Google Inc.

Mahesh Keralapura Manjunatha

Ramanathan V Guha

Vivek Raghunathan

Ramakrishnan Srikant

Kiran S Panesar

Andrew M Dai

Pablo Bellver

Vineet Gupta

Amay Nitin Champaneria

Aparna Chennapragada

Frederick Peter Brewin

Carolyn F Au

Karthik Lakshminarayanan

Carl Lischeske

Ravi Kolluri

Chiu Wah So

Chiu Wah Kelvin So

Radu Jurca

Shen Wang

Xi Xiong

Vinay Vyas Vemuri

Muruo Liu

Jakub Ocwieja

Md Mahbubul Hasan

Daniel Svonava

David Fan

Yao Liu

Andrei Dragus

At least one aspect of the present disclosure is directed to systems and methods of pruning retrieval tokens from sets of retrieval tokens based on criteria. The system can receive a plurality of retrieval tokens including a second retrieval token. The system can retrieve an indication of a first token and a plurality of requests. The system can construct a first bit string based on the predicted requests and the first retrieval token. The system can retrieve a second bit string corresponding to the second retrieval token. The system can compare the first bit string to the second bit string to determine a similarity value. The system can determine the similarity value is greater than a predetermined threshold. The system can remove the first and second retrieval token from the plurality to create a pruned set of retrieval tokens. The system can provide the pruned set to a content.

Content provider recommendations to improve targetting and other settings

At least one aspect of the present disclosure is directed to systems and methods of secure and privacy preserving device classification. A server can maintain a plurality of data records, each including an indication of a request and a known classification value. The server can train a context obfuscation model using each of the plurality of requests and known classification values. The server can train a classification model using resources and category information from a data structure in the memory of the client device. The server can transmit the context obfuscation model to a different plurality of client devices. The server can receive a request for classification including a classification vector and request metadata. The server can determine the classification of the device responsible for the request using the classification model. The server can transmit the device classification to the device responsible for the request.

Systems and methods for secure and privacy preserving device classification

At least one aspect of the present disclosure is directed to systems and methods of pruning retrieval tokens from sets of retrieval tokens based on criteria. The system can receive a plurality of retrieval tokens including a second retrieval token. The system can retrieve an indication of a first token and a plurality of requests. The system can construct a first bit string based on the predicted requests and the first retrieval token. The system can retrieve a second bit string corresponding to the second retrieval token. The system can compare the first bit string to the second bit string to determine a similarity value. The system can determine the similarity value is greater than a predetermined threshold. The system can remove the first and second retrieval token from the plurality to create a pruned set of retrieval tokens. The system can provide the pruned set to a content provider.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for a content platform that receives a request to provide a digital component. The request includes information about users to which the digital component is directed. Futurized queries are obtained from serving logs of a serving system that is configured to execute an existing digital component using serving code that directs digital content to the set of users. The futurized queries are loaded as data structures in memory of a forecasting system. The system uses an instruction set derived from the serving code to determine that similarity between the particular futurized query and the request exceeds a threshold similarity. The system then generates a forecast output as a response to the request based on futurized queries that exceed the threshold similarity. The forecast output includes data describing future provision of the digital component.

Modeling distribution of digital components over a network

Methods and apparatus for providing travel-related information for a location to a user based on activity indications of the user that are related to the location. The location may be determined based on a set of one or more related activity indications and a travel-related score may be determined for the location that is indicative of likelihood that the user has interest in travelling to the location. The user may be provided the travel-related information for the location based on the travel-related score.

Automatic continued search

Growing community of inventors

Sunnyvale, CA, United States of America

Mahesh Keralapura Manjunatha