Amazon Technologies, Inc.

Oath Inc.

Inscape Data Corporation

Yahoo! Inc.

Yahoo Assets LLC

Verizon Media Inc.

Cognitive Media Networks, Inc.

Tvi Interactive Systems, Inc.

Edo Liberty

Zeev Neumeier

Zohar Karnin

Yoelle Smadja Maarek

Maxim I Sviridenko

Ron Belmarch

Justin Thaler

Jerry Shen

Prerit Uppal

Stefano Stefani

Thomas Albert Faulhaber, Jr

Alexander Johannes Smola

Craig Wiley

Yoelle Maarek Smadja

Swaminathan Sivasubramanian

Aasish Kumar Pappu

Joel Tetreault

Meizhu Liu

Ellie Pavlick

Steven Andrew Loeppky

Liangliang Cao

Guy Halawi

David Wajc

Madhav Jha

Gilad Tsur

Leo Parker Dirac

Andrei Z Broder

Ronny Lempel

Edward Bortnikov

Michal Aharon

Vishwanath Tumkur Ramarao

Fabian Moerchen

Idan Szpektor

Mark Risher

Gowda Dayananda Anjaneyapura Range

Yehuda Arie Koren

Sahika Genc

Taylor Goodhart

Mu Li

Amir Sadoughi

Leyuan Wang

Konstantin Shmakov

Kevin J Lang

Systems and methods for identifying which video segment is being displayed on a screen of a television system. The video segment is identified by deriving data from the television signals, the derived data being indicative of the video segment being displayed on file screen. This feature can be used to extract a viewer's reaction (such as changing the channel) to a specific video segment (such as an advertisement) and reporting the extracted information as metrics. The systems and methods may further provide contextually targeted content to the television system. The contextual targeting is based on not only identification of the video segment being displayed, but also a determination concerning the playing time or offset time of the particular portion of the video segment being currently displayed.

Methods for identifying video segments and displaying contextually targeted content on a connected television

Techniques for intelligent compute resource selection and utilization for machine learning training jobs are described. At least a portion of a machine learning (ML) training job is executed a plurality of times using a plurality of different resource configurations, where each of the plurality of resource configurations includes at least a different type or amount of compute instances. A performance metric is measured for each of the plurality of the executions, and can be used along with a desired performance characteristic to generate a recommended resource configuration for the ML training job. The ML training job is executed using the recommended resource configuration.

Intelligent compute resource selection for machine learning training jobs

In an aspect, fantasy sports data is analyzed to generate and/or manage the delivery of and user experience of fantasy sports games and contests, including timing and payout structure(s) for short-term fantasy contests, such as daily, weekly, etc. fantasy sports contests, that motivate and appeal to contestants and is/are reasonably structured and concisely represented. In yet another aspect, a daily fantasy contest system is provided to automatically adjust to user demand without overshooting profitable levels and offer a variety of different contests, as well as offer contests with guaranteed and non-guaranteed payouts.

Fantasy sports data analysis for game structure development

An electronic message composition support system, method and architecture including machine learning and natural language processing techniques for extending message composition capability and support and to provide feedback to a user regarding an error, condition, etc., detected in the user's message before the user sends the message, e.g., while the user is composing the message using a messaging application's user interface.

Electronic message composition support method and apparatus

Techniques for dynamic accuracy-based experimentation and deployment of machine learning (ML) models are described. Inference traffic flowing to ML models and the accuracy of the models is analyzed and used to ensure that better performing models are executed more often via model selection. A predictive component can evaluate which model is more likely to be accurate for certain input data elements. Ensemble techniques can combine inference results of multiple ML models to aim to achieve a better overall result than any individual model could on its own.

Dynamic accuracy-based deployment and monitoring of machine learning models in provider networks

Techniques for determining and utilizing optimal aggregation schedules are described are described. A deep machine learning model can be trained using multiple processing elements implemented in one or multiple computing devices and that are interconnected using one or multiple types of links. An optimal aggregation schedule for such arbitrary topologies can be determined automatically. The determination may include solving a linear program on the spanning tree polytope. The optimal aggregation schedule can be utilized by the multiple processing elements to train the deep machine learning model.

Optimal message scheduling for aggregation

Techniques for auto-scaling hosted machine learning models for production inference are described. A machine learning model can be deployed in a hosted environment such that the infrastructure supporting the machine learning model scales dynamically with demand so that performance is not impacted. The model can be auto-scaled using reactive techniques or predictive techniques.

Auto-scaling hosted machine learning models for production inference

Disclosed is a system and method for email management that leverages information derived from automatically generated templates in order to identify types of message and message content. The systems and methods discussed herein involve identifying messages matching specific template types and structures, and automatically extracting important data from email messages matching those templates. The extracted data enables improvements for a user's experience and increased monetization. That is, templates can be analyzed to determine a type of email message, which in turn can be presented to a receiving user within an automatic folder or tag designation. Additionally, email snippets or previews can be generated from the extracted data for display within a user's inbox. Also, the extracted data can be used for monetization purposes, by serving targeted advertisements based upon the data extracted from such messages.

System and method for identification of subject line templates

Disclosed are systems and methods for improving interactions with and between computers in content providing, generating and/or hosting systems supported by or configured with personal computing devices, servers and/or platforms. The systems interact to identify and retrieve data within or across platforms, which can be used to improve the security and quality of data used in processing interactions between or among processors in such systems. The disclosed systems and methods provide added security features and functionality to messaging platforms. Message content within communicated or to be communicated messages can be subject to such security functionality through the identification of selected message portions having an identifier applied therewith that not only hides the selected message portions from being viewed within a message interface, but also modifies the message thereby rendering the selected message portion as unreadable and/or inaccessible by a user or computing device without the required security credentials.

Computerized system and method for modifying a message to apply security features to the message's content

Bandwidth-efficient (i.e., compressed) representations of audio data can be utilized for near real-time presentation of the audio on one or more receiving devices. Persons identified as having speech represented in the audio data can have trained speech models provided to the devices. These trained models can be used to classify the compressed audio in order to improve the quality to correspond more closely to the uncompressed version, without experiencing lag that might otherwise be associated with transmission of the uncompressed audio. The uncompressed audio is also received, with potential lag, and is used to further train the speech models in near real time. The ability to utilize the uncompressed audio as it is received prevents a need to store or further transmit the audio data for offline processing, and enables the further trained model to be used during the communication session.

Online training with delayed feedback

Techniques for performing image-augmentation based simulations on are described. An exemplary embodiment of such performances includes for each tuple of timestamped image and movement data, generating a next image using an image generation neural network based on the timestamped image and movement data, the image being input into the image generation neural network as a non-rendered image, and generating a reward using a reward generating neural network based on the timestamped image and movement data.

Training machine learning models for physical agents and robotic controls with simulations

A computer system, computer program product, and computer-implemented method for communicating electronic messages over a communication network coupled thereto are provided. The computer system comprises a network interface for receiving messages sent over the network and addressed to a user of the computer system; and computer executable electronic message processing software. The software comprises instructions for directing the computer system to receive a message over the network, and to identify whether a sender of the received electronic message is a human or a machine. The identifying includes first and second phases of operation. The first phase includes an offline phase employing information and activities resident on the computer system. The second phase includes an online phase employing resources remotely accessible over the network. The software further includes instructions for providing the user with the results of the identification as human or machine; and for performing automatic data extraction from the message if the message was identified to be from a machine.

Method and system for classifying man vs. machine generated e-mail

Growing community of inventors

Or-Yehuda, Israel

Edo Liberty