Amazon Technologies, Inc.

Sean Donahoe

Paul Anthony Kotas

David Neil Turner

Oliver Masting

Graham Ritchie

Sebastian Koch

Ramesh Ravipati

Christopher Walton

José Lucas Lemos Mendonça

Systems, methods, and computer-readable media are disclosed for dynamically determining relative product performance using quantitative values. In one embodiment, an example method may include determining, by one or more computer processors coupled to at least one memory, a first product of a product catalog, determining a first set of actual performance values for the first product during a first time period, generating a first expected performance model based at least in part on a first set of product attributes for the first product and the first set of actual performance values, determining a first expected performance value for the first product based at least in part on the first expected performance model and the first set of product attributes, determining a first average actual performance value for the first product over the first time period based at least in part on the first set of actual performance values, and determining a first relative performance value for the first product based at least in part on the first average actual performance value and the first expected performance value.

Dynamically determining relative product performance using quantitative values

Systems, methods, and computer-readable media are disclosed for dynamically determining audience response to presented content using a video feed. In one embodiment, an example method may include receiving video data for users to which content is presented over a time period, generating, using the video data, a set of frames corresponding to a first user, wherein the set of frames includes a first frame corresponding to a first time during the time period and a second frame corresponding to a second time during the time period, determining, using the first frame and the second frame, a first engagement value for the first user, determining, using the first frame, a first emotional classification for the first user at the first time, determining, using the second frame, a second emotional classification for the first user at the second time, and determining first aggregate user response data for the first user.

Dynamically determining audience response to presented content using a video feed

Growing community of inventors

Kirkland, WA, United States of America

Sean Donahoe