Google Inc.

Qwest Communications International Inc.

James Robert Koehler

Sheng Ma

Evgeny Skvortsov

Jiayu Peng

Xichen Huang

Benjamin Kreuter

Craig Wright

Laura Grace Book

Arthur Asuncion

Yueqing Wang

Yuxue Jin

Yunting Sun

David Chan

Ruixue Fan

Matthew Strecker Burriesci

Lu Zhang

Joseph Kelly

Aiyou Chen

Nicolas Remy

William Martin Halpin, Jr

Harrison Mark Gordon

Sean Patrick Miller

Thomas A Vaughan

Antonella Pavese

Zachary Charles Frazier

Andrew Smith

Michael Perry

Xiaojing Wang

Kirsten Paige Yee

Xiaojing Huang

Yijia Feng

Conor Sontag

Elizabeth Van Alstine

Shi Zhong

Luis Gonzalez Perez

David L Bosley

Sheng Na Ma

Benjamin R Kreuter

Systems and methods for generating min-increment counting bloom filters to determine count and frequency of device identifiers and attributes in a networking environment are disclosed. The system can maintain a set of data records including device identifiers and attributes associated with device in a network. The system can generate a vector comprising coordinates corresponding to counter registers. The system can identify hash functions to update a counting bloom filter. The system can hash the data records to extract index values pointing to a set of counter registers. The system can increment the positions in the min-increment counting bloom filter corresponding to the minimum values of the counter registers. The system can obtain an aggregated public key comprising a public key. The system can encrypt the counter registers using the aggregated shared key to generate an encrypted vector. The system can transmit the encrypted vector to a networked worker computing device.

Secure multi-party reach and frequency estimation

Systems and methods are disclosed herein for improved per-frequency counting systems that record interactions between individuals and a group of providers while maintaining differential privacy. A protocol may be defined that specifies frequency bins to categorize identifiers corresponding to individuals. A provider may generate a plurality of private sketches, each corresponding to a plurality of frequencies defined in the protocol. Frequency data is determined for each identifier. Identifiers are encoded into the private sketches corresponding to the identifiers' associated frequency. The plurality of private sketches from each provider in the group of providers are combined to generate a deduplicated distribution across the group. In one implementation, the private sketches of each provider are sequentially merged until all sketches have been combined, from which the total distribution can be estimated.

Differentially private frequency deduplication

The present disclosure provides systems and methods for entropy balanced population measurement. Entropy balancing is a statistical technique for preprocessing data to achieve covariate balance. Weighting coefficients may be dynamically adjusted to satisfy balance conditions or constraints to adjust for inequalities in representation, thereby improving covariate moments. Using entropy balancing and linear regression analysis with panel content exposure and results data provides a mechanism to estimate the effects of multiple exposure vectors simultaneously, including instances where panelists are exposed to a vector multiple times. Data may be obfuscated or anonymized for preprocessing via a double-encrypted intersection-based extraction mechanism, allowing both measurement systems and panel providers to retain confidential information.

Systems and methods for entropy balanced population measurement

Systems, methods, and computer-readable storage media that may be used to generate a category Bayesian hierarchical model. One method includes receiving a brand data set for each of a plurality of brands within a category, each brand data set comprising content input for a particular brand of the plurality of brands for a plurality of media channels at a plurality of times and a response for the particular brand of the plurality of brands at the plurality of times. The method includes determining a plurality of informative priors by generating a category Bayesian hierarchical model based on the plurality of brand data sets and a plurality of weak priors. The method further includes generating a brand Bayesian hierarchical model that models response for the particular brand for each of the plurality of media channels based on the brand data set for the particular brand and the plurality of informative priors.

Systems and methods for generating a brand Bayesian hierarchical model with a category Bayesian hierarchical model

Systems, methods, and computer-readable storage media that may be used to generate a Bayesian hierarchical model. One method includes generating a plurality of geographic regions by grouping one or more geographic sub-regions into each of the plurality of geographic regions. The method further includes receiving data for the geographic sub-regions, the data including responses, content inputs, content types, and location identifiers. The method further includes generating geo-level data from the received data by grouping the responses and content inputs of the received data based on a correlation of the location identifiers of the received data to the plurality of geographic regions. The method includes fitting a Bayesian hierarchical model based on at least the geo-level data, the content types, and the geographic regions and determining a content input mix for the content types for each geographic region based on the Bayesian hierarchical model and a content input constraint.

Systems and methods for generating a geo-level hierarchical Bayesian model

Methods, systems, and apparatus include computer programs encoded on a computer-readable storage medium, including a method for discovering unique entities over multiple devices. A virtual pool of entities is created and divided into subpools, each including fewer than all entities. Subpools are subdivided into delta pools. Cookies are recorded for each delta pool when the particular portion of content is presented to or accessed by entities in the delta pool. Recorded cookies are divided into cookie types based on cookie characteristics. Machine learning and statistical analysis algorithms are used to automatically determine sizes of delta pools and probabilities of each cookie type being classified as belonging to particular delta pools. Virtual entities are assigned from the virtual pool to each of the recorded cookies that were recorded when the particular portion of content was presented. A number of unique entities that accessed the particular portion of content is determined.

Discovery of unique entities across multiple devices

Systems, methods, and computer-readable storage media that may be used to generate causal models and calculate a selection bias in mixed media. In some embodiments, the selection bias calculation is in search sponsored content in the context of mixed media modeling. In some embodiments, a method for search bias correction is based on the back-door criterion from causal inference.

Selection bias correction for paid search in media mix modeling

Methods and systems of reducing computer resource utilization to update rankings based on dynamically generated content are described. The method includes identifying a ranking of content experience types of a list of content experience types. Each content experience type including parameters to generate content experience documents. The method also includes updating the ranking of content experience types. The updating includes iteratively selecting content experience types from the ranking of content experience types to generate an update subset of the ranking of content experience types, generating, for a client device, a feedback request document including a plurality of content experience documents, and receiving feedback information related to the plurality of content experience documents included in the feedback request document to update the ranking of the content experience types included in the update subset.

Dynamic resampling for ranking viewer experiences

Methods, systems, and apparatus, including computer program products, for determining a mixture ratio for allocating portions of an advertising budget among different advertising channels (e.g., print, online, radio, television) to optimize a performance measure, such as cost-per-action. A mixture space is used to define the available advertising channels and any constraints placed on those channels, such as no more than fifty percent of the advertising budget being allocated to a particular channel, and test mixture ratios are selected according to an optimality criterion. The selected test mixture ratios are used during a testing period on live traffic. The performance measures from the test mixture ratios are used to select a preferred mixture ratio from the mixture space.

Determining advertisement channel mixture ratios

A method and system for generating an accurate model of noise created through near end crosstalk (NEXT) when qualifying DSL loops. The model is configured such that it may account for loop characteristics such as fill and gauge of a particular loop, and disturbers which may create a noise environment within the particular binder group. The noise model may further be employed in an automated system which, based on stored characteristics for a particular loop, is employed to generate a noise value which is then further employable in a qualification analysis.

Method and system for modeling near end crosstalk in a binder group

A system and method provides for the identification of termination telephone numbers in a telephone network which are being employed as a data network modem. When a local exchange carrier (LEC) routes a call to another telecommunications network, such as one used by a competitive local exchange carriers (CLECs), it may be advantageous for reciprocal compensation purposes to identify which termination telephone numbers serviced by the CLECs are acting as modems. As such, all telephone traffic to a particular CLEC is monitored over a predetermined period of time. The information gathered is then processed and segregated according to terminating telephone number. At that point, comparisons may be made against reference information to determine whether the terminating telephone number is a modem or a non-modem.

System and method for identifying modem connections in a telephone network

Growing community of inventors

Mountain View, CA, United States of America

James Robert Koehler