Google Inc.

Other

Thomas Price

Tuna Toksoz

Justin Lewis

Anurag Agrawal

James C Roberts

Sarvjeet Singh

Prachi Gupta

Deepak Ramamurthi Sivaramapuram Chandrasekaran

Bo Fu

Sue Yi Chew

Kunal Jain

Jierui Xie

Systems and methods are described for providing an interface and facilitating selection of sponsored content networks that provide sponsored content items. This may include providing, by a mediation server, a user interface to a publisher server, the user interface configured to provide access to data stored on the mediation server; receiving a metric associated with a rule for filtering content items associated with the publisher; applying the metric to a content network list associated with the publisher using the user interface to generate an updated content network list; and transmitting mediation code including the updated content network list to the publisher server, wherein the mediation code, when executed by a user device, (i) causes the user device to control display of content items according to the updated content network list and (ii) allows the user device to flag at least one content item for modifying the updated content network list.

Publisher tool for controlling sponsored content quality across mediation platforms

Balancing content distribution between a machine learning model and a statistical model provides a baseline assurance in combination with the benefits of a well-trained machine learning model for content selection. In some implementations, a server receiving requests for a content item assigns a first proportion of the received requests to a first group and assigns remaining requests to a second group. The server uses a machine learning model to select variations of the requested content item for responding to requests assigned to the first group and uses a statistical model to select content variations for requests assigned to the second group. The server obtains performance information, e.g., acceptance rates for the different variations, and compares performance of the different models used for content selection. Audience share assigned to the machine learning model is increased when it outperforms the statistical model and decreased when it underperforms the statistical model.

Assessing accuracy of a machine learning model

Training data for a machine learning model is generated. Generating the training data includes generating first training input, the first training input including first contextual information associated with a first user device of multiple user devices associated with a user account. The first contextual information identifies first location information corresponding to the first user device. A first target output for the first training input is generated. The first target output identifies an indication of a preference associated with the user account to cancel a first transmission of one or more media items to the first user device. The training data is provided to train the machine learning model on (i) a set of training inputs including the first training input, and (ii) a set of target outputs including the first target output.

Using machine learning and other models to determine a user preference to cancel a stream or download

Systems and methods are described for providing an interface and implementing rules and metrics received from the interface regarding the selection of sponsored content networks that provide sponsored content items. This may include providing mediation code to a publisher for inclusion in publisher content provided to a user device, the mediation code associated with a table of sponsored content networks, receiving an image of a sponsored content item and a sponsored content network identifier from the user device, analyzing the image of the sponsored content item, the analysis generating extracted image data from the sponsored content item, categorizing the sponsored content item based on the extracted image data and the sponsored content network identifier, receiving an instruction to filter the sponsored content networks exceeding a metric based on a category, and updating the table of sponsored content networks to remove a particular sponsored content network.

Systems and methods for dynamic sequence-based adjustment of prompt generation are provided. The system can receive a first interaction and a second interaction via a client device and identify a first sequence based on the first interaction and the second interaction. The system can map the first sequence to a node data structure and identify a node in the node data structure that matches the first sequence. The system can generate an adjusted parameter for a first digital component object responsive to a match with an attribute of the node in the node data structure. The system can execute a real-time digital component selection process among a plurality of digital component objects including the first digital component object to select the first digital component object. The system can transmit a prompt with the first digital component object to a client device to cause the client device to present the prompt.

Dynamic sequence-based adjustment of prompt generation

A system and method are disclosed for training a machine learning model using information pertaining to transmissions of one or more media items to user devices associated with a user account. Generating training data for the machine learning model includes generating first contextual information associated with a first user device and generating a first target output that identifies an indication of a preference of a user preference to cancel the first transmission. The method includes providing the training data to train the machine learning model.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for reducing latency in presenting content. In one aspect, a system includes a native application that presents an interactive item and a latency reduction engine. The latency reduction engine detects interaction with the interactive item that links to a first electronic resource that is different from the native application and provided by a first network domain and in response to the detecting, reduces latency in presenting the first electronic resource, including executing a first processing thread and a second processing thread in parallel. The first processing thread requests a second electronic resource from a second network domain and loads the second electronic resource and, in response to the loading, stores a browser cookie for the second network domain. The second processing thread requests the first electronic resource and presents the first electronic resource.

Reducing latency in downloading electronic resources using multiple threads

Systems and methods of reducing latency in online network traffic are provided. The system can include a processor and a memory couple to the processor. The memory stores computer-executable instructions, which when executed by the processor, cause the processor to identify an indication of an interaction with a content item presented within a content slot of an information resource, generate a click identifier that uniquely identifies the interaction, transmit a first request comprising a click resource link including the click identifier to a click server, and transmit a second request to access the landing page of the content item independent of transmitting the first request. The second request comprises a generated resource link that identifies a resource identifier of the landing page of the content item and the click identifier.

Systems and methods for latency reduction in content item interactions using client-generated click identifiers

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for persisting state of a streaming application are disclosed. In one aspect, a method includes the actions of receiving data indicating interaction with third-party content that is displayed with first-party content at a client device. The interaction with the third-party content initiates an application request for a streaming version of an application ('streaming application'). The actions further include generating a representation of the streaming application in response to the interaction with the third-party content at the client device. The actions further include receiving data indicating a user interaction with the representation of the streaming application at the client device. The actions further include generating and storing data indicating a first state of the streaming application at a first time based on the user interaction with the representation of the streaming application.

Persisting state of a streaming application

A method includes receiving, at a user device, a plurality of signals pushed to the user device from a plurality of store devices located in a plurality of physical storefronts. Each signal includes a store identifier identifying the associated physical storefront. Store visit data is generated for each signal, the store visit data including the store identifiers and time data representing times associated with visits to the physical storefronts. Store visit data is aggregated for the plurality of signals pushed to the user device. Aggregated store visit data is transmitted to an analytics server configured to combine aggregated store visit data from a plurality of user devices and generate analytics data from the combined aggregated store visit data. At least a portion of the time data is removed from the aggregated store visit data before transmission to the analytics server.

Store visit data creation and management

A system and method are disclosed for training a machine learning model using information pertaining to transmissions of one or more media items to user devices associated with a user account. In an implementation, training data for the machine learning model includes first contextual information associated with a first user device and second contextual information associated with a second user device. The number of the transmissions to the user devices for the user account exceeds a threshold number of transmissions allowed for the user account. Training data further includes generating a first target output that identifies an indication of a preference of a user preference to keep or cancel each of the transmission. The method includes providing the training data to train the machine learning model. The trained machine learning model may be used to determine which of the new transmissions is to be canceled.

Systems and methods for maintaining session identifiers across multiple webpages for content selection are described herein. A server can generate a session identifier corresponding to the client device for use in selecting content to be served across the information resources of a content publisher. The server can then receive, from the client device, a request for instructions to insert into a header of an information resource for selecting a content provider. The server can determine that the request includes the session identifier that matches the one generated for a previous request. The server can obtain a parameter value for each content provider with the session identifier. The server can select a content provider using the parameter values. The server can provide a script to the client device for the header. The script can be configured to cause the client device to obtain a content item from the selected content provider to insert into the information resource.

Maintaining session identifiers across multiple webpages for content selection

Systems and methods are described for improving mediated sponsored content networks using a stack trace. One implementation relates to a method running on a sponsored content network and includes receiving a request for a sponsored content item from a user device executing code from a mediation platform. An item is sent to fulfill the request. If the identity of the mediation platform is not known, the sent item includes stack trace implementation code. An identifier is received that is associated with the mediation platform consequent to the user device executing the stack trace implementation code and the identity of the mediation platform is determined using the identifier. A performance metric associated with the determined identity of the mediation platform is identified. The method may further receive a second request for a sponsored content item from the user device and use the performance metric to select the second item.

Systems and method for improving mediated sponsored content networks using a stack trace

A client system presents, within an execution environment of an application, a third-party media stream distinct from the application, received from a remote host server via a network. The client system detects interaction events during presentation of the third-party media stream, and transmits descriptions of the detected interaction events to the remote host server. The application may be pre-cued prior to presentation, e.g., to minimize start-up time. In some implementations, a side-band message channel is established to facilitate communication between the client system and the remote host server.

Growing community of inventors

San Francisco, CA, United States of America

Thomas Price