On the Understanding of Interdependency of Mobile App Usage

With the rising popularity of smartphones and the rapid growth of mobile applications, understanding the app usage behavior of mobile users is of growing importance for both app designers and service providers. Different from previous studies mining the correlation between apps and physical world factors, e.g. location, time, etc., in this paper we focus on the interdependency among apps and try to address a series of research problems: what apps are frequently used together? What are the relations between apps (strengthen or undermine the use of each other) belonged to each category? To answer these questions, we employ the frequent pattern mining algorithm to a large-scale real-world dataset, which includes more than 1.7 million users and 5 billion app usage logs, and find out frequent app-sets and association rules with interesting insights. These results are usefulness in app marketing for service providers and in understanding different mobile users for app designers.