Abstract
Notifications on devices such as smartphones, tablets, smartwatches, laptops, etc. provide users a quick view of updates from various applications. Some notification interfaces also include quick action buttons for the user to take a corresponding action in the application that delivers the notification, e.g., mute a chat conversation, accept a calendar invite, etc. However, the suggested actions are limited to a preconfigured set and do not account for user context. This disclosure describes techniques to provide an enhanced notification UI with a wide range of suggested actions that are delivered to the user taking into account the user’s context. A large language model (LLM) is utilized to analyze the notification payload and determine suggested actions and associated applications. Deep links for the suggested actions are generated and included in the user interface. A lightweight, on-device LLM is utilized to filter the notification payload and adjust the suggested actions based on the user’s context.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Habus, Rabia; Allalouf, Tomer; Jonas, Nadav; Berkowitz, Yaniv; Omari, Sari; Segal, Aviram; and Eckman, Brad, "Personalizing Suggested Actions Embedded Within Device Notifications", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/10923