Context-Adaptive Notification Management in Mobile Language Learning

Abstract

This project investigates whether or not context-adaptive notification timing can improve responsiveness to notifications in a mobile language learning app. The app developed for this project has been created to send notifications that contain built-in quizzes. The goal is to provide users with a low-threshold form of learning that is adaptively initiated by the system. The app uses eight independent contextual variables, such as location and time, to make the system context-aware. We compared 30 users using a context-adaptive notification management model to 28 users using a context-unadaptive notification management model over a period of 11 days. We have found that users of the context-adaptive model respond significantly more often to notifications as opposed to the users using the context-unadaptive model.