Abstract:

Mobile phone technology has been completely revolutionized in recent years. From being used for calls and text messages, the usage pattern has now shifted, and involves heavy applications (apps) based on platforms such as Android and iOS. Therefore, people have become exceedingly dependent on smartphones, with most of them suffering from a smartphone addiction disorder. In this study, the authors collect app usage patterns and categorize all apps into five buckets: social, entertainment, utility, gaming, and shopping/food and beverage. The data were collected through a self-developed app, named Activity Tracker, installed in the smartphones of individuals participating in the study. Activity Tracker also collects additional phone usage features such as call duration, number of texts/calls, and screen-on duration. The authors predict smartphone addiction using supervised machine-learning algorithms with good accuracy. This is a promising attempt to uncover multi-dimensional aspects of smartphone addiction by studying the usage patterns of installed Android applications. The study also includes gender-wise variations of smartphone usage patterns and explicit parallelism between usage patterns of addicted and non-addicted users.

Keywords: