 25-34, 35-44,
45-54, 55-64, or 65+), rather than a specific age. This information
may also be stored in the panelist household data 110. Other
demographic attributes of the members 102 may be collected, such as
occupation, income, race and/or ethnicity. Similarly, these
demographic attributes may be saved to the panelist household data
110 based on a plurality of panelist classes (e.g., income of
$0-$25,000, $25,001-$50,000, $51,000-$75,000, $75,001-$100,000 . .
. $300,000+).
[0034] The demographic attributes of individual members 102 of the
household 101 may be aggregated into household demographic
attributes and associated panelist classes, which are stored in the
panelist household data 110. For example, the income of the members
102 may be aggregated to determine a household income. Additional
household demographic attributes may be a language spoken in the
household 101, a number of members 102 of the household 101, and a
number of children of the household 101, and panelist classes may
be generated for each of the household demographic attributes. In
addition, a geographic area or location for the panelist household
101 may be stored in the panelist household data 110. The
geographic area or location for the panelist household 101 may be
saved according to a geographic market (e.g., one of 210 Designated
Market Areas (DMAs) assigned by Nielsen). The panelist household
data 110 may further include device demographic attributes, for
example, one or more of a television stratum, a presence of a
digital video recorder (DVR), a number of television sets of the
households 101, and types of the panelist devices 112 associated
with the household 101.
[0035] The demographic information for the household members 102
and/or panelist households 101 may be collected in a number of
ways. For example, the panelist households 101 may be recruited to
be part of a television viewing panel that is used to provide
panelist viewing data 116. Once the panelist household 101 is
recruited, the demographic information may be collected as part of
a registration process. In another example, the panelist household
101 may be a part of, or recruited into, an Internet usage panel
that is used to provide Internet usage data. Demographic
information of the household members 102 may be collected when the
panelist household 101 is registered to be part of the Internet
usage panel. As part of the Internet usage panel, the panelist
household 101 may have a panel application installed on one or more
of the panelist devices 112 in the panelist household 101. The
panel application may collect television and/or internet usage data
to send to the usage collection server 114. In some embodiments,
the internet usage data could be used to infer information about
household member 102, such as by comparing internet content
accessed by each member 102 with demographic or other information
about users accessing the same content. Other methods may be used
to capture or confirm information about members 102 of the panelist
household 101, such as survey data or data captured from other
household behaviors, or data provided by third party services that
attempt to determine demographic data of household members 102.
[0036] The storage device 120 may further receive population data
118 over a network 122. The population data 118 may be received
from one or more trusted third party sources, such as one or more
census bureaus and/or credit bureaus. The population data 118 may
include demographic data (e.g., age, gender, ethnicity, race,
and/or income) of constituents of households of a market. The
population data 118 may also include residential information for
the constituents, such as information on a male, aged 35 with an
income of $35 k, and living at 335 Main Street, Charleston, S.C.
24901. The population data 118 may be based on geographic markets
(e.g., according to Nielsen) and aggregated based on household
and/or demographic attributes. For example, the population data 118
may include aggregated data, such as there being 500 households in
the Charleston market with a male member aged 35-44 having an
income of $25,001-$50,000. In some instances, the population data
118 may also provide limited viewing data associate with
demographics, households, and/or constituents.
[0037] FIG. 2 illustrates an example of a system in which
household-level viewing data may be used to generate projected
market-level viewing data through demographic attribution. The
system 200 includes a reporting server 202 embodied, for example,
by a general-purpose computer capable of responding to and
executing instructions in a defined manner, a personal computer, a
special-purpose computer, a workstation, and/or a mobile device.
The reporting server 202 may receive instructions from, for
example, a software application, a program, a piece of code, a
device, a computer, and/or a computer system, which independently
or collectively direct operations. The instructions may be embodied
permanently or temporarily in any type of machine, component,
equipment, or other physical storage medium that is capable of
being used by the reporting server 202.
[0038] The reporting server 202 may have a processor that executes
instructions implemented by a pre-processing module 204, an RHU
generation module 206, a household assignment module 208, and a
report generation module 210. The reporting server 202 may be
operable to process the panelist household data 110, panelist
viewing data 116, and population data 118 to generate one or more
reports 212 that include panelist viewing data 116.
[0039] FIG. 3 is a flow chart illustrating an exemplary process 300
for generating the reports 212. The following describes the process
300 as being performed by components of the reporting server 202
with respect to data associated with the panelist household 101.
However, the process 300 may be performed by other systems or
system configurations and implemented with respect to other members
of the viewing audience.
[0040] At step 302, the pre-processing module 204 may access a
portion of the collected data, including the population data 118.
The pre-processing module 204 may perform one or more
pre-processing functions on the population data 118 as appropriate.
In some cases, the pre-processing module 204 may identify
particular demographic attributes of the population data 118, such
as age, gender, race, occupation, geographic area, and/or other
elements associated with the population. In some cases, the
pre-processing module 204 may sort the population data into
particular demographic attributes based on the particular member
102 associated with each viewing event in the panelist viewing data
116. In some cases, the pre-processing module 204 may examine the
distribution of the population data, and generate classes based on
the one or more of household demographic attributes, member
demographic attributes, and/or device demographic attributes. The
household demographic attributes may include one or more of an
income of the household, a language spoken in the household, a
number of members of the household, and a number of children in the
household. The member demographic attributes may include one or
more of an age of at least one member of the household, a race of
at least one member of the household, and an education level of at
least one more of the household. The device demographic attributes
may include one or more of a television stratum, a presence of a
digital video recorder (DVR), and a number of panelist devices 112.
The population data 118 may be received from the network 122 based
on the geographic market (e.g., assigned by Nielsen), or
alternatively, the pre-processing module 204 may categorize the
households into markets based on an associated location, street,
and/or address.
[0041] At step 304, the RHU generation module 206 may generate an
array of RHUs for each of the markets based on the population data.
The RHUs may be generated based on any number of demographic
attributes 402. As illustrated in FIG. 4, the demographic
attributes 402 of the RHUs may include a number of members 102 in
the household 101, a gender of the members 102, an income of the
household 101, a number of television sets 112 of the household
101, and/or a television stratum. Each RHU may then be assigned
classes 404 for each of the demographic attributes 402. The classes
404 may include a single value or a range of values for each of the
demographic attributes 402. For example, the classes 404 with
limited number of probable values (e.g., number of television sets
and/or gender) may be based on a single value, but the classes 404
with a larger number of probably values (e.g., age and/or income)
may be based on a range of values.
[0042] The RHU generation module 206 may then generate a quota for
each of the RHUs of the market. The quota may be a representative
number of households in each RHU proportionally based on the
distribution of demographic attributes in the population data 118.
For example, FIG. 5 illustrates a functional cross-section of the
RHUs having 2 members, a male between 25-34 years old, a female
between 25-34 years old, a household income of $25,001-50,000, and
2 television sets. As further illustrated by the outlined images of
houses, the number of panelist households to be assigned to the RHU
with DBS and DVR is 12, the number assigned to the RHU with
Cable/Telco and DVR is 18, and the number assigned to the RHU with
OTA and DVR is 6. The number of panelist households to be assigned
to the RHUs of DBS without DVR is 6, the number assigned to the RHU
with Cable/Telco without DVR is 9, and the number assigned to the
RHU with OTA without DVR is 3. The illustrated quotas for each of
the RHUs is based on the relative proportion of the population
falling within these classes. For example, as illustrated, the
population data 118 may indicate that there are about twice as many
households of (2 Members, M25-34 F25-34; $25,001-50,000; 2
television sets) with DBS that have a DVR than do not have a DVR.
The illustrated quotas are also based on the population data 118
indicating that there are about two-thirds as many households of (2
Members, M25-34 F25-34; $25,001-50,000; 2 television sets) with DBS
and a DVR than Cable/Telco and a DVR. The quotas would therefore
proportionally reflect the demographics of the market as indicated
in the population data 118.
[0043] At step 306, the household assignment module 208 may access
the household data 110 for panelist classes for demographic
attributes of a plurality of panelist households 101. As discussed
herein, the classes may be collected directly from the panelist
households 101 and stored in the storage device 120. The panelist
household data 110 may include stored data of the panelist
households 101, including classes for household demographic
attributes, member demographic attributes, and/or device
demographic attributes. The panelist household data 110 may also
include an activity log for the panelist households based on the
panelist viewing data 116. For example, the panelist household data
110 may indicate whether the panelist households 101 have been
inactive within the 7, 30, or 60 days.
[0044] At step 308, the household assignment module 208 may
determine that the panelist households 101 are active within a
predetermined period of time. For example, the household assignment
module 208 may modify a list of the panelist households of step 306
by deleting the panelist households without active viewing data
within the past 30 days. Removing inactive panelist households 101
may avoid distortion of the panelist viewing data due to
non-reporting and/or inactive panelist households 101.
[0045] At step 310, the household assignment module 208 may assign
each panelist household 101 to one of the RHUs based on at least
one of the panelist classes matching the classes for respective
demographic attributes of the RHU. For example, as illustrated in
FIG. 5, the household assignment module 208 may assign panelist
households 101 (shown as filled in houses) based on the quota
(shown as outlined houses). The household assignment module 208 may
assign panelist households 101 to RHUs until the quota of the RHU
is reached. As further illustrated in the exemplary flow chart of
FIG. 6, the household assignment module 208 may assign first,
second, and third panelist households 101 to one or more RHUs.
[0046] At step 320, the household assignment module 208 may match
panelist classes for first, second, and third demographic
attributes of the first panelist household 101 to respective
classes of a first RHU. At step 322, the household assignment
module 208 may assign the first panelist household to the first RHU
based on the matching of the first, second, and third demographic
attributes. For example, the panelists classes (e.g., 2 members, a
male between 25-34 years old, and 2 television sets) of the first,
second, and third demographic attributes for the first panelist
household may match the respective classes (e.g., 2 members, a male
between 25-34 years old, and 2 television sets) of the first RHU.
Thus, the first panelist household 101 may be assigned to the first
RHU.
[0047] At step 324, the household assignment module 208 may
determine the panelist classes of the second panelist household do
no match respective classes of any RHU for the first, second, and
third demographic attributes. However, at step 326, the household
assignment module 208 may match the panelist classes for the first
and second demographic attributes of the second panelist households
to respective classes of the first RHU. At step 328, the household
assignment module 208 may assign the second panelist household to
the first RHU. For example, the panelists classes (e.g., 2 members,
a male between 25-34 years old, and 10 television sets) of the
first, second, and third demographic attributes for the second
panelist household does not match the respective classes for any
RHU. However, the panelists classes (e.g., 2 members and a male
between 25-34 years old) of the first and second demographic
attributes for the second panelist household does match the
respective classes for the first RHU. Thus the second panelist
household is assigned to the first RHU.
[0048] At step 330, the household assignment module 208 may
determine that panelist classes for the first and second
demographic attributes of the third panelist household do not match
respective classes of any RHU. The household assignment module 208
may match the panelist class for the first demographic of the third
panelist household to the respective class of the first RHU. At
step 334, the household assignment module 208 may assign the third
panelist household to the first RHU. For example, the panelists
classes (e.g., 2 members, a male between 96-100 years old) of the
first and second demographic attributes for the third panelist
household does not match the respective classes for any RHU.
However, the panelists classes (e.g., 2 members) of the first
demographic attribute for the third panelist household does match
the respective classes for the first RHU. Thus, the second panelist
household is assigned to the first RHU.
[0049] For example, the first demographic attribute may be a
household attribute, such as one or more of an income of the
household, a language spoken in the household, a number of members
of the household, and a number of children in the household. The
second demographic attribute may be a member attribute, such as one
or more of an age of at least one member of the household, a gender
of at least one member of the household, a race of at least one
member of the household, an ethnicity of at least one of the
household, and an education level of at least one member of the
household. The third demographic attribute may be a device
attribute, such as one or more of a number of panelist devices of
the household, a television strata, and a presence of a digital
video recorder (DVR).
[0050] Perfect matches of panelist households may be prioritized to
provide a more accurate representation of the RHU. After
determining the number of perfect matches (e.g., matching classes
for the first, second, and third demographic attributes) is not
sufficient to meet the quota of the RHU, the household assignment
module 208 may selectively "relax" or disregard demographic
attributes to assign the panelist households 101. Thus, one or more
device demographic attributes (e.g., number of television sets in
the panelist household 101) may be removed from consideration, as
illustrated in steps 324-326. Then, if necessary, one or more
member demographic attributes (e.g., the ethnicity of one of the
members 102) may be disregarded, as illustrated in steps 330-332.
Then one or more household demographic attributes (e.g., income of
household 101) may potentially be removed from consideration.
[0051] Although FIG. 6 illustrates first, second, and third
demographic attributes, the process 300 may include any number of
demographic attributes. The process may include just the first and
second demographic attributes. The process 300 may, additionally or
alternatively, include a plurality of one or more of the first,
second, and third demographic attributes. For example, the process
300 may include two first demographic attributes, two second
demographic attributes, and two third demographic attributes, and
proceed similar to steps 320-334, iteratively removing one of the
third, second, and third demographic from consideration in order to
assign the panelist households 101 to the RHUs. Although FIG. 6
illustrates the first, second, and third panelist households 101
being assigned to the first RHU, the panelist households 101 may be
assigned in any arrangement. For example, the first panelist
household 101 may be assigned to a third RHU in step 334. The
second panelist household 101 may be assigned to a second RHU in
step 328, and the third panelist household 101 may be assigned to
the first RHU in the step 322.
[0052] The assignments of steps 320-334 may proceed until a quota
for the RHUs are met. For example, the quota for RHU 1 may be met
in step 322, when there are sufficient number of panelist
households 101 with matching panelist classes for the first,
second, and third demographic attributes. The assignment for RHU 1
would then stop due to the quota being met. However, the assignment
for RHU 2 may proceed through steps 324-328, for example, when RHU
2 is not as well represented in the panelist households 101 as RHU
1. Steps 320-334 may be performed for each of the RHUs of the
market in order to provide panelist household assignments that
proportionally matches the demographic attributes of the
market.
[0053] In some embodiments, the household assignment module 208 may
duplicate matching panelist households 101 of an RHU based on the
number of matching panelist households 101 assigned to the
respective RHU being less than the quota. In this instance, the
panelist households 101 with the best match to the respective RHU
(e.g., the most matching classes) may be duplicated to provide an
improved representation of the RHU. The duplication of the panelist
households may ensure that the quota is met for each of the RHUs,
while maintaining the demographic integrity of the RHUs.
[0054] At step 312, the report generation module 210 may access
panelist viewing data 116 representing viewing events associated
with the panelist households 101. At step 314, the report
generation module 210 may generate viewership reports 212 with the
RHUs and the panelist viewing data 116 of the assigned panelist
households 101. The reports 212 may include data at any level of
aggregation, and may be specified by a demographic attributes of
the RHUs. The reports 212 may include the panelist viewing data 116
of various demographic groups as estimated through the use of
demographic attribution. For example, a household-based report 212
may indicate that 10% of households that primarily speak Spanish
watch soccer between 7 and 8 pm on Wednesday or 25% of households
with at least one child watch Peppa Pig. A member-based report 212
may indicate that 8% households with at least one member having a
Ph.D. watched PBS. A device-based report 212 may indicate that 20%
of households without a DVR watch NBC during prime-time. The
reports 212 based on the RHUs may include as many demographic
attributes as desired. The reports 212 may provide accurate viewing
data obtained directly from panel devices 112 in panelist
households 101, accurately scaled based on the demographics of the
market population.
[0055] The reports 212 may be displayed on a graphical user
interface (GUI) on any type of device. The reports 212 may be
generated from the data and displayed to provide an accurate
measure of demographic-based viewing data for the overall market to
the content providers, advertisers, and others. The reports 212 may
therefore be used to estimate the number of viewing people and/or
households of a particular demographic for a particular program,
advertisement, sporting event, and/or other content item.
[0056] In some embodiments, the reports 212 of a market may be
scaled relative to one or more other markets. For example, a first
report 212 may be generated based on the New York City market and a
second report 212 may be based on the Washington, D.C. market. The
first and second reports 212 may be combined by scaling the reports
212 based on the relative overall population of the market and
integrating. Thus, the first report 212 may be multiplied by a
factor of the population of the New York City market relative to
the Washington, D.C. market, and added to the second report 212 to
combine the two markets.
[0057] At step 316, the report generation module 210 may adapt the
reports 212 based on empirical viewing data. The reports 212 may be
compared to the projected assigned viewing data to the individual
platform targets, such as known viewing data for the market made
available as either census totals or a combination of census and
enumeration. The targets may provide a market data set with a
high-confidence level for accuracy. As a result of the comparison,
the report generation module may assign individual events into the
appropriate RHUs to reach the targets of the known empirical data.
The assigned individual events may be actual viewing events from
the panelist viewing data 116, which were previously unassigned to
an RHU. The adaption of step 316 may assure that the projected
results of the reports 212 match individual-currency reported
results of known empirical data.
[0058] Although specific examples using various equations of
probability are described herein, the methods described herein can
be used with a variety of probability and statistical techniques
and are not limited to only the equations and examples shown.
[0059] The techniques described herein can be implemented in
digital electronic circuitry, or in computer hardware, firmware,
software, or in combinations of them. The techniques can be
implemented as a computer program product, such as a computer
program tangibly embodied in an information carrier, e.g., in a
machine-readable storage device, in machine-readable storage
medium, in a computer-readable storage device or, in
computer-readable storage medium for execution by, or to control
the operation of, data processing apparatus, e.g., a programmable
processor, a computer, or multiple computers. A computer program
can be written in any form of programming language, including
compiled or interpreted languages, and it can be deployed in any
form, including as a stand-alone program or as a module, component,
subroutine, or other unit suitable for use in a computing
environment. A computer program can be deployed to be executed on
one computer or on multiple computers at one site or distributed
across multiple sites and interconnected by a communication
network.
[0060] Process steps of the techniques can be performed by one or
more programmable processors executing a computer program to
perform functions of the techniques by operating on input data and
generating output. Process steps can also be performed by, and
apparatus of the techniques can be implemented as, special purpose
logic circuitry, e.g., an FPGA (field programmable gate array) or
an ASIC (application-specific integrated circuit).
[0061] Processors suitable for the execution of a computer program
include, by way of example, both general and special purpose
microprocessors, and any one or more processors of any kind of
digital computer. Generally, a processor will receive instructions
and data from a read-only memory or a random access memory or both.
The essential elements of a computer are a processor for executing
instructions and one or more memory devices for storing
instructions and data. Generally, a computer will also include, or
be operatively coupled to receive data from or transfer data to, or
both, one or more mass storage devices for storing data, such as,
magnetic, magneto-optical disks, or optical disks. Information
carriers suitable for embodying computer program instructions and
data include all forms of non-volatile memory, including by way of
example semiconductor memory devices, such as, EPROM, EEPROM, and
flash memory devices; magnetic disks, such as, internal hard disks
or removable disks; magneto-optical disks; and CD-ROM and DVD-ROM
disks. The processor and the memory can be supplemented by, or
incorporated in special purpose logic circuitry.
[0062] A number of embodiments of the techniques have been
described. Nevertheless, it will be understood that various
modifications may be made. For example, useful results still could
be achieved if steps of the disclosed techniques were performed in
a different order and/or if components in the disclosed systems
were combined in a different manner and/or replaced or supplemented
by other components.
* * * * *