Amazon Technologies, Inc.

Roopesh Ranjan

Pragyana K Mishra

Nikolaos Chatzipanagiotis

Graton Marshal Randal Gathright

Neela Sawant

Pooja Ashok Kumar

Cagri Ozcaglar

Vishnu Parimi

A prediction model may be created to predict future actions likely to be performed by users while interacting with electronic content via user devices. The predictions may be used to streamline access to interface controls or other information to enable the users to facilitate or expedite performance of the predicted actions, while reducing computational demands on computing devices that provide the electronic content by, for example, reducing unnecessary intervening computing actions.

Predictive action modeling to streamline user interface

Systems are provided to select targets for communications, designed to cause or provoke the targets to perform a desired action, using bias-corrected models that predict the effect of such communications on the targets. Data regarding previous actions taken by potential targets for communications&#x2014;also referred to as 'candidates'&#x2014;can be analyzed to determine whether there is a difference in actions taken by candidates who have received prior communications in comparison with candidates who have not received the prior communications. Biases in the selection of candidates to receive the prior communications can be corrected by weighting the data associated with the selected candidates to more closely match the distribution of candidates not selected to receive the prior communications.

Uplift modeling with importance weighting

Systems and methods are provided for determining or predicting user demographic information using user behaviors through a Bayesian inference. A computing system may determine demographic information (such as age or gender) of a user based on a Bayesian update and a purchase or other user action by the user. In some embodiments, the computing system may determine the household composition of a user account based on multiple purchases by the user account. The computing system may generate recommendations for the user or the user account based on the demographic information of the user or the household composition of the user account.

Inferring user demographics from user behavior using Bayesian inference

Systems and methods are described for attributing credit to individual channels for messages delivered via multi-channel communications. A message may be delivered to a recipient via multiple delivery channels, and the recipient may engage in a conversion event or activity associated with receipt of the message. Each delivery of the message may partially contribute to causing the conversion event, and the incremental contribution of each delivery may be determined. A probability of conversion may be determined based on past message deliveries involving the same or similar messages, recipients, and channels. An impressions path may be generated based on the past message deliveries, and subpaths of the impressions path may be used to isolate the contribution of individual message impressions.

Attribution of multi-channel communications

Systems and methods are described for selecting recipients, channels, and messages for delivery via multi-channel communications. A message may be delivered to a recipient via multiple delivery channels, and the recipient may engage in a conversion event or activity associated with the message. Each potential delivery of the message may be associated with an incremental probability of causing the conversion event. The incremental probabilities may be used to determine the channel and recipient with the highest incremental probability of causing a conversion event for a given message, the recipient and message having the highest incremental probability of causing a conversion event when delivered via a given channel, and other combinations. Profiles may be used to compare messages, channels, and recipients in order to predict incremental probabilities, and messaging resources may be allocated according to budgets, cost-benefit analyses, or other criteria.

Growing community of inventors

Redmond, WA, United States of America

Roopesh Ranjan