Abstract: Encouraging broader engagement with a target publisher's content by balancing on-site topic engagement. Responsive to receiving notification of a recipient's request of an item of content from the target publisher, the on-site and off-site content consumption history of the recipient is analyzed. For each of a plurality of topics, historic engagement with on-site and off-site content is measured. Deficient topics having better off-site engagement than on-site engagement are detected. Content comprising links to items of content available from the target publisher which are characterized by deficient topics is selected and sent to the recipient. In an embodiment, the supplemental content comprises a plurality of links to other web pages on the target publisher's website which are sent to the recipient browser before the requested item of content finishes loading.

Abstract: Method and system for assessing the suitability of an entity using a proxy. A description of a behavior associated with a desirable audience is received. A proxy behavior estimated to be characteristic of the desirable audience is selected. The proxy behavior comprises the performance of proxy events related to the consumption of media received by an entity over a network, which can be found in an entity's consumption history. An entity can be assessed for inclusion in a proxy audience, by examining the entity's consumption history for proxy behaviors. A behavioral model is built using a training set comprising the proxy audience. By applying the behavioral model to the consumption history of a specified entity, the specified entity's suitability for selection can be determined. Advantageously, in an embodiment, the invention enables the use of behavioral modeling techniques even when the complete behavior of the desirable audience is not available.

Abstract: Protected audience selection system. Media consumption histories of browsers which have converted are received at a modeling system where targeting of browsers is prohibity. A model is built by determining a frequency of each respective media consumption event among the histories and comparing each determined frequency of a respective media consumption event to a frequency of the respective media consumption event among a population of browsers without the conversion event. The model is sent to a targeting system which excludes conversion events. A description of the conversion event is received at the targeting system. A history of a targetable browser is received at the targeting system. The model is applied to the history of the targetable browser at the targeting system, where conversion events have been excluded from the history. Advertising content is sent to the targetable browser according to a result of applying the model.

Abstract: Privacy centric feature analysis. A secure set of multiple mapped features is selected and provided to a mobile device. Each mapped feature maps a sharable feature to a matching criterion for an item of protected information and no combination of mapped features for a secure set are unique to an individual item of protected information. Privacy compliance instructions enable the mobile device to select a mapped feature from a received set of mapped features by identifying an item of protected information available to the mobile device which corresponds to a matching criterion found in the received set of mapped features. The sharable feature of the selected mapped feature is provided. Advantageously, the analysis system protects the privacy of the mobile device user because it does not require the mobile device to relay protected information for the selection of customized content or relevant advertisements.

Abstract: A control system determines whether a browser is retrieving supplemental content over a slow or faulty network using a tracking indicator. Based on determining that browser is retrieving supplemental content over a slow or faulty network, the control system modifies responses to opportunities to provide supplemental content to the browser. The control system may modify responses by declining to bid, responding with a lower bid amount, or responding with alternate supplemental content. Based on determining that browser is no longer retrieving supplemental content over a slow or faulty network, the control system modifies responses to opportunities to provide supplemental content to the browser by responding with a higher bid amount or responding with the original supplemental content. Thus, the control system maximizes the use of control system and campaign manager resources by responding to opportunities according to the likelihood that they will result in an impression or a conversion.

Abstract: A tag manager system provides access to a domain in a container in the tag manager system. A tag user navigates to elements in the domain (such as site pages, text elements, graphic elements, or video elements) and selects elements for tagging from within the container. The user selects tag options from within the tag manager system. The tag manager system uses the selected tag options to generate the corresponding tag codes for insertion in the domain HTML code. The tag manager saves previously defined tag codes and the tag user updates stored tag options in the tag manager system. The tag manager system provides pointer codes to insert in the HTML code so the user can update stored tag codes without updating the HTML code. The tag manager further highlights elements of the domain in the container for suggested tagging, and automatically selects option based on the element type.

Abstract: Embodiments of the invention generate metrics quantifying the mobility of a mobile device without persisting information related to the device's specific location at any given time. Specifically, at multiple intervals, a value of a mobility metric is computed based on the distance between the current location of the mobile device and a previously identified origin location of the mobile device.

Abstract: Adaptive control of exposure. A proportional exposure cap is a maximum fraction applicable to a recipient's total viewable attention in a time window. The total viewable attention represents all viewable advertising content which will be provided to the recipient. A notification of availability of an opportunity to expose a specified recipient to advertising content is received during the time window. The specified recipient's consumed viewable attention is detected. The specified recipient's total viewable attention for the time window is predicted. Responsive to the maximum fraction of the specified recipient's predicted total viewable attention for the time window being greater than the consumed viewable attention of the specified recipient, the advertising campaign's advertising content is sent to the specified recipient and the consumed viewable attention is updated.

Abstract: A data operations system receives compressed data and a search term. The data operations system completes a modified decoding of the compressed data, resulting in distinguishable data terms that are smaller than the corresponding data terms, and loads modified decoded terms into a data register. The data operations system generates a truncated search term and loads instances of the truncated search term into a query register. The data operations system performs a parallel data operation, such as query operation, by comparing each of the modified decoded terms to an instance of the truncated search term. The data operations system returns the results of the operation.

Abstract: A mobile device's location and an identifier corresponding to the mobile device is received. The mobile device's location is privatized by mapping it to landmarks proximate to the mobile device's location, storing the proximate landmarks in association with the mobile device's identifier, and discarding the received location data. The proximate landmarks are featurized to generate a model which is used to determine a value of the advertising opportunity corresponding to a target identifier.

Abstract: Adaptive Sampling. Data comprising pairings of data value with lists of data keys are received. The range of possible values of the data keys is partitioned into unbalanced buckets, with at least two of the unbalanced buckets representing different fractions of the range. Each unbalanced bucket is assigned to a respective processing unit selected from a plurality of processing units. The pairings are processed by the processing units, with each processing unit generating an intermediate result. The intermediate results are combined to generate a comprehensive result. A sampling error is determined by scaling an unbalanced bucket's intermediate result according to its corresponding fraction and comparing the scaled intermediate result to the comprehensive result. An unbalanced bucket having a sampling error less than a sampling error threshold is selected. The selected unbalanced bucket's corresponding fraction is selected as a sampling rate for a second data processing job.

Abstract: Selective regulation of information transmission from mobile applications to an external system. A privacy policy is configured for and mapped to each of a multiplicity of mobile application concerns, with each privacy policy comprising rules regulating the transmission of information to an external system. Instrumentation instructions can be integrated with a mobile application or mobile operating system. The instrumentation instructions direct the mobile device to submit a privacy policy request comprising an identifier from the mobile device to a third-party privacy compliance system and enable sending information from the mobile device to the external system, subject to the privacy policy. The privacy policy request is received at the third-party privacy compliance system which selects the privacy policy based on an identifier and sends the privacy policy to the mobile device for implementation.

Abstract: Online advertising campaigns are operated responsive to keyword trends. The keywords used by a group of browsers is analyzed periodically over time. A list of the most frequently used keywords is separated into those that have previously appeared on the list, referred to herein as the stable keywords, and those that are newly emerging, referred to herein as the trending keywords. The advertiser selects at least one advertising creative that the advertiser associates with the stable keywords, referred to herein as the stable creative, and at least one advertising creative that the advertiser associates with the trending keywords, referred to herein as the trendy creative. The advertising system then operates the online advertising campaign to deliver the respective stable and trendy creatives in proportion to the frequency of use of the trending versus the stable keywords.

Abstract: A distributed system is adapted to manage the performance of distributed processes. In one aspect, multiple stripes associated with a data item are stored in a distributed storage. The stored stripes may include one or more stripes of redundancy information for the data item. The stored stripes may include one or more copies of stripes for the data item. A distributed process including at least one task is performed. During performance of the distributed process, a determination is made as to whether to perform an accelerated data retrieval operation. Responsive to a determination to perform an accelerated data retrieval operation, at more than the minimal number stripes information required to reconstruct the data item is requested from the distributed storage. After a sufficient subset of stripes associated with the data item is received, the data item is reconstructed using the subset.

Abstract: A dynamically regulated advertising delivery control system. A campaign is operated by sending bids to an exchange responsive to receiving bid requests from the exchange, each bid request representing an opportunity to expose a browser to content. Won bid notifications are received from the exchange and exposure notifications are received from exposed browsers. Failed exposures are detected by detecting won bid notification identifiers without corresponding exposure notification identifiers. Responsive to the failed exposures exceeding an upper limit, the campaign is operated in a throttled mode by sending bids to the exchange in response to a fraction of the suitable bid requests received from the exchange and ignoring some suitable bid requests. Responsive to detecting successful exposures in the throttled mode, the operation of the campaign is dynamically regulated by increasing the fraction.

Abstract: A tag manager system provides access to a domain in a container in the tag manager system. A tag user navigates to elements in the domain (such as site pages, text elements, graphic elements, or video elements) and selects elements for tagging from within the container. The user selects tag options from within the tag manager system. The tag manager system uses the selected tag options to generate the corresponding tag codes for insertion in the domain HTML code. The tag manager saves previously defined tag codes and the tag user updates stored tag options in the tag manager system. The tag manager system provides pointer codes to insert in the HTML code so the user can update stored tag codes without updating the HTML code. The tag manager further highlights elements of the domain in the container for suggested tagging, and automatically selects option based on the element type.

Abstract: An influence system for predicting advertisement impact for audience selection. An advertising probe campaign is operated by sending an advertisement to each entity in a treatment group of entities. A control group of entities which excludes the treatment group entities is selected and no campaign advertising content is sent to the treatment group entities. An influence model is created by comparing features of the treatment group converters to features of the control group converters. An individual frequency cap is selected for each entity that is a candidate for the advertising campaign based on a result of applying the influence model to the features of the candidate entity. The entity may be selected to receive an advertisement based on the individual frequency cap. Some embodiments are integrated with a real time bidding (RTB) exchange and a bid response may be configured based on the results of applying the influence model.

Abstract: Automatic performance triggered campaign adjustment. A hierarchical feature tree is generated. Each child node's feature is more specific than its respective parent node's feature. The discovery system creates a behavioral model comprising features of the feature tree which is used in the operation of an advertising campaign. A degraded model feature is detected at the discovery system by comparing a performance metric of a model feature from two different time windows. The discovery system matches a node of the feature tree with the degraded feature and selects a prospective model feature from an ancestor node of the matching feature's node. An estimated performance metric for the prospective model feature is determined and the results are used to decide if the prospective model feature should be incorporated into an updated model or not. The model can be updated with a new model feature selected from one or more prospective model features.

Abstract: Embodiments of the invention generate metrics quantifying the mobility of a mobile device without persisting information related to the device's specific location at any given time. Specifically, at multiple intervals, a value of a mobility metric is computed based on the distance between the current location of the mobile device and a previously identified origin location of the mobile device. The values of the mobility metric computed over a period of time quantify the overall mobility of the mobile device. The mobility metric does not provide any information regarding the specific location of the mobile device at any given time.

Abstract: A distributed system processing a publisher's streaming data. The distributed system comprises multiple workers and publisher data stores, each publisher data store dedicated to a worker and a publisher. A sampling ratio (the fraction of data items for storage in the publisher's data store) is selected by a publisher data store's worker based on historical information. At least two workers select different sampling ratios. Data items representing an interaction between an entity and the publisher are received. Each data item is assigned to a worker for processing. A hash function is applied to the data item's identifier, resulting in a key value falling within the hash function's range. The scope of the publisher's data store is equal to the hash function's range multiplied by the sampling ratio of the publisher's data store. A data item with a key value within the scope of the publisher's data store is stored therein.