Amazon Technologies, Inc.

Kirtan Modi

Robert James Victor

Qie Hu Huang

Xinyu Zhou

Amit Gupta

Vykunth Ashok

Shuang Li

Vamshi Krishna Surabhi

Jae In Cho

Yongzhen Lu

Ying Zhang

Jiahui Jeff Wang

Xing Zhao

Xuefei Wang

Devices, systems, and methods are provided for managing advertisement exposure. The method may include receiving, by a first device, from a second device, an audio fingerprint based on presentation of first media content using a first media channel at a first time; determining, using automatic content recognition, that the audio fingerprint represents a portion of a first advertisement; determining that the first advertisement was presented at a second time using a second media channel and the second device; determining that a first number of times that the first advertisement has been presented to the user exceeds a first threshold number and that a second number of times that the first advertisement has been presented to the user does not exceed a second threshold number; refraining from selecting the first advertisement at a third time; and selecting the first advertisement for presentation at a fourth time.

Enhanced cross-media channel deterministic and modeled presentation frequency limits

Techniques for performing a cross-platform media content personalization are described. According to some examples, a computer-implemented method includes receiving an indication, at a content delivery service from a support service for streaming media player devices, that indicates a set of one or more users of the support service that have a threshold similarity of a record of interactive activity with the support service to a record of interactive activity of a target user with the support service; determining, by the content delivery service, a subset of the users, from the set of one or more users of the support service, that have a record of interactive activity with the content delivery service; determining, by a content recommendation service of the content delivery service, a media file to provide to the target user based on the record of interactive activity with the content delivery service of the subset of the users; and presenting, by a streaming media player device of the target user and that is supported by the support service, the media file to the target user.

Computer-implemented methods for cross-platform knowledge transfer for video content personalization

Systems, devices, and methods are described herein for improving inventory management. As used herein, 'inventory' refers to digital space at an inventory providers webpage at which content can be delivered. The disclosed techniques utilize reinforced machine learning and an offline training process to train various models with which a content request corresponding to the inventory can be classified according to historical requests and a selection process identified for the request (e.g., a direct or an indirect selection process). If an indirect selection process is chosen, the content request may be optimized for that process utilizing additional machine learning models trained using reinforced machine learning and the offline training process. The disclosed techniques enable the inventory provider to optimize content selections according to a preferred objective. The training operations are performed offline, in a training system configured to simulate the run time environment.

Simulation augmented reinforcement learning for real-time content selection

Devices, systems, and methods are provided for on-target rate optimization for video. A method may include receiving streaming video advertisement impression data; receiving user activity data indicative of day-parts when viewers watch content; generating, based on the streaming video advertisement impression data and the survey data, using a machine learning model, a demographic probability vector, wherein each entry of the demographic probability vector is indicative of a probability that a viewer is in a respective age range of the non-overlapping demographic groups; generating, using the machine learning model, an audience recognition model with the demographic probability vector; generating a synthetic audience model predicting future advertisement viewing behavior; generating an assignment of an advertisement bid to a respective demographic group of the non-overlapping demographic groups; and generating, based on the assignment, a list of target demographic groups of the non-overlapping demographic groups for a bid request associated with the advertisement bid.

Enhanced on-target rate optimization for video using machine learning

Devices, systems, and methods are provided for audience recognition. A method may include receiving over-the-top (OTT) advertisement impression data comprising metadata and content of advertisement bid requests, the metadata indicative of scheduled OTT media presentation; receiving user activity data indicative of day-part times when viewers watch content absent from the OTT advertisement impression data; generating, based on the OTT advertisement impression data, a first demographic probability vector; generating, based on the user activity data, a second demographic probability vector; generating, based on a combination of the first demographic probability vector and the second demographic probability vector, a third demographic probability vector, each entry of the third demographic probability vector indicative of a third probability that a viewer is in a respective age range; and generating an indication of the third demographic probability vector.

Enhanced recognition of content audiences

Growing community of inventors

Seattle, WA, United States of America

Kirtan Modi