Abstract: Methods, apparatus, systems, and articles of manufacture to create a reference signature asset from matched signature fragments are disclosed. Example apparatus disclosed herein form a group of matched signature fragments corresponding to a same piece of media. Disclosed example apparatus also select a subset of matched signature fragments from the group to form a path of matched signature fragments that collectively represent a continuous segment of the piece of media, the continuous segment identified by a segment start time and a segment end time, the segment start time corresponding to a fragment start time associated with an initial matched signature fragment in the path, and the segment end time corresponding to a fragment end time associated with a last matched signature fragment in the path. Disclosed example apparatus further use the path to combine the subset of matched signature fragments and create a new reference signature asset.

Abstract: A disclosed example includes determining a plurality of cross-demographic total census parameter values based on ratios of third-party impression counts and third-party audience sizes, selecting a first cross-demographic total census parameter value as an upper bound constraint from the plurality of cross-demographic total census parameter values based on comparisons between ones of the plurality of cross-demographic total census parameter values, selecting a second cross-demographic total census parameter value from the plurality of cross-demographic total census parameter values based on the second cross-demographic total census parameter value being less than the upper bound constraint, determining a first census-level audience size based on the second cross-demographic total census parameter value, and determining a first census-level impression count based on the second cross-demographic total census parameter value.

Abstract: A computer-implemented method for efficiently estimating the number of unique elements in a collection of elements comprises generating, via hash logic, hash values associated with the elements. The hash values specify bit positions within an array of bits. Hash values output from the hash logic conform to a geometric distribution such that bit positions of the array of bits corresponding to lower orders bits are more likely to be generated than bit positions corresponding to higher-order bits. Bits of the array of bits corresponding to the bit positions are set. The number of bits of the array of bits that are set is counted. Estimation logic estimates the number of unique elements of the collection of elements as a function of the number of bits of the array of bits that are set.

Abstract: Disclosed examples access first impression data representative of first impressions collected by an impression monitor system, the first impressions corresponding to the media accessed at a plurality of client devices; generate a panelist impressions composition by removing at least one duplicate impression from second impression data, the second impression data representative of second impressions logged by meters installed at the client devices, the duplicate impression corresponding to one or more accesses to the media represented in both the first impression data and the second impression data for a same audience member; determine an error value for the media based on the panelist impressions composition and a database proprietor impression composition, the database proprietor impression composition provided by a database proprietor for the media; and determine an error-corrected impression composition based on the error value and the panelist impressions composition.

Abstract: A system for prospectively identifying media characteristics for inclusion in media content is disclosed. A neural network database including media characteristic information and feature information may associate relationships among the media characteristic information and feature information. Personal characteristic information associated with target media consumers may be used to select a subset of the neural network database. A first set of nodes, representing selected feature information, may be activated. The node interactions may be calculated to detect the activation of a second set of nodes, the second set of nodes representing media characteristic information. Generally, a node is activated when an activation value of the node exceeds a threshold value. Media characteristic information may be identified for inclusion in media content based on the second set of nodes.

Abstract: An example apparatus includes a distance matrix generator to determine a number of remapping combinations corresponding to a first and a second household demographic model, generate (a) a first age-based household demographic model and a second age-based household demographic model and (b) a third age-based household demographic model and a fourth age-based household demographic model, generate a first distance matrix corresponding to the first and the third age-based household demographic model, and generate a second distance matrix corresponding to the second and the fourth age-based household demographic model, a distance matrix identifier to determine a first and a second reassignment distance and a person identification assigner to assign a person identification number of a first person corresponding to the first household to a second person corresponding to the second household based on the first distance matrix or the second distance matrix.

Abstract: An example apparatus includes an audience size calculator circuitry to determine a first census-level audience size, and an impression count calculator circuitry to determine a first census-level impression count.

Abstract: Methods and apparatus to generate reference signatures are disclosed. An example method includes collecting a first signature for media being presented to a plurality of households; crediting the media when the first signature matches a reference signature in a reference signature database; in response to determining that the first signature does not match a reference signature in the reference signature database and the first signature does not match an unidentified signature in an unknown signature database, storing the first signature in the unknown signature database; in response to determining that a second signature does not match the reference signature in the reference signature database and the second signature matches the unidentified signature in the unknown signature database, and increasing a count associated with the unidentified signature.

Abstract: Methods, apparatus, systems, and articles of manufacture are disclosed to extract information from uniform resource locators (URLs). An example system includes at least one memory, programmable circuitry, and instructions to cause the programmable circuitry to collect first uniform resource locator (URL) information from a server of an Internet-based media publisher, the first URL information corresponding to first media accessed by first users, determine first URL components in the first URL information, and determine feature-to-user assignment rules based on the first URL components.

Abstract: Methods and apparatus to capture images are disclosed. An example apparatus disclosed herein includes an image sensor to observe a first image of a face, and a logic circuit to determine whether the face in the first image is in a first orientation with respect to the image sensor, and when the face in the first image is in the first orientation: trigger the image sensor to capture a second image of the face, and perform a facial recognition process on the second image, but not on the first image.

Abstract: Methods, apparatus, systems, and articles of manufacture are disclosed to determine unique audience size via clustered data. An example apparatus includes memory, machine readable instructions, and processor circuitry to at least one of execute or instantiate the machine readable instructions. The example processor circuitry is to form a first matrix identifying matching and non-matching instances of first values of respective demographics of an audience of media and second values of the respective demographics. Also, the processor circuitry is to normalize the first matrix to generate an account sharing adjustment matrix. Additionally, the processor circuitry is to apply the account sharing adjustment matrix to a vector of one or more counts of impressions of the media to correct for one or more members of the audience that share a user account registered with a database proprietor, the one or more counts of the impressions corresponding to the respective demographics.

Abstract: Methods, apparatus, systems, and articles of manufacture are disclosed to determine whether audience measurement meters are co-located. An example apparatus is to, based on a difference between a first sampling time of a first entry of a first log and a second sampling time of a corresponding entry of a second log satisfying a first threshold, determine at least one matching instance of at least one first device identifier of the first entry and at least one second device identifier of the corresponding entry. Additionally, the example apparatus is to populate a variable with the at least one matching instance. The example apparatus is also to, based on a metric satisfying a second threshold, cause transmission of an alert indicating that a first meter and a second meter were co-located during generation of the first log and the second log, the metric based on the at least one matching instance.

Abstract: A computer-implemented method for efficiently estimating the number of unique elements in a collection of elements comprises generating, via hash logic, hash values for each element of the collection of elements. The method further comprises specifying, in a sketch-frequency table, a set of discrete statistical values associated with the hash values and, for each discrete statistical value of the set of discrete statistical values, information indicative of a frequency at which binary representations of the hash values are associated with the discrete statistical value. The cardinality of the collection of elements is estimated based on the sketch-frequency table.

Abstract: An apparatus and method to automatically determine the location of unknown media devices is disclosed. An example apparatus includes a media device detector to detect an unknown media device identified in monitoring data collected by an audience measurement device at a location determined to have an “on” status, a media transmission detector to detect a media transmission associated with a device address and an association storer to store an association of the device address, the unknown media device and the location. A probability determiner determines a probability that the unknown media device is located at the location.

Abstract: Methods, apparatus, systems and articles of manufacture are disclosed to optimize reference signature matching using watermark matching. An example disclosed herein includes selecting first unhashed reference signatures to evaluate for a second time segment of monitoring data based on watermark coverage of a first time segment of the monitoring data, the monitoring data obtained from a meter monitoring media presented by a media device and comparing the first unhashed reference signatures with first unhashed monitored signatures from the second time segment of the monitoring data.

Abstract: Disclosed example apparatus to determine volume of switching (VoS) among television programs examine first viewing data associated with a first time period and second viewing data associated with a second time period to identify a first set of panelists represented in both the first and second viewing data; in response to a size of the first set of panelists satisfying both first and second thresholds, estimate the VoS based on a first subset of the first viewing data and a second subset of the second viewing data associated with the first set of panelists; and in response to the size of the first set of panelists satisfying the first but not the second threshold, estimate the VoS based on the first and second subsets, and a third subset of the first viewing data and a fourth subset of the second viewing data associated with a second set of panelists.

Abstract: Example apparatus disclosed herein are to determine a location of a portable device based on location data collected by the portable device, determine whether a person is within a threshold distance of the portable device based on proximity data collected by the portable device, and, when the person is determined to be within the threshold distance of the portable device: (i) infer at least one of a presence or an absence of the person in a media exposure environment based on the location of the portable device, (ii) determine a confidence level that the person corresponds to a member of the panelist household, and (iii) determine an accuracy of presence information collected by a people meter based on the confidence level and the inference of the at least one of the presence or the absence of the person, the people meter different than the portable device.

Abstract: Methods, apparatus, systems and articles of manufacture to identify media sources are disclosed. Example apparatus disclosed herein include an exclusivity determiner, a unique asset identifier, and a signature matcher. The exclusivity determiner is to determine whether media is exclusive to a provider. The unique asset identifier is to associate a first signature of the media with the provider if the media is exclusive to the provider. The signature matcher is to identify the provider based on a second signature, the second signature matching the first signature, the second signature extracted from media presented at a media presentation location.

Abstract: Methods and apparatus to generate audience measurement data from population sample data having incomplete demographic classifications are disclosed.

Abstract: Apparatus, systems, articles of manufacture, and methods to identify media using watermarks and signatures are disclosed. An example apparatus includes at least one memory, instructions, and processor circuitry to execute the instructions to: determine a first time and a category indicated by a watermark detected in the media, determine a search window based on the first time indicated by the watermark, the search window to at least one of (i) extend a first duration of time before the first time indicated by the watermark or (ii) extend a second duration of time after the first time indicated by the watermark, and compare (i) a query signature associated with a second time within the search window with (ii) a subset of reference signatures associated with the category to identify the media, the query signature generated from the media, the subset of reference signatures generated from reference media associated with the category.