The Nielsen Comapny (us), LLC

Josh Gaunt

Albert Ronald Perez

Kumar Nagaraja Rao

Paul Donato

Mainak Mazumdar

Seema Varma Srivastava

James R Oliver

Juliette Aurisset

Yutao Peng

A disclosed example includes accessing computer-generated impression records, the computer-generated impression records based on network communications received at a server of a database proprietor from client devices, the computer-generated impression records indicative of accesses to media at the client devices; accessing self-reported demographic data and behavioral data from a database of the database proprietor, the self-reported demographic data and the behavioral data corresponding to user accounts registered with the database proprietor and associated with the client devices; comparing the self-reported demographic data with a probability distribution of higher-accuracy demographic data; determining different adjustments for corresponding ones of self-reported demographics of the self-reported demographic data based on the comparison; adjusting the corresponding ones of the self-reported demographics based on corresponding ones of the different adjustments to generate adjusted user demographic data; and assigning the adjusted user demographic data to corresponding ones of the computer-generated impression records.

Methods and apparatus to analyze and adjust demographic information

An example includes generating panelist-user data based on reference demographic information and self-reported demographic information, the reference demographic information and the self-reported demographic information corresponding to audience members of an audience member entity panel that are also registered users of a database proprietor, the reference demographic information from a panelist database of an audience measurement entity, and the self-reported demographic information from a user database of the database proprietor; generating a first training model and a second training model, the first training model based on a first portion of the panelist-user data, the second training model based on a second portion of the panelist-user data; selecting the first training model based on outputs of the first and second training models; and generating a third model by making an adjustment to a demographic category of the first training model, the third model to adjust third demographic information.

An example includes generating a first model based on reference demographic data corresponding to panelists, and based on second demographic data and behavioral data from a database proprietor that is separate from a first entity, the reference demographic data collected from the panelists by the first entity, and the second demographic data and the behavioral data corresponding to ones of the panelists having user accounts with the database proprietor; analyzing an output of the first model including the second demographic data partitioned into a plurality of nodes, each node representing a respective subset of the second demographic data; redistributing some of the second demographic data between some of the nodes to generate a second model; applying the second model to third demographic data of the database proprietor; and determining whether to adjust at least some of the third demographic data based on an output of the second model.

An example method to adjust demographic data involves generating a first model based on reference demographic data corresponding to panelists, and based on first age data and a quantity of mobile webpage views from a database proprietor that is separate from a first entity, the reference demographic data collected from the panelists and stored by the first entity, and the first age data and the quantity of mobile webpage views corresponding to ones of the panelists having user accounts with the database proprietor; and based on determining that redistributing at least some of the first age data will increase an age data accuracy, redistributing the at least some of the first age data across age categories to generate a second model, the second model to facilitate adjusting second age data of the database proprietor to correct inaccuracies and/or misrepresentations in at least some of the second age data.

An example method to adjust demographic data involves generating a first model based on reference demographic data corresponding to panelists and based on first self-reported demographic data and behavioral data from a database proprietor. The first self-reported demographic data and the behavioral data correspond to ones of the panelists having user accounts with the database proprietor. The first self-reported demographic data is provided to the database proprietor by the ones of the panelists having the user accounts with the database proprietor. The first model is used to partition the first self-reported demographic data into nodes, each node representing a respective subset of the first self-reported demographic data. At least some of the first self-reported demographic data is redistributed between at least some of the nodes to generate a second model that is to facilitate correcting at least one of inaccuracies or misrepresentations in the second self-reported demographic data.

Examples to determine media impressions are disclosed. An example method includes determining an exposure of first media, wherein the exposure occurs after a cookie identifier is received at a computing device, wherein the cookie identifier is received from a database proprietor, determining a first panelist identifier associated with the exposure based on the cookie identifier, determining a second panelist identifier associated with the exposure based on a determination of a user identity by a panelist meter associated with the computing device, determining an adjustment factor by comparing the first panelist identifier and the second panelist identifier, determining a number of impressions of second media from data logged by the database proprietor, the data logged in response to requests transmitted to the database proprietor in response to clients accessing the second media, and applying the adjustment factor to the number of impressions to determine an adjusted number of impressions.

Methods and apparatus to determine an adjustment factor for media impressions

Examples to determine media impressions are disclosed. An example method includes detecting a cookie identifier established by a database proprietor at a computing device, determining an impression of media, wherein the impression occurs after the cookie identifier is established, determining a first panelist identifier associated with the impression based on the cookie identifier, determining a second panelist identifier associated with the impression based on determination of a user identity by a panelist meter associated with the computing device, and storing an adjustment factor determined by comparing the first panelist identifier and the second panelist identifier.

Methods, apparatus, and articles of manufacture to analyze and adjust demographic information of audience members are disclosed. An example method involves generating a first model based on reference demographic data corresponding to panelists and based on second demographic data and behavioral data from a database proprietor. The second demographic data and the behavioral data correspond to ones of the panelists having user accounts with the database proprietor. The method also involves using the first model to partition the second demographic data into a plurality of nodes, each node representing a respective subset of the second demographic data. The method also involves redistributing at least some of the second demographic data between at least some of the nodes to generate a second model.

Growing community of inventors

Sunnyvale, CA, United States of America

Josh Gaunt