Applications and services on mobile devices often generate notifications. Alerts are likely to be more effective if delivered at times when the user is in contact with the phone when stationary or while moving at a slow pace. This disclosure describes techniques that dynamically adjust the manner and/or time of notification alerts based on the current position of the device relative to the user and/or user activity, determined based on user-permitted data such as device sensor readings. The data is provided as input to a suitably trained machine learning model. The model indicates the likely position of the device in relation to the user, along with the likely user activity. Based on this information, the delivery mode and/or timing is dynamically adjusted. Implementation of the techniques can optimize the delivery of notification alerts, thus increasing the likelihood of catching the user’s attention and enhancing the utility derived from the notification content.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.